Article Text

This article has a correction. Please see:

Original article
The urokinase plasminogen activator receptor (uPAR) controls macrophage phagocytosis in intestinal inflammation
  1. Marco Genua1,2,
  2. Silvia D'Alessio1,
  3. Javier Cibella1,
  4. Alessandro Gandelli1,
  5. Emanuela Sala1,
  6. Carmen Correale1,
  7. Antonino Spinelli2,3,
  8. Vincenzo Arena4,
  9. Alberto Malesci1,2,
  10. Sergio Rutella4,
  11. Victoria A Ploplis6,
  12. Stefania Vetrano1,
  13. Silvio Danese1
  1. 1IBD Center, Humanitas Clinical and Research Center, Rozzano, Italy
  2. 2Department of Translational Medicine, University of Milan, Milan, Italy
  3. 3Department of Surgery-IBD Surgery Unit, Humanitas Clinical and Research Center, Rozzano, Italy
  4. 4Department of Pediatric Hematology/Oncology, IRCCS Bambino Gesù Children's Hospital, Rome, Italy
  5. 5Department of Internal Medicine, Catholic University of Rome, Rome, Italy
  6. 6W. M. Keck Center for Transgene Research, University of Notre Dame, Notre Dame, Indiana, USA
  1. Correspondence to Dr Silvio Danese, Humanitas Clinical and Research Center—Via Manzoni, 56—Rozzano, MI 20089, Italy; sdanese{at}hotmail.com

Abstract

Objective Inflammation plays crucial roles in the pathogenesis of several chronic inflammatory disorders, including Crohn's disease (CD) and UC, the two major forms of IBD. The urokinase plasminogen activator receptor (uPAR) exerts pleiotropic functions over the course of both physiological and pathological processes. uPAR not only has a key role in fibrinolysis but also modulates the development of protective immunity. Additionally, uPAR supports extracellular matrix degradation and regulates cell migration, adhesion and proliferation, thus influencing the development of inflammatory and immune responses. This study aimed to evaluate the role of uPAR in the pathogenesis of IBD.

Design The functional role of uPAR was assessed in established experimental models of colitis. uPAR deficiency effects on cytokine release, polarisation and bacterial phagocytosis were analysed in colonic macrophages. uPAR expression was analysed in surgical specimens collected from normal subjects and patients with IBD.

Results In mice, uPAR expression is positively regulated as colitis progresses. uPAR-KO mice displayed severe inflammation compared with wild-type littermates, as indicated by clinical assessment, endoscopy and colon histology. The absence of uPAR led to an increased production of inflammatory cytokines by macrophages that showed an M1 polarisation and impaired phagocytosis. In human IBD, CD68+ macrophages derived from the inflamed mucosa expressed low levels of uPAR.

Conclusions These findings point to uPAR as an essential component of intestinal macrophage functions and unravel a new potential target to control mucosal inflammation in IBD.

  • Macrophages
  • Integrins
  • Bacterial Translocation
  • Experimental Colitis
  • Molecular Mechanisms

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Significance of this study

What is already known on this subject?

  • Despite the emerging evidence supporting a central role for uPAR in inflammation, studies on its role in IBD pathogenesis are incomplete. To the best of our knowledge, few reports analysed the expression of uPAR in patients with CD or UC.

  • Gibson and colleagues analysed uPAR expression only in colonic crypts by ELISA and found no differences between healthy subjects and patients with IBD.

  • Laurem and colleagues showed increased expression of uPAR in the enteric nerves of inflamed intestine as it is likely involved in nerve repair.

  • Finally, two recent works from Lönnkvist and Kolho evaluated the interesting possibility that soluble uPAR level in blood may be used as an inflammatory marker for IBD.

  • Taken together, a comprehensive characterisation of uPAR expression and function in IBD pathogenesis is still missing.

What are the new findings?

  • uPAR expression is modulated during colitis development. Indeed, uPAR expression is specifically increased in CX3CR1+F4/80+CD11c phagocytes.

  • In mice, uPAR absence leads to detrimental intestinal inflammation, with increased mucosal expression of Th1 pro-inflammatory cytokines.

  • The absence of uPAR alters macrophage function and activation. In particular, uPAR is required for bacterial phagocytosis via MAC-1 interaction and RAC-1 activation, and its absence leads to the loss of tolerance to bacterial product inducing a classic M1 polarisation.

  • Remarkably, macrophages infiltrating the inflamed mucosa of patients with IBD showed reduced uPAR expression compared with healthy controls.

How might it impact on clinical practice in the foreseeable future?

  • Numerous studies over the past 20 years have provided a strong rationale for developing therapeutic agents that target uPAR in inflammatory disorders.

  • However, no uPAR-targeted therapeutic agent has been advanced into clinical evaluation. A number of obstacles have retarded progress on the discovery and development of uPAR inhibitors for clinical evaluation, including a rapidly evolving landscape on the relevant activities and epitopes of uPAR to target. A deeper understanding of uPAR functions, particularly in the intestine, will result in the identification of several novel therapeutic approaches that will overcome some of these hurdles and show promise as clinical candidates in IBD.

