• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Microbes Infect. Author manuscript; available in PMC Sep 24, 2007.
Published in final edited form as:
PMCID: PMC1993839

Outer membrane protein A of Escherichia coli K1 selectively enhances the expression of intercellular adhesion molecule-1 in brain microvascular endothelial cells


Escherichia coli K1 meningitis is a serious central nervous system disease with unchanged mortality and morbidity rates for last few decades. Intercellular adhesion molecule 1 (ICAM-1) is a cell adhesion molecule involved in leukocyte trafficking toward inflammatory stimuli at the vascular endothelium; however, the effect of E. coli invasion of endothelial cells on the expression of ICAM-1 is not known. We demonstrate here that E. coli K1 invasion of human brain microvascular endothelial cells (HBMEC) selectively up-regulates the expression of ICAM-1, which occurs only in HBMEC invaded by the bacteria. The interaction of outer membrane protein A (OmpA) of E. coli with its receptor, Ecgp, on HBMEC was critical for the up-regulation of ICAM-1 and was depend on PKC-α and PI3-kinase signaling. Of note, the E. coli-induced up-regulation of ICAM-1 was not due to the cytokines secreted by HBMEC upon bacterial infection. Activation of NF-κB was required for E. coli mediated expression of ICAM-1, which was significantly inhibited by over-expressing the dominant negative forms of PKC-α and p85 subunit of PI3-kinase. The increased expression of ICAM-1 also enhanced the binding of THP-1 cells to HBMEC. Taken together, these data suggest that localized increase in ICAM-1 expression in HBMEC invaded by E. coli requires a novel interaction between OmpA and its receptor, Ecgp.

Keywords: Meningitis, Brain microvascular endothelial cells, Intercellular adhesion molecule-1, Cell adhesion molecules

1. Introduction

Intercellular adhesion molecule-1 (ICAM-1 or CD54), a member of the immunoglobulin supergene family, plays an important role in the immune and inflammatory response [1]. ICAM-1 is a 55-kDa protein that can be expressed on the cell surface or secreted extracellularly by many cell types [2]. The expression of ICAM-1 is strongly up-regulated by several inflammatory mediators, such as cytokines (interleukin (IL)-1β, tumor necrosis factor (TNF)-α, and interferon-γ), oxidative stress, and sheer stress [3]. The primary role of ICAM-1 is the facilitation of the extravasation of leukocytes to migrate towards the sites of infection. In addition, ICAM-1 serves as a receptor for rhinoviruses and is up-regulated by a variety of microorganisms such as Chlamydia pneumoniae, Mycobacterium tuberculosis, and Haemophilus influenzae [46]. Several studies have shown that transcription factor NF-κB is the key regulator of ICAM-1 gene expression in endothelial cells [1]. NF-κB comprises of a family of closely related transcription factors that play a key role in the expression of genes involved in inflammation and immune response [7]. NF-κB is a dimeric transcription factor composed of homo- or heterodimers of the Rel family of proteins, of which p50/p65 is the most frequent dimer and hence considered the prototype. NF-κB is usually present in an inactive form in the cytoplasm bound to a member of the IκB family, generally IκB-α. In this complex, IκB blocks the nuclear translocation signal of NF-κB, and the NF-κB–IκB complex needs to be disrupted for NF-κB to become activated [8]. Specific phosphorylation of serine 32 and 36 on IκB-α by IκB kinase (IKK) triggers the disruption of NF-κB–IκB complex, which renders IκB for ubiquitination and subsequent degradation by the proteosomal complex, thereby releasing NF-κB from IκB and allowing nuclear translocation [911].

Neonatal meningitis is a serious central nervous system disease and Escherichia coli K1, a gram-negative bacterium, is the leading cause of this disease [12]. A surge in E. coli K1 infections has been reported in recent years surpassing the infections caused by other meningitis associated pathogens such as Group B Streptococcus and Listeria monocytogenes [13]. The disease is fatal in 5–40% of infected neonates and cause neurological sequelae in up to 30% of survivors, which include deafness, cortical blindness, and mental retardation [12]. Incomplete understanding of the pathogenesis of the disease is attributed to the unchanged mortality and morbidity rates associated with neonatal meningitis for last few decades despite the availability of active antimicrobial agents. For example, the role of bacterial invasion of blood-brain barrier on the expression of cell adhesion molecules is not clearly known. Due to the presence of cytokines in the circulation, it has been speculated that ICAM-1 expression on brain microvascular endothelial cells (BMEC), which constitute the lining of the blood–brain barrier (BBB), might be up-regulated during these bacterial infection [14]. For most Gram-negative organisms, lipopolysaccharide (LPS) serves as a potent pro-inflammatory mediator. One target for LPS to stimulate inflammatory response is the vascular endothelium. LPS, both directly and through the induction of host cytokines, causes endothelial cells to up-regulate expression of adhesion molecules for the recruitment of leukocytes. Although cytokine induced expression of cell adhesion molecules has been well studied, the expression of these molecules and the induction of cytokine profile in HBMEC by E. coli infection is poorly understood.

