OUP user menu

Listeria monocytogenes-infected human umbilical vein endothelial cells: internalin-independent invasion, intracellular growth, movement, and host cell responses

Lars Greiffenberg , Zeljka Sokolovic , Hans-Joachim Schnittler , Andrea Spory , Regine Böckmann , Werner Goebel , Michael Kuhn
DOI: http://dx.doi.org/10.1111/j.1574-6968.1997.tb12768.x 163-170 First published online: 1 December 1997

Abstract

The interaction of Listeria monocytogenes with human umbilical vein endothelial cells was studied. We show that L. monocytogenes invades human umbilical vein endothelial cells independently of internalin A, internalin B, internalin C, and ActA. L. monocytogenes replicates efficiently inside the cells and moves intracellularly by the induction of actin polymerization. We further show that L. monocytogenes-infection of human umbilical vein endothelial cells induces interleukin-6 and interleukin-8 expression during the first 6 h of infection. The expression of MCP-1 and the adhesion molecules VCAM-1 and ICAM-1 was not altered under the experimental conditions used here.

Keywords
  • Listeria monocytogenes
  • Human umbilical vein endothelial cells
  • Invasion
  • Internalin
  • Interleukin-6
  • Interleukin-8

1 Introduction

The Gram-positive, facultative intracellular bacterial species Listeria monocytogenes is known to infect many mammalian cell types in vitro [1]. Once inside the host cell, the bacteria gain access to the cytoplasm, due to the action of the pore forming cytolysin, called listeriolysin O, and the phospholipase PI-PLC encoded by the plcA gene [1]. The subsequent intracytoplasmic multiplication is accompanied by intracellular movement and cell to cell spread. Most of the virulence genes of L. monocytogenes identified so far, which are required for the intracellular life cycle, are located on the chromosome in the virulence gene cluster and their expression is regulated by the transcriptional activator PrfA [1, 2]. The members of the internalin gene family are distributed on the chromosome and organized either in clusters (inlAB and inlC2DE) or as single genes (inlC and inlF) [35]. The gene products of the inlA and inlB genes are involved in invasion of epithelial cells and hepatocytes [3, 6, 7], whereas the role in host cell invasion, if any, of the other members of the internalin family is not clear. L. monocytogenes infections of various mammalian cell types, including macrophages, epithelial cells and endothelial cells, cause massive alterations of gene transcription and cellular metabolism in the host cell. These alterations include induction of proinflammatory cytokines [8], repression of cytokine receptor expression [8], activation of transcription factors [9], stimulation of lipid mediator synthesis [10], and induction of cell adhesion molecule expression [11]. The last two types of host cell responses, which were analyzed in human umbilical vein endothelial cells (HUVECs), show that this cell type can efficiently respond to L. monocytogenes infections.

The final outcome of human infections with L. monocytogenes is either meningitis or meningo-encephalitis [12], which suggests that L. monocytogenes is able to cross the blood-brain barrier. During passage across the blood-brain barrier, L. monocytogenes most likely interacts with endothelial cells. Due to the difficult experimental access to human brain microvascular endothelial cells for infection assays with L. monocytogenes, freshly isolated HUVECs have been used as a suitable model for Listeria-endothelial cell interaction.

L. monocytogenes was recently shown to invade HUVECs in vitro by two different mechanisms [13]. The bacteria either infect the HUVECs directly or enter the cells by a heterologous spreading mechanism from an infected macrophage.

2 Materials and methods

2.1 Bacterial strains and culture

Listeriae used for cell culture infections were grown in brain-heart infusion broth (Difco) at 37°C with aeration. Mid-exponential-phase cultures were washed once in phosphate-buffered saline (PBS) and stored at −80°C in aliquots in PBS, with glycerol added to 15% (v/v).The bacterial strains used for infection are shown in Table 1.

View this table:
Table 1

L. monocytogenes strains used

StrainGenotypeCharacteristicsSource or reference
Sv 1/2a EGDWTlocal strain collection
NCTCa 7973WT, hyperhemolyticlocal strain collection
ΔinlCΔinlCin-frame deletionF. Engelbrecht [5]
A42ΔprfAin-frame deletionB. Middendorf [21]
A49ΔactAin-frame deletionZ. Sokolovic [9]
A76ΔinlAin-frame deletionthis study
WL-111ΔinlBin-frame deletionthis study
WL-112ΔinlAinlBin-frame deletionsthis study
WL-113ΔinlBinlCin-frame deletionsthis study
  • aNCTC: National Cancer Tissue Cultures, NIH.