Introduction

IBD denotes chronic inflammatory disorders that affect the gastrointestinal tract.1 There are two main clinical forms of IBD, namely, Crohn's disease (CD), which can affect any part of the gastrointestinal tract, and UC, in which pathology is restricted to the colonic mucosa.1 Although the precise aetiology of IBD is only partially understood, it has been elucidated that chronic inflammation implies a substantial degree of tissue remodelling, which heavily augments the thickness of the intestinal wall and actively participates in disease pathogenesis.2 Indeed, extracellular matrix (ECM) degrading protease systems are upregulated in inflamed and injured tissues, and crucially control tissue repair.2

Among the factors that control tissue remodelling, the urokinase-type plasminogen activator receptor (uPAR) has been extensively studied in both physiological and pathological conditions.3–7 Moreover, uPAR has been implicated in several activities regulating cell migration,8 chemotaxis,9 cell–cell interaction,10 and phagocytosis of apoptotic cells through its interaction with integrins.5 ,11 In addition, uPAR expression is augmented in cytokine or bacteria-activated cell populations, including macrophages and monocytes, contributing to the infiltration of inflammatory cells into infected tissues or organs.3 ,8 ,12–14 Therefore, uPAR exerts pleiotropic functions controlling not only ECM degradation but also modulating the development of the immune response.

However, uPAR's role in IBD pathogenesis is still unclear. Few studies analysed uPAR expression level in IBD patients15–17 without attempting to fully characterise its function in IBD pathogenesis. Indeed, uPAR was first identified in the colonic crypts of healthy controls and patients with IBD, without finding any difference in expression levels.16 Moreover, Laerum and colleagues located uPAR in the enteric nerves of the intestine and found increasing levels in patients with UC undergoing enteric nerve repair.17 Recently, two different works by Lönnkvist 18 and Kolho 19 evaluated soluble uPAR level in blood as a diagnostic marker for adult and paediatric IBD, respectively. However, a thorough characterisation of uPAR expression and function in IBD pathogenesis is still lacking.

The present study explored the contribution of uPAR to the development of experimental colitis by characterising its expression and function in the inflamed intestine and evaluating the effects of uPAR deficiency on gut inflammation. Herein, we show that during colitis development uPAR expression is induced in macrophages where it exerts a protective role by controlling macrophage responses against the intestinal flora.

Materials and methods

Human tissues

Actively inflamed and non-inflamed intestinal tissue specimens were obtained from patients with CD and UC receiving surgery or colonoscopy. Normal tissue specimens (tumour-uninvolved areas) were obtained from patients receiving bowel resection for colon cancer. Clinical disease activity was assessed by the Harvey–Bradshaw Activity Index and the Colitis Activity Index, as previously reported.20 All diagnoses were confirmed by clinical, radiologic, endoscopic and histological criteria. Tissues were frozen on dry ice and stored at -80°C for western blot analyses, or were embedded in optimum cutting temperature (OCT) compound for immunofluorescence studies or used immediately for isolation of LP mononuclear cells, as described in the online supplementary section. Human studies were approved by the ethical committee of Humanitas Clinical and Research Center (authorisation no. 1049, Humanitas Clinical and Research Center, Milan, Italy).

Mice

uPAR knockout (KO) mice21 were kindly provided by Dr Nicolai Sidenius (IFOM-IEO, Milan, Italy) and maintained on C57BL6/N genetic background. Animals were kept under specific pathogen-free conditions. Six- to ten-week-old, age-matched and sex-matched wild-type (WT) and KO littermates were used for all experiments. Typically, 4–6 mice were used for each group of treatment, and the experiment was repeated at least three times. Animal experiments adhered to the requirements of the Commission Directive 86/609/EEC and to the Italian legislation (Decreto Legislativo 116; 27 January 1992). The studies were approved by the Animal Care and Use Committee (authorisation no. 192/2012-B, Humanitas Clinical and Research Center, Milan, Italy).

Mouse models of colitis and colitis evaluation

Colitis was induced in C57BL/6 mice as previously described.20 Details on the experimental models are elucidated in the online supplementary section.

Isolation and stimulation of murine macrophages

Macrophages were obtained from either intestinal mucosa (LP-MΦ) or peritoneal cells (PEC-MΦ) as described in the online supplementary section.

RNA isolation and qRT-PCR

RNA was isolated from colons of healthy and colitic WT and KO mice. Biopsies were weighed and homogenised using Tissue Lyser II (Qiagene) following the manufacturer's instructions. Total RNA was extracted using RNeasy Lipid Tissue kit (Qiagen), treated with DNase I and retro-transcribed with Reverse Transcription Reagent and random primers (Applied Biosystems). Total RNA was also extracted from macrophage cultures using TRiZol Reagent (Ambion), following the manufacturer's instructions. The sequences of primer pairs used are provided in the online supplementary section.