Our studies have shown that the expression of outer membrane protein A (OmpA) of E. coli K1 is critical for the invasion of cultured human BMEC (HBMEC) [15]. Concurrent to this observation, OmpA+ E. coli traverses across the blood brain barrier very efficiently in our newborn rat model of hemotogenous meningitis compared to OmpA− E. coli despite the presence of equal number of bacteria in the circulation [16]. Subsequently we identified that OmpA interacts with Ecgp, a gp96-like receptor molecule on the surface of HBMEC [17]. Of note, due to the selective expression of Ecgp on HBMEC but not on non-brain endothelial cells, the OmpA-mediated E. coli invasion is significantly greater in HBMEC when compared to the invasion of non-brain endothelial cells [17]. The OmpA-mediated entry of E. coli into HBMEC increases the activation of PKC-α, which subsequently interacts with caveolin-1, a protein present in caveolae, for the uptake of the bacteria [18,19]. Thus, overexpression of either dominant negative-PKC-α or -caveolin-1 significantly inhibited the entry of E. coli into HBMEC. The internalization of E. coli into HBMEC also induces the activation of FAK, PLC-γ, and PI3-kinase and over-expression of dominant negative forms of these molecules significantly blocked the invasion [2022]. In contrast, over-expression of dominant negative-Src kinase had no effect on the invasion (unpublished results). In this study we sought to evaluate the effect OmpA+ E. coli entry of HBMEC on the expression of cell adhesion molecules. A better understanding of E. coli K1 pathogenesis may lead to novel treatment strategies, thereby significantly reducing the morbidity associated with this disease.

2. Materials and methods

2.1. Bacterial strains and culture conditions

Escherichia coli K1 (serotype O18:K1:H7) strain, RS218, was isolated from the cerebrospinal fluid of a neonate with meningitis and invades HBMEC in a cell culture model [15]. OmpA− E. coli is a non-invasive mutant lacking the entire OmpA gene. pOmpA+ E. coli is an OmpA-complemented OmpA− E. coli and pOmpA− E. coli is a vector control. All bacteria were grown in brain heart infusion broth with appropriate antibiotics as necessary.

2.2. Antibodies and reagents

Antibodies to ICAM-1, VCAM-1 and β-actin were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Anti-E-selectin antibody was obtained from Biodesign (Saco, ME). Antibodies to IκB and phospho-IκB were obtained from Cell Signaling (Beverly, MA). Cy-2- and Cy-3-conjugated secondary antibodies were obtained from Rockland Immunochemicals (Gilbertsville, PA). Normal goat serum and the Vectashield mounting medium with 4,6-diamino-phenylindole were obtained from Vector Laboratories Inc. (Burlingame, CA). Riboquant multiprobe ribosomal protection assay system was obtained from Pharmingen (San Diego, CA). Supersignal chemiluminescence reagent was obtained from Pierce (Rockford, IL). T4 polynucleotide kinase, TriZOL reagent and Lipofectamine 2000 were obtained from Invitrogen (Carlsbad, CA). Phorbol myristate acetate (PMA), poly dI-dC, and all other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO).

2.3. E. coli invasion assays and transfections

The isolation of HBMEC, maintenance of the cells, and the gentamycin protection assays were described previously [15]. Primary cells and cells transfected with SV-40 large T-antigen (THBMEC) were used between passages 12 and 16 for all experiments. Both primary and transfected HBMEC behave similarly in bacterial invasion assays [23,24]. Primary HBMEC were transfected with mammalian expression vectors using Lipofectamine 2000 according to the manufacturer’s instructions. The dominant negative constructs of PKC-α, p85 and c-Src were previously described [18,22].

2.4. Semi-quantitative PCR analysis

The relative abundance of mRNA encoding ICAM-1, VCAM-1 and E-selectin were quantified by RT–PCR. The primers used were as follows: 5′-CAGTGACCATCTA CAGCTTTCCGG-3′ and 5′-GCTGCTACCACAGTGAATGA TGACAA-3′ for ICAM-1; 5′-GATACAACCGTCTTGGTCA GCCC-3′ and 5′-CGCATCCTTCAACTGGCCTT-3′ for VCA M-1; 5′-AGAAATATGTGGTTTCCACGATGA-3′ and 5′-AA ACTGGAGATTGGTTTCCAATTG-3′ for E-selectin.

2.5. Preparation of cell lysates, Western blot analysis, immunocytochemistry, and ELISA

Confluent HBMEC cultures in 100 mm dishes were incubated with either OmpA+ or OmpA− E. coli (MOI of 1:100, cell: bacteria ratio) for 1.5 h at 37 °C in experimental medium without antibiotics. The cells were then washed with RPMI and further incubated with experimental medium containing 20 μg/ml of gentamycin for varying periods. After indicated periods of incubation, cell lysates were prepared and subjected to SDS–PAGE followed by Western blotting. Immunocytochemistry of infected monolayers of HBMEC grown in eight-well chamber slides was performed by differential staining as previously described [18]. Briefly, HBMEC monolayers were incubated with OmpA+ or OmpA− E. coli for 1.5 h as described. The monolayers were then washed and further incubated in medium containing 30 μg/ml gentamycin to prevent the growth of extracellular bacteria for 8 and 12 h. The eight-well chamber slides were washed with cold RPMI and incubated with anti-S-fimbria antibody for 1 h at 4 °C. The monolayers were further washed and incubated with excess amounts of secondary antibody coupled to horseradish peroxidase for 1 h. The chamber slides were washed, brought to room temperature, and fixed. The invaded OmpA+ E. coli were then stained with the same antibody after permeabilization of the monolayers with 1% Triton X-100, followed by secondary antibody coupled to Cy3. The slides were mounted in Vectashield anti-fade solution containing 4,6-diamidino-2-phenylindole. Cells were viewed with a Leica DMRA microscope (Wetzlar, Germany). The levels of E-selectin expression were assessed by ELISA using HBMEC plated in 96 well plates as described previously [25].