2.2 Generation of the in-frame deletion mutants L. monocytogenes ΔinlA, L. monocytogenes ΔinlB, L. monocytogenes ΔinlA/ΔinlB, and L. monocytogenes ΔinlB/ΔinlC

A 377 bp fragment from the 5′ end of the inlA gene [3] was amplified from L. monocytogenes chromosomal DNA by PCR with the primers 5′-TCGCAAAGAATTCTAGACCAAG-3′ and 5′-AATCTGTCCCGGGATTACCAAA-3′. The primers were designed to generate restriction sites for EcoRI and XmaI, respectively. A second 440 bp fragment from the 3′ end of the inlA gene was amplified with the primers 5′-CCAATATCCCGGGAATAGCTAT-3′ and 5′-TAGAGCGAATTCTCGCTATCGCC-3′. These primers were designed to generate restriction sites for XmaI and EcoRI, respectively. The inserts were then ligated into pTZ18R resulting in plasmid pTZ18R-1/4–13. The EcoRI insert was cut out of this plasmid as an EcoRI/BamHI fragment and then ligated into the shuttle vector pLSV-1 [14] resulting in plasmid pLSV1/7.

A 263 bp fragment from the 5′ end of the inlB gene [3] was amplified from chromosomal DNA by PCR with the primers 5′-GGAAGAATTCTTTAATCTCAGG-3′ and 5′-ATACTGGGTACCTTGAACGGATT-3′. The primers were designed to generate sticky end restriction sites for EcoRI and KpnI, respectively. A second 301 bp fragment from the 3′ end of the inlB gene was amplified with the primers 5′-AATAAGGTACCGGTCTACCAAGG-3′ and 5′-CGATGGATCCTTGACAATCAAC-3′. These primers were designed to generate restriction sites for KpnI and BamHI, respectively. The inserts were then ligated into pUC18 in a two-step ligation resulting in plasmid pUC18-AB. The EcoRI/BamHI insert from pUC18-AB was then ligated into the shuttle vector pLSV-1 [14] resulting in plasmids pLSV-AB.

By using the plasmids pLSV1/7 and pLSV-AB, allelic exchange was performed with L. monocytogenes EGD to generate the inlA and the inlB deletion strains L. monocytogenesΔinlA (A76) and L. monocytogenesΔinlB (WL-111). The ΔinlA/ΔinlB (WL-112) and the ΔinlB/ΔinlC (WL-113) double mutants were constructed by allelic exchange with plasmid pLSV-AB and L. monocytogenesΔinlA and L. monocytogenesΔinlC[5] as recipients, respectively. The correct deletions on the chromosomes of the respective strains were confirmed by PCR amplification of the deleted alleles and by sequencing the deletion sites.

2.3 Isolation and culture of HUVECs

Endothelial cells were isolated from human umbilical cord veins and cultured as described elsewhere [15]. Cells were grown to confluence in medium 199 (Gibco BRL, Eggenstein, Germany) supplemented with 20% pooled human serum from healthy donors (obtained from a local blood bank) and 50 U ml−1 penicillin G and 50 μg ml−1 streptomycin sulfate (Sigma, Deisenhofen, Germany). 24 h prior to infection cells were washed and further cultured in antibiotic free media for 12 h. Cells of passage one were used for the experiments.

2.4 Infection of HUVECs with L. monocytogenes and determination of invasion and intracellular growth

HUVECs, isolated and cultured as described, were seeded in 6 well plates at 5×105 HUVECs per well and infected with L. monocytogenes. After centrifugation of the bacteria onto the endothelial cell monolayer for 10 min at 1000×g and the 60 min infection period, the cultures were washed five times and further incubated in the presence of 50 μg ml−1 gentamicin for 1 h. After lysis of the monolayer by the addition of ice cold distilled water, appropriate dilutions were plated on brain-heart infusion agar. To count only intracellular replicating bacteria, washed cultures were further incubated for up to 24 h with medium containing 10 μg ml−1 gentamicin, then washed again, lysed and plated as described. All experiments were done in triplicate and repeated up to three times.

2.5 Immunofluorescence

HUVEC monolayers were infected as described above. At 6 h post infection, the cells were fixed and stained with 7-nitrobenz-2-oxa-1,3-diazole-phalloidin (NBD-phalloidin) as previously described [13] and analyzed microscopically.