Evaluation of uPAR protein expression in colon biopsies

uPAR protein expression was evaluated in tissue specimens from patients with IBD and in colon sections from healthy and colitic mice. Details are provided in the online supplementary section.

Cytokine measurements

Proteins were extracted from colons of healthy and colitic mice as described above. The presence of IL-4, IL-6, IL-10, IL-13, IL-17, TGF-β, TNF-α and IFN-γ was quantified by ELISA following the manufacturer's instructions (Duo Set ELISA, R&D systems). The same pattern of cytokines along with the chemokines MIP1α and 1β was also evaluated in the supernatants of macrophage cultures, treated as described in the online supplementary section.

Flow cytometer analyses

Flow cytometer analyses were conducted in LP-MC isolated from colon biopsies of either human or murine origin. Details on the staining procedure are described in the online supplementary section.

Phagocytosis and bacterial killing

Phagocytosis and bacterial killing activities were evaluated in WT and KO LP-MΦ using the Gentamycin Protection Assay. In parallel, the assay was also developed in macrophages from peritoneal cells (PEC-MΦ). Details are described in the online supplementary section.

Statistical analyses

Statistical significance was evaluated by the non-parametric, two-tailed Mann–Whitney U test for the analysis of variables that were not normally distributed. Statistical significance was defined as p<0.05.

Results

uPAR expression in experimental models of colitis

We first investigated the expression pattern of uPAR in the colon of healthy and colitic mice using either dextran sodium sulfate (DSS) or 2,4,6 trinitrobenzene sulfate (TNBS) administration, two widely used models of acute intestinal inflammation.

uPAR protein expression was increased by both treatments (figure 1) in a time-dependent manner that mirrored colitis onset. In particular, DSS administration increased uPAR protein levels at day 6 and day 9. Similarly, uPAR protein expression peaked after 48 h from TNBS instillation and subsequently decreased after 72 h (figure 1A), reflecting the recovery of body weight. To evaluate whether colitis-induced uPAR protein upregulation was due to either reduced degradation or increased synthesis, we assessed mRNA levels using quantitative PCR experiments. uPAR mRNA levels reflected protein expression in both the DSS and TNBS models (figure 1B). Overall, these data suggest that de novo synthesis of uPAR is induced in the inflamed colon.

Figure 1

Macrophages drive increased urokinase plasminogen activator receptor (uPAR) expression during experimental colitis. Colitis was induced in C57BL/6 mice as described in ‘Materials and methods’. At the indicated time points, mice were sacrificed and colon excised in order to obtain both mRNA and proteins. uPAR expression was estimated by western blot analyses (A) and quantitative RT-PCR (B). Representative images of immunofluorescence analyses of colonic mucosa in healthy and colitic mice are showing F4/80+ cells (C, in red), CD11c+ cells (D, in red) and uPAR+ cells (panels C and D, in green). Co-localisation of uPAR with either the macrophage or the dendritic cell marker is shown in merged pictures. Scale bar, 40 µM.

Inflammatory cells drive uPAR expression changes

uPAR is known to be expressed by several cell types, including epithelial cells, endothelial cells, fibroblasts and leucocytes.22 In order to readily visualise which cell type drives uPAR expression in experimental colitis, we performed co-localisation analyses by confocal fluorescence.

uPAR was expressed by all the cell types observed, broadly by endothelial cells (CD31+ cells, see online supplementary figure S1A) and weakly by epithelial cells (pan-cytokeratin+ cells, see online supplementary figure S1B) and fibroblast (α-SMA+ cells, see online supplementary figure S1C). However, after colitis induction we observed a specific co-localisation of uPAR with F4/80 macrophage markers (figure 1C, merge panel) and to a lesser extent with CD11c+ cells (figure 1D, merge panel).

In order to strengthen the results obtained with immunofluorescence analyses and to accurately quantify cell-type changes in uPAR expression, we performed flow cytometry experiments. Only a small percentage of CD45 cells expressed uPAR, and this percentage did not change in colitic mice (see online supplementary figure S1D). In sharp contrast, uPAR was found to be expressed by 42.5% on average of CD45-expressing leucocytes in healthy mice, whereas this percentage significantly increased in colitic mice (72.5% for the DSS and 68.2% for the TNBS model; see online supplementary figure S1E). Altogether, these data clearly suggested that leucocytes, and in particular mononuclear phagocytes, might incite changes in uPAR expression during experimental colitis.

uPAR expression is enhanced by CX3CR1+ macrophages

In the intestine, mononuclear phagocytes could be distinguished in monocyte, macrophages and dendritic cells.23 We therefore extended our previous analyses to these different cell subpopulations by flow cytometry.