2.6. Enzyme mobility shift assay (EMSA) and ribonuclease protection assays (RPA)

Nuclear extracts were prepared from cells treated with bacteria for different periods and were analyzed by EMSA, essentially as described [26]. RPA analysis was performed using a Riboquant multiprobe system with the total RNA isolated from infected HBMEC. In brief, the isolated RNA (10 μg) was hybridized with 32P-labeled probes overnight at 56 °C followed by RNase digestion, according to the manufacturer’s instructions. After digestion, the protected fragments were resolved on a 5% denaturing polyacrylamide gel, dried and later exposed to X-ray film. The intensity of bands corresponding to TNF-α, MCP-1, IL-8 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA was analyzed using ImageJ software.

2.7. THP-1 binding assay

Binding of THP-1 monocytes to activated endothelial cells was performed as described previously [27]. HBMEC in eight-well chamber slides (Nunc Inc., Naperville, IL) were treated with E. coli for 12 h, washed and 500 μl fresh experimental media containing 2 × 105 THP-1 monocytes expressing GFP were added to each well. After 30 min incubation at room temperature with continuous rocking to allow binding, the top portion of the chambers and the gasket were removed. The slide was then dipped twice in HEPES-buffered saline to remove unbound cells and mounted. Slides were visualized under both fluorescence and light microscope and the average number of cells bound in 10 fields in triplicate wells was determined.

2.8. Determination of cytokine levels

HBMEC in 24-well culture plates were infected with bacteria for different periods and the supernatants were collected, centrifuged to remove bacteria, and then stored at −80 °C until further use. Levels of TNF-α and IL-8 in the clarified supernatants were assayed using a specific DuoSet enzyme-linked immunosorbent assay (ELISA) development system (R&D Systems, Minne apolis, MN), according to the manufacturer’s instructions.

3. Results

3.1. Selective up-regulation of ICAM-1 expression in HBMEC infected with OmpA+ E. coli

The endothelium expresses multiple leukocyte adhesion molecules when activated by inflammatory mediators [14]. This phenomenon promotes transmigration of leukocytes, which in conjunction with other host mediators cause neuronal death. However, it is not clear whether the bacteria themselves can induce the expression of cell adhesion molecules on endothelial cells. Therefore, the effect of E. coli invasion of HBMEC on the expression of ICAM-1, VCAM-1, and E-selectin was examined. The total cell lysates of HBMEC monolayers infected with either OmpA+ or OmpA− E. coli were subjected to Western blotting using anti-ICAM-1 antibody. HBMEC infected with OmpA+ E. coli showed an increase in ICAM-1 expression 4 h post-infection with a peak at 8 h well above background (Fig. 1A). OmpA− E. coli infected cells also showed a slight increase in the expression of ICAM-1 at 4 h when compared to control but without further increase at later time points. PMA, which robustly induced the expression of ICAM-1 at all time points of incubation, was used as a positive control. Densitometric analysis of the ICAM-1 bands normalized to β-actin from three individual experiments showed that ICAM-1 expression induced by OmpA+ E. coli was ten-fold higher than control cells and two fold that of induced by OmpA− E. coli (Fig. 1B, P < 0.05 by Student’s t-test). VCAM-1 expression was also up-regulated, however, similarly by both OmpA+ and OmpA− E. coli (Fig. 1C). Of note, PMA did not cause up-regulation of VCAM-1 in HBMEC beyond the control levels. The expression of another cell adhesion molecule, E-selectin, was assessed by cell-based ELISA as the anti-E-selectin antibody does not recognize the denatured molecule. E-selectin expression in cells inoculated with either OmpA+ or OmpA− E. coli showed no significant differences at any time point (Fig. 1D). PMA also showed similar increase in the expression of E-selectin.

Fig. 1
Increased expression of ICAM-1 in HBMEC infected with OmpA+ E. coli. Confluent HBMEC monolayers were treated with either OmpA+ or OmpA− E. coli or PMA for indicated periods. Total cell lysates were prepared, separated by SDS polyacrylamide gel ...

We have previously demonstrated that inhibitors of actin polymerization, PKC-α, or PI3-kinase significantly reduced E. coli invasion of HBMEC [18,22]. Therefore, to examine whether invasion of HBMEC by E. coli is necessary for the up-regulation of ICAM-1, confluent monolayers of HBMEC were treated with cytochalasin D (an actin polymerization inhibitor) or a cell permeable PKC-α inhibitory peptide or Wortmannin (an inhibitor of PI3-kinase) prior to performing invasion assays. An inhibitor of Src kinase, PP1, which had no effect on the invasion of E. coli, was also used as a control. Interestingly, inhibition of invasion by cytochalasin D or PKC-α inhibitory peptide or Wortmannin significantly reduced the up-regulation of ICAM-1 by OmpA+ E. coli, whereas PP1 did not show any effect (Fig. 1E), suggesting that the invasion of E. coli is important for increased expression of ICAM-1. Our previous studies also demonstrated that OmpA+ E coli invades only a certain population of HBMEC [17]. To examine whether the ICAM-1 expression occurs only in those cells infected by the bacteria, immunocytochemistry of infected HBMEC was carried out. The intracellular bacteria were stained in the experiments infected with OmpA+ E. coli, whereas total cell-associated bacteria were stained in OmpA− E. coli infected HBMEC (since no invasion occurs with OmpA− E. coli) with anti-S-fimbria antibody by differential staining. As shown in Fig. 2, only HBMEC infected with OmpA+ E. coli expressed higher levels of ICAM-1 when compared to the non-infected surrounding HBMEC (Fig. 2C,D), whereas ICAM-1 immunostaining was slightly increased in HBMEC infected with OmpA− E. coli (Fig. 2E,F). These data indicate that OmpA+ E. coli selectively up-regulates the expression of ICAM-1 in HBMEC only invaded by the bacteria and significantly greater than OmpA− E. coli induced ICAM-1 expression.