2.6 RNA isolation, cDNA synthesis and PCR analysis

At the indicated time points after infection total RNA was extracted from the cells by the guanidinium thiocyanate procedure as described by Chomczynski and Sacchi [16]. Complementary DNA was synthesized from 10 μg of total RNA with the Stratagene first-strand synthesis kit (San Diego, CA) as previously noted in detail [8]. The PCR amplification was performed as described [8]. The primers used are outlined in Table 2. Appropriate volumes of cDNA were amplified essentially as described elsewhere [8]. The reaction products were visualized by electrophoresis and the specificities of the amplified bands were confirmed by their predicted sizes (see Table 2). As an internal control, PCR amplifications with β-actin-specific primers were always included. RNA samples not treated with reverse transcriptase were also included in the amplification reactions to check for DNA contamination.

View this table:
Table 2

Primer sequences used for PCR and expected sizes of the PCR products

Gene productPrimer pairPCR products (bp)
TNF-αa5′-CCTTGGTCTGGTAGGAGACG-3′325
5′-CAGAGGGAAGAGTTCCCCAG-3′
IL-1βa5′-AATTTTTGGGATCTACACTCTCCAGCTGTA-3′331
5′-CTTCATCTTTGAAGAAGAACCTATCTTCTT-3′
IL-6a5′-ATGAACTCCTTCTCCACAAGCGC-3′628
5′-GAAGAGCCCTCAGGCTGGACTG-3′
IL-8b5′-TTTGCCAAGGAGTGCTAAAG-3′199
5′-CTCCACAACCCTCTGCACCC-3′
MCP-1b5′-CAATCAATGCCCCAGTCACC-3′479
5′-AGACCCTCAAAACATCCCAG-3′
VCAM-1b5′-TGATGACAGTGTCTCCTTCTTTG-3′450
5′-ATCCCTACCATTGAAGATACTGG-3′
ICAM-1b5′-CGTGCCGCACTGAACTGGAC-3′446
5′-CCTCACACTTCACTGTCACCT-3′
β-Actina5′-TGACGGGGTCACCCACACTGTGCCCATCTA-3′661
5′-CTAGAAGCATTTGCGGTGGACGATGGAGGG-3′
  • aPrimer sequences were described by Arnold et al. [22].

  • bPrimer sequences were selected on the basis of published cDNA sequences.

2.7 ELISA

The detection of interleukin-8 (IL-8) in the supernatant of infected HUVECs was performed with an ELISA kit from R&D Systems (Minneapolis) as recommended by the manufacturer.

3 Results

3.1 L. monocytogenes invades HUVECs, grows intracellularly and recruits actin filaments

Here we confirm the invasive capacity of L. monocytogenes for HUVECs and also demonstrate that microfilaments are important for invasion since uptake is inhibited by cytochalasin D treatment (data not shown). Using low multiplicities of infection (MOI) of two bacteria per cell we were able to keep the infected monolayers intact for up to 20 h, a period which resulted in constant increase in the numbers of intracellular bacteria. As outlined in Fig. 1, the hyperhemolytic strain L. monocytogenes NCTC 7973 grows significantly faster than the weakly hemolytic strain L. monocytogenes EGD in HUVECs over a period of 20 h without damaging the cells.

Figure 1

L. monocytogenes NCTC 7973 grows faster in HUVECs than strain L. monocytogenes EGD. HUVECs were infected at a MOI of 2 bacteria per cell and the number of intracellular bacteria was calculated at different time points post infection as described.

In parallel with the onset of intracellular multiplication, L. monocytogenes becomes associated with host cell filamentous actin inside HUVECs as demonstrated by NBD-phalloidin staining of actin filaments as shown in Fig. 2. The filamentous actin is organized in actin tails, the typical signature of ongoing intracellular movement [1]. The long-term multiplication of L. monocytogenes inside HUVECs is also associated with intercellular spread of the bacteria from the initially infected cells to their neighbors (data not shown).

Figure 2

L. monocytogenes moves inside HUVECs. HUVECs were infected with L. monocytogenes EGD as described. At 6 h post infection the cells were fixed and treated with NBD-phalloidin to stain filamentous actin. A typical actin tail is marked by an arrowhead.