We first investigated dendritic cells, defined as CD11b+F4/80CD11c+ cells. No differences were observed between healthy and colitic mice in the relative percentage of DCs over the CD45+ cells (data not shown). Moreover, since CD103 (αE integrin) has been recently emerged as functionally relevant in dendritic cell capacity to presenting the antigen,24 we assed uPAR expression in CD103+ and CD103 DCs (figure 2A). uPAR was expressed by the entire population of CD103+ DCs, whereas it was almost absent in the CD103 ones. However, no differences were observed after colitis induction (figure 2A).

Figure 2

Urokinase plasminogen activator receptor (uPAR) expression is specifically enhanced in CX3CR1+ macrophages. Mononuclear cells were isolated from healthy and colitic mice, and then analysed for uPAR expression on dendritic cells (A), macrophages and inflammatory monocytes (B–C). The percentage of uPAR-expressing LP-DCs, LP-monocytes or LP-MΦ is plotted in the histogram graphs. Data are the mean±SD of three independent experiments, *p<0.05, **p<0.01. Representative images of immunofluorescence analyses of colonic mucosa in healthy and colitic CX3CR1gfp/− mice (D) are showing uPAR+ cells (in grey), CX3CR1+ cells (in green), and F4/80+ cells (in red). Co-localisation of uPAR with CX3CR1+ macrophage is shown in merged pictures. Scale bar, 40 µM.

We then analysed monocytes and macrophages, discriminated as CD11b+F4/80Ly6C+ and CD11b+F4/80+Ly6C cells, respectively. In colitic mice, there was a substantial recruitment of monocytes, as expected (figure 2B, healthy 4.4±0.6% vs colitic 40.5±2.5%, p<0.01). Similarly, the percentage of macrophages significantly increased in the inflamed colon (figure 2B, healthy 20.8±0.3% vs colitic 38.4±0.5%, p<0.01). We next quantified uPAR+ cells on these subsets. Interesting, the relative percentage of uPAR+ monocytes was unaffected, whereas uPAR+ macrophages were significantly increased by colitis induction (figure 2C).

The fractalkaline receptor (CX3CR1) has been recently related to regulatory macrophages.25 We therefore analysed its expression in uPAR+ macrophages by confocal microscopy. Interestingly, uPAR clearly co-localised with CX3CR1 especially after colitis induction (figure 2D, merge panel).

The absence of uPAR leads to increased susceptibility to experimental colitis

Since uPAR expression was increased in intestinal CX3CR1 macrophages, we investigated its functional role during experimental colitis in uPAR-KO mice.

We first verified whether loss of uPAR could induce changes in leucocyte populations, thus influencing the inflammatory response. To this end, we quantified the number of polymorphonuclear (PMN) cells, macrophages, dendritic cells, myeloid suppressor cells, B and T cells by FACS analysis in the intestinal mucosa and mesenteric lymph nodes (MLN). Minor differences in the relative number of each cell population were found, suggesting that the relative frequency of leucocyte populations in colonic mucosa and MLN of uPAR-KO mice may be similar to that in WT littermates (see online supplementary figure 2).

Next, we compared the susceptibility of uPAR-KO and WT littermates with the induction of DSS and TNBS colitis. In contrast to WT, uPAR-KO mice were more susceptible to DSS-induced colitis, showing severe disease and enhanced weight loss, associated with higher DAI (figure 3A–B). In addition, uPAR-KO mice displayed significantly severe mucosal lesions at endoscopy compared with WT (figure 3C–D), whereas no differences were observed between WT and uPAR-KO mice at baseline (figure 3C–D). At the time of sacrifice, uPAR-KO mice showed increased colon shortening (figure 3D) and colons displayed more severe signs of macroscopic inflammation and histopathological alterations (figure 3C–D). Similarly, when we compared the susceptibility to TNBS-induced colitis in uPAR-KO and WT mice, we found that KO animals experienced severe disease, with significantly increased body weight loss (see online supplementary figure S4A), higher endoscopic and histopathological alterations (see online supplementary figure S4B–C).

Figure 3

Urokinase plasminogen activator receptor-knockout (uPAR-KO) mice display augmented susceptibility to DSS-induced colitis. uPAR-KO mice and wild-type (WT) littermates received 2% DSS in drinking water ad libitum. Colitis induction was evaluated by body weight loss (A), expressed as percentage of the initial weight, and clinical disease activity index (B, DAI). Representative pictures of endoscopic and histological analyses are shown in (C) and are quantified in (D). All data are expressed as mean±SD of three independent experiments. At the end of treatment (day 9), the severity of colitis was assessed also by evaluating colon shortening (D). *p<0.05, **p<0.01.

uPAR expression by bone marrow-derived cells plays a dominant role in regulating intestinal inflammation

To confirm the involvement of uPAR in driving leucocyte activities during intestinal inflammation, we generated four different groups of chimeric mice, namely, WT(WT), KO(KO), WT(KO) and KO(WT). The genotype inside the brackets indicates the source of bone marrow cells, while the one outside indicates the recipient mice (figure 4). Four weeks after bone marrow cell injection, mice were reconstituted and underwent DSS treatment. The absence of uPAR expression in the transplanted bone marrow cells determined an increased susceptibility to colitis, independently of the genetic background of the recipient, as observed by body weight loss percentage (figure 4A), disease activity index (figure 4B) and histological score (figure 4C–D). Interestingly, we also observed a decrease of the inflammatory parameters in chimeric mice that express uPAR on bone marrow cells, even if the recipient do not express the receptor (KO(WT) group, figure 4). Altogether, these data confirmed that uPAR activities by leucocytes control intestinal inflammation.