Fig. 2
Immunocytochemistry of ICAM-1 expression on HBMEC infected with E. coli. Confluent monolayers of HBMEC were either uninfected (A, B), infected with OmpA+ E. coli (C, D), or OmpA− E. coli (E, F) and then subjected to differential staining as described ...

3.2. OmpA expression in E. coli and its interaction with the HBMEC-specific gp96-like receptor, Ecgp, is critical for up-regulation of ICAM-1

Our previous studies have established that the expression of OmpA on the surface of E. coli K1 is one of the critical features to invade HBMEC by interacting with a specific receptor, Ecgp [17]. Therefore, first, to determine whether the presence of OmpA is necessary for ICAM-1 expression in HBMEC, OmpA− E. coli was complemented with a plasmid containing ompA gene. The strain, designated as pOmpA+ E. coli, expresses similar amounts of OmpA as the wild-type bacteria, as shown by dot-blot analysis [16]. Infection of HBMEC with pOmpA+ E. coli exhibited an increase in ICAM-1 expression similar to the levels observed in HBMEC infected with wild-type E. coli and two-fold greater than the pOmpA− E. coli infected cells (Fig. 3A,B). In contrast, the VCAM-1 expression is similar in HBMEC treated with either pOmpA+ or pOmpA− E. coli. Of note, invasion assays performed in the absence of serum also revealed up-regulation of ICAM-1 by OmpA+ E. coli suggesting that LPS may not be playing a major role (data not shown). Additionally, to demonstrate that OmpA interaction with Ecgp is critical, either the bacteria or HBMEC were pre-incubated with antibodies to OmpA or Ecgp, respectively for 1 h prior to infection. As expected, both antibodies reduced the up-regulation of ICAM-1 to background levels, whereas control isotype matched antibodies had no effect (Fig. 3C). Pretreatment of HBMEC with anti-Ecgp antibody alone or in the presence of PMA did not induce the up-regulation of ICAM-1. These data suggest that the anti-Ecgp antibody is a blocking antibody and that the signaling pathways involved in OmpA-Ecgp mediated ICAM-1 expression might be different than those induced by PMA. Taken together, these results demonstrate that OmpA-Ecgp interaction is involved for the selective up-regulation of ICAM-1.

Fig. 3
OmpA interaction with its cellular receptor, Ecgp, is involved for E. coli K1 induced ICAM-1 expression. (A) HBMEC were infected with OmpA+ E. coli, OmpA− E. coli, OmpA− E. coli transformed with a plasmid containing ompA gene (pOmpA+) ...

3.3. Cytokines produced by HBMEC upon infection with OmpA+ E. coli are not responsible for the up-regulation of ICAM-1

Cytokines have been shown to elicit multiple biological functions during the inflammatory response including ICAM-1 expression on endothelial cells [6,28]. Therefore, the cytokine and chemokine profile induced by E. coli infection of HBMEC was examined by RPA using RNA isolated from cells infected with E. coli. A commercially available riboquant probe set for the inflammatory genes TNF-α, IL-1β, IL-6, MIP-1β, IL-8, MCP-1, and RANTES was used. The experiments were carried out at least three times and a representative picture is shown in Fig. 3D. OmpA+ E. coli infected HBMEC showed a significant increase in the mRNA levels of TNF-α, MCP-1 and IL-8 over time when compared to the uninfected controls. No increase of other cytokines and chemokines was observed. Of note, no significant differences in the mRNA levels of cytokines and chemokines were observed between OmpA+ and OmpA− E. coli, as determined by densitometric analysis of the blot. The supernatants of E. coli infected HBMEC after different periods of infection were collected and subjected to a sandwich ELISA to estimate the amount of cytokines secreted (TNF-α and IL-8). OmpA+ E. coli induced increased production of TNF-α (15–45 pg/ml) with time when compared to uninfected controls. Similar quantities of TNF-α were also observed with OmpA− E. coli. Both OmpA+ and OmpA− E. coli induced the secretion of IL-8 with similar peak levels at 16 h (~400 pg/ml). Since we did not find any significant difference in the cytokine secretion profile between OmpA+ and OmpA− E. coli, it is assumed that the secretion of TNF-α and IL-8 by infected HBMEC does not contribute to the difference in ICAM-1 expression between the two strains.

3.4. OmpA+ E. coli induces ICAM-1 expression in HBMEC by activating NF-κB

To examine the role of NF-κB in E. coli-mediated increase in ICAM-1 expression, nuclear extracts were prepared from E. coli infected cells and were subjected to EMSA. As expected, OmpA+ E. coli induced a 2.5-fold increase in NF-κB activity by two-hour post-infection when compared to untreated cells while OmpA− E. coli induced NF-κB activity was 1.5-fold lower than that of the OmpA+ E. coli (Fig. 4A). To further characterize the NF-κB subunits involved, we added antibodies against the p50 and p65 NF-κB subunits to the reaction mixture before mixing with the labeled probe. Addition of both antibodies resulted in super shift of the NF-κB bands, indicating that both p50 and p65 subunits are involved (Fig. 4A).