3.2 Invasion is independent of InlA, InlB, InlC, and ActA

HUVECs are normally non-phagocytic cells and the uptake of L. monocytogenes therefore requires induced phagocytosis. It has been shown that members of the family of internalins (InlA and InlB) can trigger uptake of these bacteria by epithelial cells and hepatocytes [3, 6, 7]. We therefore studied the potential role of the gene products of the members of the internalin multigene family in HUVEC invasion by performing classical invasion assays using gentamicin to kill extracellular bacteria. Different well defined in-frame deletion mutants in the internalin genes were used and invasion assays were performed, the results of which are outlined in Fig. 3. All mutants tested (ΔinlA, ΔinlB, ΔinlC, ΔinlA/ΔinlB, ΔinlB/ΔinlC) are nearly as invasive as the otherwise isogenic wild-type strain L. monocytogenes EGD which clearly demonstrates that neither InlA nor InlB is involved in triggering L. monocytogenes uptake by HUVECs. Even a double mutant lacking InlA and InlB and a mutant lacking the small secreted InlC protein [5] were as invasive as the wild-type strain. Additionally, the strains L. monocytogenes EGD and NCTC 7973, which differ in their intracellular growth rate (Fig. 1), behave nearly identically in an invasion assay (Fig. 3). As shown in Fig. 3, the listerial surface protein ActA is also not involved in internalization of L. monocytogenes by HUVECs, since an in-frame deletion mutant in the actA gene invades HUVECs with the same efficiency as the wild-type strain (Fig. 3). Internalization is, however, PrfA-dependent as shown by the reduced invasiveness of a prfA in-frame deletion mutant.

Figure 3

Invasive capacity of different L. monocytogenes mutants. HUVECs were infected (MOI=2) with different in-frame deletion mutants of L. monocytogenes EGD and the percentage of intracellular bacteria was calculated at 1 h post infection as described. L. monocytogenes EGD (A), NCTC 7973 (B), ΔinlA (C), ΔinlB (D), ΔinlA/ΔinlB (E), ΔinlC (F), ΔinlB/ΔinlC (G), ΔprfA (H), ΔactA (I).

3.3 L. monocytogenes infection of HUVECs induces IL-6 and IL-8 expression

We have recently used RT-PCR to analyze gene expression in macrophages upon infection with L. monocytogenes[1, 8]. This approach was also applied here to study altered gene expression in L. monocytogenes-infected HUVECs. Upon infection of HUVECs with L. monocytogenes EGD (MOI=5), total RNA was isolated, cDNA synthesized, and used as a template for PCR with different sets of specific primer pairs. As shown in Fig. 4, infection of HUVECs with L. monocytogenes EGD resulted in an increase in IL-6- and especially IL-8-specific mRNAs. Transcripts specific for IL-1 β and TNF-α were not detectable in non-infected and infected cells. Having established the L. monocytogenes-induced expression of IL-6 and IL-8 upon infection, we followed the kinetics of their induction. RNA was isolated at different time points ranging from 0 to 6 h post infection and analyzed by RT-PCR. As shown in Fig. 5, IL-6 expression started around 4 h post infection. IL-8-specific mRNA first appeared at 1 h post infection and increased constantly until 6 h post infection. Low levels of IL-8 protein were detectable by ELISA in the cell culture of non-infected HUVECs as shown in Fig. 6. The amount of IL-8 increased in L. monocytogenes-infected HUVECs beginning at 2 h post infection and reached levels of more than 5000 pg ml−1 at 4 and 6 h post infection.

Figure 4

Induction of mRNA expression in HUVECs after infection with L. monocytogenes EGD (I) (MOI=5). Total RNA was isolated at 4 h post infection and transcripts analyzed by RT-PCR. β-Actin was amplified as an internal control to check for similar RNA concentrations in the two RNA preparations. NI, non-infected control; M, molecular mass marker.

Figure 5

Time course of IL-6 and IL-8 expression in HUVECs after infection with L. monocytogenes EGD (MOI=5). Total RNA was isolated at different time points post infection and β-actin-specific, IL-6-specific, and IL-8-specific transcripts were analyzed by RT-PCR. NI, non-infected control; M, molecular mass marker.

Figure 6

IL-8 concentrations in the supernatant of L. monocytogenes-infected HUVECs (MOI=5) at different time points post infection. IL-8 concentrations in the supernatant were determined by ELISA over a period of 6 h post infection. NI: non-infected control.

In addition to the cytokines, we also tested the expression of MCP-1 and the cell adhesion molecules ICAM-1 and VCAM-1. However, under the conditions used, no difference was detected in infected and non-infected HUVECs (Fig. 4). Whereas ICAM-1 mRNA was not found, MCP-1 and VCAM-1 mRNAs were already detectable in non-infected cells and their expression was obviously not altered upon infection with L. monocytogenes EGD for 4 h.