Figure 4

Urokinase plasminogen activator receptor (uPAR) expression on bone marrow-derived cells plays a dominant role in DSS-induced colitis. Recipient wild-type (WT) mice and knockout (KO) littermates were irradiated as indicated in ‘Materials and methods’. Chimeric mice were then generated by injection of WT or KO-derived bone marrow cells into the retroorbital venous plexus. The genotype inside the bracket indicates the source of bone marrow cells. Four groups were obtained: WT(WT) chimera (black squares, grey dotted line), KO(KO) chimera (red squares, black dotted line), WT(KO) chimera (red circles, black line) and KO(WT) chimera (black circles, grey line). Mice were then subjected to DSS-induced colitis and followed in time to evaluate weight loss percentage (A) and disease activity index (B). Representative pictures of histological analyses are shown in (C) and are quantified in (D).

uPAR-KO mice display an augmented mucosal production of Th1 cytokines

Because the DSS and TNBS models of colitis are associated with a skewed cytokine profile, we also evaluated whether loss of uPAR could affect the secretion of pro-inflammatory and regulatory cytokines. We first measured the amounts of IL-6, TNF-α and IFN-γ produced by colonic tissues before and after intestinal injury. Comparable levels of these cytokines were found in WT and uPAR-KO mice at baseline (figure 5A–C). By contrast, treatment with DSS induced a remarkable increase of pro-inflammatory cytokine secretion in uPAR-KO mice relative to WT mice (figure 5A–C). Similar results were obtained after TNBS instillation (figure 5A–C). We then evaluated the production of IL-4, IL-13 and IL-17 (figure 5D–F). Similar amounts of Th2 cytokines were observed in WT and uPAR-KO mice before and after colitis induction. Differently, the release of IL-17 was significantly increased in uPAR-KO mice, as especially observed after DSS administration (figure 5F). Finally, we quantified the amount of regulatory cytokines such as TGF-β and IL-10 (figure 5G–H). Mucosal production of TGF-β was almost unaffected, whereas the release of IL-10 was significantly higher in WT mice both after DSS treatment and TNBS instillation.

Figure 5

Urokinase plasminogen activator receptor-knockout (uPAR-KO) mice produce higher levels of Th1 pro-inflammatory cytokines after DSS- and TNBS-induced colitis. Wild-type (WT, grey columns) and uPAR-KO (KO, black columns) littermates were treated as indicated in ‘Materials and methods’. The mucosal production of the Th1 cytokines (TNF-α, A; IFN-γ, B; and IL-6, C) was evaluated by ELISA. Similarly, the amount of IL-4 (D), IL-13 (E), IL-17 (F) and of the regulatory cytokines (TGF-β, G; and IL-10, H) was quantified in the protein mucosal extracts obtained from healthy and colitic WT and KO mice. The graphs depict the mean±SD of three independent experiments. *p<0.05, **p<0.01.

Taken together, these data show that the release of Th1 pro-inflammatory cytokines is strongly augmented in uPAR-KO mice during colitis induction.

In intestinal macrophages uPAR deficiency amplifies the release of pro-inflammatory cytokines and influences polarisation

Since regulatory macrophages were the main cell type expressing uPAR within the inflamed gut, we reasoned that both increased susceptibility to colitis and higher production of inflammatory cytokines at mucosal sites might be attributed to macrophage dysfunction. To assess the role of uPAR in macrophage activity during colitis, we isolated macrophages from the colon of WT and uPAR-KO littermates (LP-MΦ) before and after DSS-induced colitis.

We first evaluated the release of both pro-inflammatory and regulatory cytokines in the supernatants of LP-MΦ obtained from the inflamed colon. Interestingly, the release of IL-6, TNF-α and IL-1β was significantly higher in KO LP-MΦ (figure 6A–C) that instead produced a lower amount of IL-10 (figure 6D). We also quantified IFN-γ, IL-17, IL-4 and IL-13, but the levels were undetectable (data not shown). In sharp contrast, LP-MΦ released a high amount of the chemokines MIP-1α and MIP-1β (figure 6E–F), but no differences were observed between KO and WT macrophages.

Figure 6

Urokinase plasminogen activator receptor (uPAR) is required to dampen pro-inflammatory cytokine production and to control polarisation in intestinal macrophages. LP-MΦ were obtained from either wild-type (WT, grey columns) or uPAR-knockout (KO, black columns) littermates, as described in ‘Materials and methods’. Production of TNF-α (A), IL-1β (B), IL-6 (C), IL-10 (D), MIP1α (E) and MIP1β (F) was evaluated in the supernatants of macrophage cultures, obtained from colitic mice. The expression of M1 polarisation markers (iNOS and IL-12p40, G) and M2 polarisation markers (IL-10 and MR, H) was evaluated by quantitative PCR. The graphs depict the mean±SD of three independent experiments. *p<0.05, **p<0.01.