Fig. 4
OmpA+ E. coli induces ICAM-1 expression in HBMEC by activating NF-κB. (A) HBMEC were infected with E. coli for different time periods, nuclear extracts were prepared, and subjected to EMSA using radiolabeled probe for NF-κB. In certain ...

Phosphorylation of inhibitory kappa B (IκB) by IκB-kinase results in its ubiquitination and subsequent degradation, and allowing the translocation of NF-κB to the nucleus [11]. Hence, IκB phosphorylation and degradation in the lysates of HBMEC infected with E. coli were also analyzed. Fig. 4B shows that IκB phosphorylation increases in OmpA+ E. coli infected HBMEC, which correlated with greater IκB degradation with time. In contrast, OmpA− E. coli infection did not result in major changes in either IκB phosphorylation or degradation. Densitometric analysis of phospho-IκB and IκB bands on the blots revealed a two-fold increase in the phosphorylation and degradation of IκB upon incubation of HBMEC with OmpA+ E. coli (Fig. 4C). To determine the upstream signaling components in OmpA+ E. coli induced NF-κB activation, we analyzed ICAM-1 protein expression in HBMEC transiently transfected with dominant negative NIK, IKK-α and IKK-β expressing plasmids. The transfected cells were treated with E. coli for different periods and the lysates were subjected to Western blotting for ICAM-1. As shown in Fig. 4D, the expression of ICAM-1 was significantly reduced in HBMEC expressing dominant negative (DN) IKK-β. In contrast, neither DN IKK-α nor DN NIK showed any effect on the expression of ICAM-1 induced by OmpA+ E. coli, suggesting that IKK-β plays a major role in this process. Since equal number of OmpA+ or OmpA− E. coli was associated with the cells tested, the differences observed in NF-κB or IκB activation is not due to unequal binding of bacteria. Altogether, these experiments further indicate that NF-κB plays a role in E. coli- induced ICAM-1 expression in HBMEC.

3.5. PKC-α and P13-K dependent but Src-independent induction of ICAM-1 expression by OmpA+ E. coli

Our previous studies have demonstrated that OmpA+ E. coli invasion requires the activation of PKC-α and PI3-K, which are downstream of Ecgp interaction with OmpA [18,22]. Therefore, to examine whether PKC-α and/or PI3-K are required for ICAM-1 expression upon infection with the bacteria, HBMEC were transfected with dominant negative constructs of PKC-α and the p85 subunit of PI3-K. Both PKC-α and PI3-K are shown to be necessary for E. coli invasion of HBMEC [10,14]. In addition, HBMEC were also transfected with kinase-dead Src dominant negative construct. Both DN-PKC-α and DN-PI3-K/HBMEC showed substantial inhibition of ICAM-1 expression when infected with OmpA+ E. coli at all the time points (Fig. 5A). In contrast, DN-Src/HBMEC showed no such decrease in the ICAM-1 expression, suggesting that OmpA+ E. coli activates ICAM-1 expression via PKC-α and PI3-K pathways. To examine whether PKC-α or PI3-K inhibition results in inhibition of NF-κB activity induced by E. coli, EMSA were performed with nuclear extracts from DN-transfected HBMEC. The densities of NF-κB bands were determined from three independent experiments and expressed in a graph. As seen in Fig. 5B, no increase in NF-κB activity was observed in both DN-PKC-α/HBMEC and DN-p85/HBMEC infected with OmpA+ E. coli. However, in DN-Src/HBMEC, substantial increase in NF-κB activity was detected. These results indicate that PKC-α and PI3-K-dependent pathways are involved in NF-κB activation in OmpA+ E. coli infected HBMEC.

Fig. 5
PKC-α and P13-K but not Src-kinase signaling is required for the induction of ICAM-1 expression by OmpA+ E. coli. (A) HBMEC were transfected with dominant negative constructs of PKC-α, p85 subunit of PI3-K, and a kinase dead Src in pcDNA3 ...

3.6. E. coli-induced ICAM-1 expression results in increased adhesion of THP-1 monocytes

To investigate whether E. coli induced ICAM-1 expression had an effect on the adhesive properties of HBMEC, we performed adhesion experiments using the monocytic THP-1 cells transfected with GFP plasmid. HBMEC were treated with E. coli for 8 h before incubation with GFP-expressing THP-1 cells for 30 min. Non-adherent cells were removed by washing, mounted in DAPI, and GFP-fluorescent cells were counted. The binding of THP-1 to OmpA+ E. coli infected HBMEC was significantly greater than the control and OmpA− E. coli infected cells (Fig. 6A,B, P < 0.01 by Student’s t-test). The increased binding of THP-1 cells is not due to increased association of OmpA+ E. coli with HBMEC as our previous studies showed that overexpression of ICAM-1 had no effect on the adherence of the bacteria to HBMEC [17]. In addition, THP-1 adherence to HBMEC transfected with DN-PKC-α and DN-PI3-K constructs was reduced by 50% and 70%, respectively (P < 0.05, Fig. 6C). These results are in agreement with the ICAM-1 expression data and further substantiate the role of PKC-α and PI3-K in the expression of ICAM-1 induced by E. coli. They also suggest the biological relevance of E. coli induced ICAM-1 expression in recruiting monocytes during HBMEC infection.

Fig. 6
E. coli induced ICAM-1 expression results in increased adhesion of THP-1 monocytes. (A) Non-transfected HBMEC were plated in eight-well chamber slides. After reaching confluence, the cells were treated with E. coli as described in Section 2. The cells ...

