Article Text

Download PDFPDF

Impact of toll-like receptor 4 on the severity of acute pancreatitis and pancreatitis-associated lung injury in mice
  1. R Sharif1,
  2. R Dawra1,
  3. K Wasiluk1,
  4. P Phillips1,
  5. V Dudeja1,
  6. E Kurt-Jones2,
  7. R Finberg2,
  8. A Saluja1
  1. 1
    Department of Surgery, University of Minnesota, Minneapolis, Minnesota, USA
  2. 2
    Department of Medicine, University of Massachusetts, Worcester, Massachusetts, USA
  1. Professor A K Saluja, MMC 195, 420 Delaware Street SE, Minneapolis, MN 55455, USA; asaluja{at}


Background and Aims: Acute pancreatitis is an inflammatory disease involving acinar cell injury, and the rapid production and release of inflammatory cytokines, which play a dominant role in local pancreatic inflammation and systemic complications. Toll-like receptor 4 (TLR4) initiates a complex signalling pathway when it interacts with lipopolysaccharide (LPS), which ultimately results in a proinflammatory response. We hypothesised that TLR4 is important in the pathophysiology of acute pancreatitis, independently of LPS. Using two different models of acute pancreatitis, we investigated how genetic deletion of TLR4 or its co-receptor CD14 effects its progression and severity.

Methods: We induced acute pancreatitis by administering either caerulein or l-arginine to wild-type, TLR4−/−, and CD14−/− mice. Control mice received normal saline injections. The severity of acute pancreatitis was determined by measuring serum amylase activity, quantifying myeloperoxidase (MPO) activity in the pancreatic tissue, and histologically assessing acinar cell injury.

Results: It was found that administering caerulein and l-arginine to wild-type mice resulted in acute pancreatitis (as assessed by hyperamylasaemia, oedema, increased pancreatic MPO activity, and pancreatic necrosis) and associated lung injury. The same treatment to TLR4−/− or CD14−/− mice resulted in significantly less severe acute pancreatitis, and reduced lung injury. We found no evidence of either bacteria or LPS in the blood or in pancreatic tissue.

Conclusions: The severity of acute pancreatitis is ameliorated in mice that lack either TLR4 or CD14 receptors. Furthermore, these results indicate that TLR4 plays a significant pro-inflammatory role independently of LPS in the progression of acute pancreatitis.

View Full Text

Statistics from

Acute pancreatitis is a complex disease with variable severity and an overall mortality rate of 10% to 15%.1 Infections of the inflamed pancreas and surrounding tissues are the most common cause of morbidity and mortality. The underlying mechanisms of the progression of acute pancreatitis have not yet been clearly defined, but the current understanding is that the earliest events include premature intra-acinar activation of digestive zymogens. Once activated, these enzymes cause acinar cell injury.2 Previous studies suggest that the severity and outcome of the resulting pancreatitis might be determined by events that occur subsequently to acinar cell injury. Such events include the activation of transcription factors such as nuclear factor kapp B (NF-κB); the generation and release of cytokines and other mediators of inflammation; and the recruitment and activation of inflammatory cells.3 4 Nonetheless, the exact mechanisms by which inflammation develops are still unknown. It is therefore critical to investigate the molecular mechanisms of this complex disease process, especially those which activate the innate immune response.

Toll-like receptor 4 (TLR4) is one of at least 11 mammalian pattern-recognition receptors that comprise the innate immune response. These receptors recognise conserved products of microbial metabolisms (pathogen-associated molecular patterns, or PAMPs). Several TLRs recognise different PAMPs, and thereby provide the initial specificity level of the innate immune response.5 TLR4 is essential for recognising lipopolysaccharide (LPS); mice whose TLR4 gene is either mutated or missing are hyporesponsive to LPS and yet highly susceptible to infection by Gram-negative bacteria.68 However, TLR4 is not the only protein found in the LPS receptor complex, which also includes myeloid differentiation protein 2 (MD-2), lipopolysaccharide-binding protein (LBP), and CD14. LPS is bound in serum by LBP, which delivers LPS to CD14, a glycosyl phosphatidylinositol (GPI)-linked membrane protein that exists in a soluble form and in a membrane-bound form, both of which appear to potentiate responsiveness to LPS.9 10 Binding LPS to the LPS receptor complex of macrophages results in the upregulation of NF-κB and the eventual release of proinflammatory cytokines such as tumour necrosis factor α (TNFα), interleukin 1β (IL1β), IL6 and IL8. TLR4 is the protein within the LPS receptor complex through which intracellular signalling occurs in response to binding LPS.