Macrophages participate in immune response to environmental cues in a polarised manner.26 We therefore analysed iNOS and IL12p40 mRNA levels (figure 6G), as classic M1 polarisation markers while MR and IL-10 were allowed for M2 polarisation (figure 6H). Of note, KO LP-MΦ showed a significantly higher amount of iNOS and IL12p40, as especially observed in colonic macrophages from inflamed mice.

Altogether, these data clearly indicated that in the experimental model of colitis observed, uPAR controls intestinal macrophages function by dampening pro-inflammatory cytokine production and controlling M1 versus M2 polarisation.

uPAR is required by intestinal macrophages to control in vitro the release of pro-inflammatory cytokines and macrophage polarisation.

In IBD pathogenesis, macrophages are exposed to an enriched bacterial environment.27 We therefore verified whether the functional differences observed between WT and KO macrophages are due to an impaired tolerance to bacterial products.

uPAR is regulated by several TLR ligands,8 ,28 and therefore we first analysed whether incubation with LPS or heat-inactivated Escherichia coli was able to induce uPAR expression in intestinal macrophages. Remarkably, both LPS and E. coli were similarly able to significantly increase uPAR expression in LP-MΦ, as observed by PCR analyses (see online supplementary figure S4A). Similar results were obtained with PEC-MΦ, even though a stronger effect was observed upon LPS stimulation (data not shown). Moreover, to assess in vitro the role of uPAR in cytokine production by macrophages, we evaluated IL-6, TNF-α, and IFN-γ production in the supernatant of LP-MΦ cultures. Interestingly, upon LPS or E. coli stimulation no difference in TNF-α production was observed between WT and KO macrophages (data not shown). In sharp contrast, KO macrophages produced a significantly lower amount of IL-10 (see online supplementary figure S4B) and a higher amount of IL-6 and IFN-γ (see online supplementary figure S4C–D).

Finally, we analysed M1 and M2 polarisation marker, as previously done in macrophages from colitic mice. As expected, WT LP-MΦ were poorly responsive to LPS stimulation, as especially observed in the induction of classic M1 polarisation markers. In contrast, KO LP-MΦ were highly responsive to LPS, leading to increased expression of both iNOS and IL-12p40 mRNA (see online supplementary figure S4E). Moreover, LPS induced the expression of MR and IL-10 in WT LP-MΦ but not in KO LP-MΦ (see online supplementary figure S4E).

Taken together, these data confirmed that intestinal macrophages exposed to bacterial products required uPAR to control the release of cytokines and macrophage polarisation.

uPAR is required for controlling bacterial phagocytosis via MAC-1 interaction and RAC-1 activation

It has recently been shown that uPAR is required for the recognition and phagocytosis of several species of bacteria by macrophages.5 ,12 ,29 To investigate whether the loss of uPAR affects these functions in the intestine, LP-MΦ or PEC-MΦ were infected for 1 h with heat-inactivated E. coli and, following gentamicin treatment, viable intracellular bacteria colony forming units (CFUs) were counted after 10, 30, 60 and 120 min to evaluate bacterial phagocytosis and killing activity. We found that WT and uPAR-KO MΦ differed in their ability to phagocytose E. coli (Figure7A and see online supplementary figure S5A) and partially in their capacity to kill bacteria (figure 7B and see online supplementary figure S5B).

Figure 7

Urokinase plasminogen activator receptor (uPAR) is required for intestinal macrophage phagocytic activity. Bacterial phagocytosis (A) and bacterial killing (B) were evaluated by Gentamycin Protection Assay in wild-type (WT, grey) and uPAR-knockout (KO, black) LP-MΦ after infection with heat-inactivated Escherichia coli at the indicated time points. uPAR interaction with CD11b and RAC-1 activation (C) was evaluated after stimulation with either LPS or heat-inactivated E. coli (1 µg/mL for LPS, MOI 1:10 for E. coli, 30 min) by immunoprecipitation. CD11b interaction and RAC-1 activation (D) were evaluated in the presence or absence of α-325 (75 µM) using a scramble (s-325) as control. Phagocytosis was also evaluated by GPA assay in WT LP-MΦ in the presence or absence of -325 (E). The bacterial amount in colon tissue and mesenteric lymph node was evaluated by CFU rescue in both colitic and healthy WT and uPAR-KO mice (F). All data are represented as mean±SD of three independent experiments. *p<0.05, **p<0.01.