4. Discussion

An increasing body of evidence suggests that bacterial invasion of endothelial cells can initiate profound alterations in their functions. For example, several pathogens known to invade the central nervous system (CNS), including Listeria and Borrelia, have been shown to up-regulate the expression of adhesion molecules in HUVEC and to stimulate leukocyte adhesion [29,30]. For E. coli K1, we have previously shown that it is far more efficient at invading HBMEC than HUVEC, mainly because of the critical interaction between OmpA and its receptor, Ecgp, which is expressed in HBMEC but not in HUVEC [17]. Furthermore, we demonstrated that the tight junctions of only those HBMEC invaded by OmpA+ E. coli were selectively disrupted but not HUVEC infected with the same bacteria [31]. In agreement with the in vitro data, leakage of injected horseradish peroxidase in newborn rats infected with OmpA+ E. coli was observed only in certain areas of the brain, suggesting that some parts of the brain are more affected (unpublished results). Therefore, we speculate that the invasion of E. coli in selective areas of the blood–brain barrier might increase the expression of cell adhesion molecules to which sites larger numbers of leukocytes would be recruited. The transmigration of leukocytes might contribute to the damage to these areas of brain such as hippocampal dentate gyrus and leakage of the blood–brain barrier.

Here, we demonstrate that OmpA+ E. coli invasion of HBMEC selectively up-regulates the expression of ICAM-1, but not other cell adhesion molecules tested, VCAM-1 or E-selectin. Studies on F. tularensis LVS [32] reveal a similar selective expression of ICAM-1 and VCAM-1 without affecting E-selectin, whereas periodontopathic bacteria, Porphyromonas gingivalis in cultured HEp-2 epithelial cells induces the expression of ICAM-1 but not VCAM-1 [33]. The LPS present on Gram-negative bacteria has been shown to be a major factor in inducing cell adhesion molecule expression in endothelial cells. However, OmpA− E. coli despite the presence of LPS could not induce ICAM-1 expression parallel to the levels induced by OmpA+ E. coli regardless of similar binding to HBMEC, indicating that invasion of E. coli is critical for selective expression of ICAM-1. Of note, the enhanced expression of ICAM-1 induced by OmpA+ E. coli either in the absence or presence of serum in the incubation medium, implying that LPS plays a very minimal role (data not shown). Other E. coli K1 proteins besides OmpA have been shown to play a role in the invasion of bacteria into HBMEC, such as IbeA, IbeB, and CNF, although neither their cognate receptors nor modes of action are fully characterized yet [34]. Nonetheless, antibodies to OmpA or Ecgp significantly inhibited the OmpA+ E. coli induced expression of ICAM-1, demonstrating the requirement of OmpA–Ecgp interaction for the selective expression of ICAM-1. These finding indicate a novel bacterial ligand-HBMEC receptor interaction for the up-regulation of ICAM-1 expression.

Several pathogens also have been shown to induce expression of pro-inflammatory cytokines in endothelial cells, which subsequently contribute to the up-regulation of ICAM-1, VCAM-1 and E-selectin [35]. Our data suggest that both OmpA+ and OmpA− E. coli K1 induce the production of similar levels of cytokines. Thus, pro-inflammatory cytokines secreted by HBMEC infected with the bacteria play an insignificant role in enhanced expression of ICAM-1, at least in culture and OmpA has a marginal role in inducing cytokine production. The expression of ICAM-1 occurs only in HBMEC infected with OmpA+ E. coli but not in all cells further suggesting that the released cytokines might be playing no role in increased ICAM-1 expression. NF-κB elements within the proximal ICAM-1 promoter region mediate the increased expression of ICAM-1 in HUVEC and epithelial cells exposed to cytokines (TNF-α and IL-1β) [36]. Similarly, our results demonstrate that OmpA+ E. coli induced the activation of NF-κB when compared to OmpA− E. coli for up-regulation of ICAM-1 expression. In the classical or canonical pathway, activation of the IKK complex leads to the phosphorylation at the serine or tyrosine residues of the IKK-β subunit, which in turn phosphorylates two N-terminal specific serines in IKK-β, thus targeting it for ubiquitination. Our studies with transiently transfected NIK, IKK-α and IKK-β dominant negative HBMEC revealed that activation of NF-κB was reliant on IKK-β and not on NIK and IKK-α. Elewaut et al. showed that the inhibition of IKK-β alone was sufficient to block NF-κB activity induced by enteroinvasive bacteria in intestinal epithelial cells [37]. Thus, there is precedence for a key role for IKK-β acting upstream of ICAM-1 expression initiated by bacterial invasion. We further demonstrated using dominant negative constructs, that PKC-α and PI3-kinase participate in the signaling pathways leading to OmpA+ E. coli-induced up-regulation of ICAM-1 expression while Src-kinase does not. In contrast, PMA induced activation of ICAM-1 was blocked by dominant negative constructs of PKC-α and Src-kinase (data not shown). Previous studies have reported that both TNF-α and interferon-γ mediates ICAM-1 up-regulation, which is dependent on PKC-α and Src-kinase activity [28,38]. In contrast, OmpA–Ecgp interaction, which is required for invasion of HBMEC, appears to trigger specific signaling pathways leading to up-regulation of ICAM-1 and therefore may differ from the signaling events induced by PMA and cytokines.