4 Discussion

In the present study we demonstrate the ability of L. monocytogenes to enter HUVECs by a microfilament-dependent pathway which is followed by efficient intracellular multiplication as judged by gentamicin-based survival assays. Two widely used L. monocytogenes strains, NCTC 7973 and EGD, were compared for their ability to invade and multiply intracellularly. It was found that the invasive capacities are nearly identical, but strain NCTC 7973 grows faster over a period of 20 h. The prfA genes encoding the transcriptional regulator PrfA of strain NCTC 7973 and strain EGD are slightly different. These differences are most likely the reason for the enhanced expression of the L. monocytogenes virulence factors by strain NCTC 7973 already under extracellular conditions [17]. The way in which the differences in gene expression are linked to altered intracellular multiplication is, however, not understood. As shown by the cytochalasin D sensitivity of the L. monocytogenes invasion process, uptake is clearly microfilament-dependent as it is in other cell types tested [1]. Once within the cytoplasm, the bacteria polymerize cellular actin and move intracellularly as was analyzed in detail in other cellular systems [1].

The cloning of the inlAB operon and the recent identification of five additional members of the internalin multigene family [35] have raised the question whether the individual members of the family are linked to invasion of different cell types. InlA and InlB were shown to be crucial for epithelial and hepatocyte invasion, respectively. Drevets and colleagues, working with a Tn1545 mutant in the inlAB operon [3], postulated a role for InlA in HUVEC invasion [13]. We analyzed the contribution of some internalins in triggering L. monocytogenes uptake by HUVECs. Our data obtained with in-frame deletion mutants in some of the internalin genes clearly show that neither InlA, InlB nor InlC plays a role in invasion of HUVECs. Even double mutants are nearly as invasive as the wild-type strain. One explanation for this apparent discrepancy is the source of the endothelial cells. We used freshly isolated primary HUVECs at passage one while Drevets et al. used a HUVEC ATCC cell line with a life expectancy of more than 50 doublings [11, 13]. The listerial surface molecule ActA which is transcribed from PrfA-dependent promoters and which was shown to bind to heparan sulfate proteoglycan receptors [18] is obviously also not involved in invasion of HUVECs by L. monocytogenes since a deletion in the actA gene does not reduce the invasive capacity. As shown here, invasion is clearly dependent on the positive transcriptional regulator PrfA since a mutant with an in-frame deletion in the prfA gene shows a 5–6-fold reduction in invasiveness. This result implies that a yet unknown but PrfA-dependent factor (protein) must exist in L. monocytogenes which mediates uptake by HUVECs and probably also by other endothelial cells.

Invasion of endothelial cells by L. monocytogenes was recently shown to induce the generation of lipid mediators triggered by the virulence factors LLO and PI-PLC [10]. Here we show that L. monocytogenes infection of HUVECs also results in the upregulation of IL-6 and IL-8 gene expression whereas other proinflammatory cytokines like tumor necrosis factor-α and IL-1β are not induced. IL-8, a potent chemokine active on monocytes and macrophages [19], should, upon release from L. monocytogenes-infected endothelial cells, result in the attraction of lymphocytes to the site of infection. The potential role of IL-6 expressed during invasion is not clear, but recent data show that IL-6 together with its soluble receptor stimulates endothelial cells resulting in STAT3 activation, chemokine expression and augmentation of ICAM-1 [20]. Messenger RNAs specific for adhesion molecules like ICAM-1 and VCAM-1 as well as mRNA specific for the macrophage chemoattractant protein 1 (MCP-1) were not induced under the conditions used while a recent publication showed the accumulation of ICAM-1 and VCAM-1 molecules on the surface of L. monocytogenes infected HUVECs between 4 and 18 h post infection [11]. The different results might again be due to the different source of the endothelial cells or to a lack of correlation between the accumulation of adhesion molecules on the cell surface (as observed by Drevets [11]) and the induced expression of the respective genes described here.

Acknowledgements

This work was supported by the Deutsche Forschungsgemeinschaft through Sonderforschungsbereiche 165-B4 and 355-B5 and by the European Union through the BIOMED 2 Project ‘Listeria Eurolab’, Grant CT950659. We also thank M. Koch for skilful technical assistance and J. Kreft for critically reading the manuscript.

Footnotes

  • 1 Both authors contributed equally.

References

View Abstract