Phagocytosis is a complex multifactorial event that requires the modification of cytoskeletal proteins mediated by intracellular signalling, involving RAC-1 activation.30 In addition, integrins play a central role in the phagocytic process31 ,32 and it has been shown that integrin MAC-1 (CD11b/CD18) interacts with uPAR, forming a functional unit in monocytes.33 We therefore hypothesised that uPAR may modulate integrin activity and signalling in intestinal MΦ during the phagocytic process. We thus investigated whether uPAR was able to interact with MAC-1 in LP-MΦ after stimulation with LPS or heat-inactivated E. coli. We found that uPAR clearly interacted with CD11b in WT LP-MΦ upon both LPS and E. coli stimuli, as observed by immunoprecipitation (figure 7C) and immunofluorescence (see online supplementary figure S5C). Moreover, in the absence of uPAR, such stimuli did not lead to RAC-1 activation, in contrast to what we observed in WT LP-MΦ (figure 7C). Furthermore, LP-MΦ were not able to activate RAC-1 after incubation with alpha 325, an inhibitor of uPAR–integrin interaction (figure 7D), and displayed an impaired E. coli phagocytosis (Figure. 7E). Finally, when the ability to phagocytose bacteria in the absence of uPAR was investigated in vivo, CFU rescue was significantly higher both in colon and MLN of KO mice (figure 7F), especially after colitis induction. These observations were also confirmed by FISH analyses (see online supplementary figure S5D).

Taken together, these data clearly indicated that the phagocytic process of bacterial products required uPAR expression, allowing interaction with MAC-1 and activation of RAC-1.

uPAR expression in patients with IBD

Based on the above observations, we reasoned that uPAR may exert protective roles in intestinal inflammation and therefore we predicted that patients with IBD would have decreased levels of uPAR.

uPAR is a cell-surface receptor composed of three domains (D1–D3). In pathological conditions, several proteases can cleave uPAR,34 leading to the generation of a truncated cell-surface form and the release of a soluble form (suPAR). The truncated form of the receptor is unable to interact with integrins,34 while suPAR is used as a clinical marker for several pathologies.34 ,35 Since western blot analyses could detect both the intact and cleaved forms, we determined uPAR protein expression. Interesting, the results obtained showed that in actively inflamed mucosa samples the total amount of uPAR increased. However, the intact form of uPAR is dramatically downregulated with a parallel significant increase of the cleaved uPAR (figure 8A). This is in agreement with the results obtained by Lönnkvist and Kolho that showed suPAR levels are increased in the blood of IBD patients.18 ,19 Finally, we looked at uPAR expression specifically in LP-MΦ cells by both immunofluorescence and flow cytometry. Immunofluorescence data clearly indicated that uPAR expression is macrophage specific, also in the human intestine (figure 8B). Moreover, flow cytometry analyses showed that macrophages from the inflamed area of patients with IBD tend to lose uPAR expression (figure 8C, CD non-inflamed=51.2±5.2 vs CD inflamed=15.9±6.7, p<0.01; UC non-inflamed=56.8±12.2 vs UC inflamed=38.9±11.7, p<0.05).

Figure 8

Urokinase plasminogen activator receptor (uPAR) is specifically downregulated on the macrophage surface of patients with IBD. uPAR protein expression was by western blot (A). Numbers overlying the blots represent densitometric values of uPAR total amount. The graphs represent mean±SD of the densitometry analyses for the intact and cleaved form of uPAR, respectively (*, p<0.05). Results are representative of four patients with Crohn's disease (CD). Immunofluorescence analysis (B) of colonic mucosa showed CD68+ cells depicted in red, uPAR+ cells in green. Merged pictures show antigen co-localisation in yellow. The panels are representative of four controls, four CD and four UC patients. The images were acquired with an oil immersion objective (40×, 1.4 NA Plan-Apochromat; Olympus). Scale bar, 40 μm. Finally, flow cytometry studies (C) were conducted in LP cells, isolated as described in the supplementary section. LP-MΦ were gated as CD45+CD68+CD3. uPAR-positive macrophages were quantified by ATN658 antibody and compared with IgG-isotype control. The graphs depict the mean±SD of five CD and five UC patients.

Discussion

This study explored for the first time the contribution of uPAR to the pathogenesis of intestinal inflammation and unravelled a previously unrecognised role for this receptor in controlling intestinal macrophage functions.

Degradation of extracellular matrix and ulceration of the mucosa are major features of IBD.2 Among the factors controlling ECM degradation, uPAR has emerged as a crucial player since its main function is to stimulate uPA proteolysis at the leading edge of cells.4 ,7 Additionally, uPAR is induced during tissue remodelling36 insinuating the hypothesis that it could control intestinal epithelial barrier integrity. However, our data clearly showed that during experimental colitis uPAR was poorly expressed in the epithelial layer overlying the mucosa. These results were in agreement with data provided by Gibson and colleagues,16 which did not find any differences in uPAR expression levels in the colonic crypt of IBD patients compared with healthy controls.