There are two salient features of this study. First, the invasion of E. coli into HBMEC is a requisite for selective up-regulation of ICAM-1 expression. Second, there is a requirement for PKC-α and PI3-kinase signaling, which are stimulated proximally to Ecgp after interacting with OmpA for up-regulation of ICAM-1 expression. Other intermediate signaling molecules that relay the signals for the up-regulation of ICAM-1 are not clearly known at present; however, they appear to be distinct from that of either TNF-α or PMA-induced activation of ICAM-1. Our present data suggest that selective enhanced expression of ICAM-1 in HBMEC invaded by E. coli could behave as ‘‘hot spots’’ for the recruitment of greater number of leukocytes, which in turn cause damage to the brain. In addition, the cumulative effect of leukocyte-induced and E. coli-induced blood–brain barrier disruption might also be responsible for the damage of selective areas in the brain. Therefore, prevention of E. coli K1 invasion-induced recruitment of leukocytes seemed partially important for the control of the morbidity associated with neonatal meningitis.


We thank Sarah Parsons for providing Src kinase mutants and June Li for providing dominant negative constructs of IKK-α, IKK-β and NIK. Our sincere thanks to Drs. Martine Torres and Barbara Driscoll for critical reading of the manuscript. This work was supported by grants from PHS (AI40567 to N.V.P.) and partially by CHLA research career development award (to S.K.S.).


outer membrane protein A
human brain microvascular endothelial cells
endothelial cell glycoprotein
protein kinase C-α
phosphatidylinositol 3-kinase
nuclear factor-κB