Although CD14 is not the signalling component of the LPS receptor complex, its two forms (soluble and membrane-bound) are important in the host response to LPS. CD14 has also been implicated in severe pancreatitis.11 CD14 expression is not restricted to myeloid cells; however, its function in non-myeloid cells is still unknown, and the level of expression of CD14 appears to be influenced by exposure to LPS.12 In addition to being a major LPS-binding receptor, CD14 can serve as a ligand for other PAMPs, is important in the activity of other TLRs,13 and has been implicated in the heat shock protein response.14 15

The aims of our current study were to determine (1) whether TLR4 receptors are important to the pathological progression of acute pancreatitis and its associated lung injury; and (2) whether LPS is involved in the activation of TLR4 during acute pancreatitis.


Caerulein, the decapeptide analogue of the potent pancreatic secretagogue cholecystokinin (CCK), was purchased from Bachem Peninsula Labs (San Carlos, California, USA); collagenase from Worthington Biochemical (Freehold, New Jersey, USA); amylase reagent from Diagnostic Chemicals (Charlottetown, Canada); protein assay reagent from Thermo Electron (Louisville, Colorado, USA); protease inhibitor cocktail tablets from Roche Diagnostics (Mannheim, Germany); and all other chemicals and reagents from Sigma Chemical (St. Louis, Missouri, USA).


Homozygous TLR4-deficient (TLR4−/−) mice or CD14-deficient (CD14−/−) mice were backcrossed to a C57BL/6 background, for >10 generations.16 The mice were housed and bred in microisolators under specific-pathogen free conditions, in a climate-controlled room with an ambient temperature of 22 (SD 1)°C and a 12:12 h light/dark cycle. They were fed standard laboratory chow, and drinking water was available ad libitum. All studies were performed using wild-type, TLR4−/−, and CD14−/− mice that weighed between 25 and 30 g.

Induction of pancreatitis

Secretagogue-induced pancreatitis was developed by administering hourly intraperitoneal injections of a supramaximally stimulating concentration of caerulein (50 μg/kg in a total volume of 0.2 ml) for 12 h to wild-type, TLR4−/− or CD14−/− mice. Control mice of all strains received injections of normal saline (0.2 ml) at the same time intervals.17 One hour after the final caerulein or saline injection, mice were killed by CO2 asphyxia and blood and tissue samples were rapidly collected for analysis. Pancreatitis was induced in wild-type and TLR4−/− mice by giving two intraperitoneal injections of l-arginine, each at concentrations of 400 mg/100 g body weight, with a 1 h interval between injections. Mice were killed 72 h after the second injection.18

Blood and tissue culture

Blood and pancreatic tissue were collected aseptically. Harvested blood was cultured directly on blood agar plates in order to detect the presence of microbes, and incubated them at 37°C for 12 and 24 h. The blood agar plates were then observed for microbial growth. Pancreatic tissue was homogenised in sterile saline and aliquots of the homogenate were plated on blood agar. For histology, pancreatic tissue was fixed in 10% neutral phosphate-buffered formalin. Pancreatic and lung tissue samples were snap-frozen (for later measurement of tissue myeloperoxidase (MPO) activity) in liquid nitrogen, and stored at −80°C until analysis. Part of the harvested pancreatic tissue was used to measure oedema according to methods described in a previous study.2

Morphological examination

After fixation, 5-μm sections of pancreatic and lung tissue samples were stained with haematoxylin/eosin, and subsequently examined by an experienced morphologist who was not aware of their identity. The extent of acinar cell injury or necrosis was quantified by computer-assisted morphometry, as previously described.19 For the morphological examination, 10 microscopic fields (×100) were randomly chosen for each tissue sample, and the extent of acinar cell injury or necrosis in each sample was expressed as a per cent of the total acinar tissue.

Preparation and use of acini

Dispersed pancreatic acini were prepared from wild-type, TLR4−/−, and CD14−/− mice by collagenase digestion, as previously described.20 Acini were suspended in oxygen-saturated HEPES–Ringer buffer (pH 7.4) containing 0.1% bovine serum albumin. Viability of the acini was assessed by trypan blue exclusion and preparations with viability of >95% were used for further studies. Afterwards, the secretion of pancreatic acini was evaluated from wild-type, TLR4−/−, and CD14−/− mice in response to caerulein (concentration from 10–12 to 10–7 mol/l) by measuring amylase in incubation buffer over 30 min at 37°C. Amylase secretion was expressed as a percentage of total amylase in acini.