The observation that uPAR is expressed by cells of haematopoietic lineage4 ,13 ,28 supports the idea that this receptor modulates innate and adaptive immune response independently of its role in controlling ECM degradation. Indeed, it has been shown that uPAR participates in the initiation of the innate immune response by regulating cell adhesion and migration.37 Additionally, in uPAR-deficient mice, macrophages and neutrophils failed to infiltrate the lungs of mice infected with Streptococcus pneumonia6 or Pseudomonas aeruginosa,12 or to migrate to the inflamed peritoneal cavity of thioglycollate-treated mice.9

Therefore, uPAR might favour innate immune responses mostly by promoting inflammatory cell activation and migration rather than through its fibrinolytic function.

Consequently, we reasoned that also during IBD pathogenesis uPAR might controls immune cell activities. Recently, Ly6Chi monocytes have been recognised as precursors of either regulatory macrophages or pro-inflammatory DCs in the intestine, depending on the local environmental conditions.38 Moreover, a new organisational scheme of intestinal DCs and macrophages has been proposed, in relationship to the expression of the non-overlapping marker CD103 (αE integrin) and of CX3CR1 (fractalkaline receptor).24 ,25 ,38 Remarkably, we proved that even if uPAR is broadly expressed by different cell types, during intestinal inflammation it is specifically enhanced in CX3CR1+ regulatory macrophages, crucially involved in sampling bacteria and with a distinct anti-inflammatory phenotype.23 ,25

To further discern the functional role of uPAR in intestinal inflammation, we took advantage of uPAR-KO mice, undoubtedly proving that the absence of uPAR leads to an augmented susceptibility to both DSS and TNBS models of colitis. Furthermore, we verified that uPAR function is strictly related to its expression in the leucocyte compartment. Indeed, we demonstrated that bone marrow chimera with uPAR expression maintained in the recipient tissue but not in the donor bone marrow cells displayed a severe intestinal inflammation. Similarly, uPAR expression in the bone marrow cells was sufficient to reduce intestinal inflammation even if the receptor was not expressed in the host tissue.

An increasing number of studies are elucidating the role of the intestinal macrophages in the gut since it has been shown that in this specific milieu they displayed a distinct phenotype with a profound inflammatory anergy despite avid phagocytic and bactericidal activity.27 ,39 ,40 Additionally, macrophages control immune response by polarisation, which in turn regulates their activities and functions.26

In agreement with the functional characterisation of uPAR as a protective molecule expressed by CX3CR1+ macrophages, we found that its absence in LP-MΦ amplifies the release of Th1 pro-inflammatory cytokines, dampens the production of IL-10 and influences macrophages polarisation. These activities were underlined ex vivo after colitis induction and in vitro after exposure to LPS or E. coli.

Indeed, we observed that the absence of uPAR in LP-MΦ affects their tolerance to LPS or E. coli, with a significantly higher release of pro-inflammatory cytokines and inducing a classical M1 polarisation. Additionally, uPAR-KO LP-MΦ are defective in phagocytosis due to the missing interaction between uPAR and MAC-1, which leads to inhibition of RAC-1 activation, an important step in the phagocytic process. This was further strengthened by in vivo data showing that both healthy and colitic uPAR-KO mice had an augmented bacterial flora both in the colon and the mesenteric lymph node.

Therefore, it is reasonable to expect that uPAR deficiency might alter microbiota composition, conferring different susceptibility to the experimental colitis models. However, cross-fostering and/or co-housing experiments are required to confirm or confute this intriguing possibility.

Altogether, these results suggest that uPAR exerts a protective role during experimental intestinal inflammation. In agreement with these findings, in mucosal biopsies of patients with CD and UC, uPAR expression was significantly downregulated on the surface of intestinal macrophages, suggesting that low uPAR expression might lead to an impaired bacterial handling and increased M1 inflammatory phenotype.

The results presented herein point to the urokinase receptor as a novel molecule involved in IBD pathogenesis. Modulating uPAR activities in the gut could be beneficial to dampen intestinal inflammation through effects on the innate immune response and on bacterial handling.

Acknowledgments

The authors wish to thank Dr Marco Erreni and Dr Giulia Marelli (Humanitas Research Hospital) for kindly providing healthy and colitic CX3CR1 GFP/− colonic specimens.

References

Footnotes

  • Contributors MG: conception and design, acquisition, analysis and interpretation of data. SD'A: conception and design, analysis and interpretation of data. JC, AG, ES and VA: acquisition of data. AS, CC: acquisition, analysis and interpretation of data. AM: revising the article critically for important intellectual content. SR and VAP: drafting the article and revising it critically for important intellectual content. SV: conception and design, analysis and interpretation of data. SD: conception and design, drafting the article and revising it critically for important intellectual content.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Funding Innovative Medicines Initiative IMI-funded project “BeTheCure” (#115142-2) Associazione Italiana per la Ricerca sul Cancro—AIRC Investigator Grant (#IG-14490) and the Eli and Edythe Broad Foundation (BMRP#IBD-0345R2).

Linked Articles

  • Correction
    BMJ Publishing Group Ltd and British Society of Gastroenterology