1. Hubbard AK, Rothlein R. Intercellular adhesion molecule-1 (ICAM-1) expression and cell signaling cascades. Free Radic Biol Med. 2000;28:1379–1386. [PubMed]
2. Gearing AJ, Newman W. Circulating adhesion molecules in disease. Immunol Today. 1993;14:506–512. [PubMed]
3. Turowski P, Adamson P, Greenwood J. Pharmacological targeting of ICAM-1 signaling in brain endothelial cells: potential for treating neuro-inflammation. Cell Mol Neurobiol. 2005;25:153–170. [PubMed]
4. Kaukoranta-Tolvanen SS, Ronni T, Leinonen M, Saikku P, Laitinen K. Expression of adhesion molecules on endothelial cells stimulated by Chlamydia pneumoniae. Microb Pathog. 1996;21:407–411. [PubMed]
5. Frick AG, Joseph TD, Pang L, Rabe AM, St Geme JW, 3rd, Look DC. Haemophilus influenzae stimulates ICAM-1 expression on respiratory epithelial cells. J Immunol. 2000;164:4185–4196. [PubMed]
6. Nasreen N, Mohammed KA, Ward MJ, Antony VB. Mycobacterium-induced transmesothelial migration of monocytes into pleural space: role of intercellular adhesion molecule-1 in tuberculous pleurisy. J Infect Dis. 1999;180:1616–1623. [PubMed]
7. Rothwarf DM, Karin M. The NF-kappa B activation pathway: a paradigm in information transfer from membrane to nucleus. Sci STKE. 1999;1999:RE1. [PubMed]
8. Baeuerle PA, Baltimore D. I kappa B: a specific inhibitor of the NF-kappa B transcription factor. Science. 1988;242:540–546. [PubMed]
9. Karin M, Ben-Neriah Y. Phosphorylation meets ubiquitination: the control of NF-[kappa]B activity. Annu Rev Immunol. 2000;18:621–663. [PubMed]
10. Krappmann D, Wegener E, Sunami Y, Esen M, Thiel A, Mordmuller B, Scheidereit C. The IkappaB kinase complex and NF-kappa-B act as master regulators of lipopolysaccharide-induced gene expression and control subordinate activation of AP-1. Mol Cell Biol. 2004;24:6488–6500. [PMC free article] [PubMed]
11. Ghosh S, Karin M. Missing pieces in the NF-kappaB puzzle. Cell. 2002;109(Suppl):S81–S96. [PubMed]
12. Leib SL, Tauber MG. Pathogenesis of bacterial meningitis. Infect Dis Clin North Am. 1999;13:527–548 v-vi. [PubMed]
13. Stoll BJ, Hansen N, Fanaroff AA, Wright LL, Carlo WA, Ehrenkranz RA, Lemons JA, Donovan EF, Stark AR, Tyson JE, Oh W, Bauer CR, Korones SB, Shankaran S, Laptook AR, Stevenson DK, Papile LA, Poole WK. Changes in pathogens causing early-onset sepsis in very-low-birth-weight infants. N Engl J Med. 2002;347:240–247. [PubMed]
14. van der Flier M, Geelen SP, Kimpen JL, Hoepelman IM, Tuomanen EI. Reprogramming the host response in bacterial meningitis: how best to improve outcome? Clin Microbiol Rev. 2003;16:415–429. [PMC free article] [PubMed]
15. Prasadarao NV, Wass CA, Weiser JN, Stins MF, Huang SH, Kim KS. Outer membrane protein A of Escherichia coli contributes to invasion of brain microvascular endothelial cells. Infect Immun. 1996;64:146–153. [PMC free article] [PubMed]
16. Prasadarao NV, Wass CA, Kim KS. Endothelial cell GlcNAc beta 1-4GlcNAc epitopes for outer membrane protein A enhances traversal of Escherichia coli across the blood-brain barrier. Infect Immun. 1996;64:154–160. [PMC free article] [PubMed]
17. Prasadarao NV. Identification of Escherichia coli outer membrane protein A receptor on human brain microvascular endothelial cells. Infect Immun. 2002;70:4556–4563. [PMC free article] [PubMed]
18. Sukumaran SK, Prasadarao NV. Regulation of protein kinase C in Escherichia coli K1 invasion of human brain microvascular endothelial cells. J Biol Chem. 2002;277:12253–12262. [PubMed]
19. Sukumaran SK, Quon MJ, Prasadarao NV. Escherichia coli K1 internalization via caveolae requires caveolin-1 and protein kinase C alpha interaction in human brain microvascular endothelial cells. J Biol Chem. 2002;277:50716–50724. [PubMed]
20. Reddy MA, Wass CA, Kim KS, Schlaepfer DD, Prasadarao NV. Involvement of focal adhesion kinase in Escherichia coli invasion of human brain microvascular endothelial cells. Infect Immun. 2000;68:6423–6430. [PMC free article] [PubMed]
21. Sukumaran SK, McNamara G, Prasadarao NV. Escherichia coli K-1 interaction with human brain micro-vascular endothelial cells triggers phospholipase C-gamma1 activation downstream of phosphatidylinositol 3-kinase. J Biol Chem. 2003;278:45753–45762. [PubMed]
22. Reddy MA, Prasadarao NV, Wass CA, Kim KS. Phosphotidyl inositol Kinase activation and its interaction with focal adhesion kinase is required for E. coli invasion of human brain endothelial cells. J Biol Chem. 2000;275:36769–36774. [PubMed]
23. Stins MF, Prasadarao NV, Ibric L, Wass CA, Luckett P, Kim KS. Binding characteristics of S fimbriated Escherichia coli to isolated brain microvascular endothelial cells. Am J Pathol. 1994;145:1228–1236. [PMC free article] [PubMed]
24. Stins MF, Prasadarao NV, Zhou J, Arditi M, Kim KS. Bovine brain microvascular endothelial cells transfected with SV40-large T antigen: development of an immortalized cell line to study pathophysiology of CNS disease. In Vitro Cell Dev Biol Anim. 1997;33:243–247. [PubMed]
25. McEvoy LM, Sun H, Tsao PS, Cooke JP, Berliner JA, Butcher EC. Novel vascular molecule involved in monocyte adhesion to aortic endothelium in models of atherogenesis. J Exp Med. 1997;185:2069–2077. [PMC free article] [PubMed]
26. Selvaraj SK, Prasadarao NV. Escherichia coli K1 inhibits proinflammatory cytokine induction in monocytes by preventing NF-kappa B activation. J Leukoc Biol. 2005;78:544–554. [PubMed]
27. Zhang L, Tarleton RL. Persistent production of inflammatory and anti-inflammatory cytokines and associated MHC and adhesion molecule expression at the site of infection and disease in experimental Trypanosoma cruzi infections. Exp Parasitol. 1996;84:203–213. [PubMed]
28. Bernot D, Peiretti F, Canault M, Juhan-Vague I, Nalbone G. Upregulation of TNF-alpha-induced ICAM-1 surface expression by adenylate cyclase-dependent pathway in human endothelial cells. J Cell Physiol. 2005;202:434–441. [PubMed]
29. Lopez S, Prats N, Marco AJ. Expression of E-selectin, P-selectin, and intercellular adhesion molecule-1 during experimental murine listeriosis. Am J Pathol. 1999;155:1391–1397. [PMC free article] [PubMed]
30. Ebnet K, Brown KD, Siebenlist UK, Simon MM, Shaw S. Borrelia burgdorferi activates nuclear factor-kappa B and is a potent inducer of chemokine and adhesion molecule gene expression in endothelial cells and fibroblasts. J Immunol. 1997;158:3285–3292. [PubMed]
31. Sukumaran SK, Prasadarao NV. Escherichia coli K1 invasion increases human brain microvascular endothelial cell monolayer permeability by disassembling vascular-endothelial cadherins at tight junctions. J Infect Dis. 2003;188:1295–1309. [PubMed]
32. Forestal CA, Benach JL, Carbonara C, Italo JK, Lisinski TJ, Furie MB. Francisella tularensis selectively induces proinflammatory changes in endothelial cells. J Immunol. 2003;171:2563–2570. [PubMed]
33. Ansai T, Yamamoto E, Awano S, Yu W, Turner AJ, Takehara T. Effects of periodontopathic bacteria on the expression of endothelin-1 in gingival epithelial cells in adult periodontitis. Clin Sci (Lond) 2002;103(Suppl 48):327S–331S. [PubMed]
34. Kim KS. Pathogenesis of bacterial meningitis: from bacteraemia to neuronal injury. Nat Rev Neurosci. 2003;4:376–385. [PubMed]
35. Hess DC, Bhutwala T, Sheppard JC, Zhao W, Smith J. ICAM-1 expression on human brain microvascular endothelial cells. Neurosci Lett. 1994;168:201–204. [PubMed]
36. Kuldo JM, Westra J, Asgeirsdottir SA, et al. Differential effects of NF-{kappa}B and p38 MAPK inhibitors and combinations thereof on TNF-{alpha}- and IL-1{beta}-induced proinflammatory status of endothelial cells in vitro. Am J Physiol Cell Physiol. 2005;289:C1229–C1239. [PubMed]
37. Elewaut D, DiDonato JA, Kim JM, Truong F, Eckmann L, Kagnoff MF. NF-kappa B is a central regulator of the intestinal epithelial cell innate immune response induced by infection with enteroinvasive bacteria. J Immunol. 1999;163:1457–1466. [PubMed]
38. Melrose J, Tsurushita N, Liu G, Berg EL. IFN-gamma inhibits activation-induced expression of E- and P-selectin on endothelial cells. J Immunol. 1998;161:2457–2464. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...