Pancreatic fragments were harvested and weighed to determine the tissue wet weight. Those samples were then desiccated by incubating them in a drying oven overnight, and reweighed to determine tissue dry weight. Tissue water content was calculated as the difference between wet and dry weight, and expressed as a percent of wet weight. Pancreatic oedema in our model of l-arginine-induced acute pancreatitis was calculated by a morphometric examination of the interstitial space. Serum amylase activity was quantified using 4,6-ethylidene-(G7)-1,4-nitrophenyl-(G1)-α-d-maltoheptaoside as substrate, as previously described.21 The extent of PMN sequestration was measured in both pancreatic and lung tissue by quantifying MPO, as previously described.22 The protein in the supernatant was quantified using a microprotein assay kit (Thermo Electron) according to the manufacturer’s instructions, and MPO activity expressed per 50 μg protein. LPS was quantified using a standard, commercially available QCL-1000 chromogenic Limulus amebocyte lysate assay kit (BioWhitaker, Walkersville, Maryland, USA) according to the manufacturer’s instructions.

Statistical analysis

Results are expressed as of the mean with the SEM for multiple determinations from at least three or more separate experiments. The significance of changes was analysed using the Student t test when comparing only two groups, or by analysis of variance (ANOVA) when comparing three or more groups, with significance at p<0.05. The photographs shown represent the results obtained from at least three independent experiments.


Effect of TLR4 deletion on caerulein-induced pancreatitis and associated lung injury

All parameters used to quantify the severity of acute pancreatitis were reduced in TLR4−/− mice as compared with wild-type mice. Serum amylase activity (fig 1A) and oedema (fig 1B) were decreased in TLR4−/− mice as compared with wild-type mice. There was also a marked reduction in neutrophil sequestration as evidenced by decreased pancreatic tissue MPO activity (fig 1C) in TLR4−/− mice as compared with wild-type mice. Quantitative analysis of acinar cell injury or necrosis revealed significantly less necrosis in TLR4−/− mice as compared with wild-type mice (fig 1D). Morphological changes seen in pancreatitis included acinar cell vacuolisation, oedema and acinar cell necrosis, all of which were significantly less severe in TLR4−/− mice as compared with wild-type mice (fig 2). In lungs of TLR4−/− mice, there was significantly less MPO activity as compared with wild-type mice, indicating reduced sequestration of neutrophils (fig 3).

Figure 1

Effects of TLR4 deficiency on the severity of caerulein-induced pancreatitis. Results show that serum amylase (A), pancreatic water content (B), pancreatic myeloperoxidase (MPO) activity (C), and acinar cell necrosis or injury (D) were all significantly decreased in TLR4−/− mice, as compared with wild-type mice. Results are expressed as means with the SEM of at least three separate experiments with statistical significance at *p<0.05.

Figure 2

Morphological effects of pancreatitis in TLR4−/− mice. Results demonstrate a marked reduction in acinar cell injury or necrosis in pancreatic tissue from TLR4−/− mice treated with caerulein (D), as compared with wild-type mice treated with caerulein (B). Controls were injected with saline alone for wild-type mice (A) and TLR4−/− (C). TLR, toll-like receptor.

Figure 3

Effects of deleting TLR4 on the severity of associated lung injury in caerulein-induced pancreatitis. There was significantly less lung injury, as measured by lung tissue MPO activity, in TLR4−/− mice, as compared with wild-type mice. Results are expressed as the mean with the SEM of MPO activity per 50 μg total protein of at least three separate experiments with statistical significance at *p<0.05. MPO, myeloperoxidase; TLR, toll-like receptor.

Effects of CD14 deletion on caerulein-induced acute pancreatitis

To confirm that our findings in TLR4−/− mice were not merely a phenomenon of genetically deleting an LPS-binding protein within the LPS receptor complex, the experiments were repeated with CD14−/− mice. When the parameters of severity were assessed in caerulein-induced pancreatitis both in CD14−/− mice and in their wild-type counterparts, serum amylase values were not significantly different in CD14−/− mice as compared with wild-type mice (fig 4A). However, in accordance with the data obtained with TLR4−/− mice, there was significantly less pancreatic oedema (fig 4B), significantly less neutrophil pancreatic infiltration, as reflected by MPO activity (fig 4C), and significantly less acinar cell injury or necrosis in CD14−/− compared with wild-type (fig 4D). Morphological changes associated with acute pancreatitis were less severe in CD14−/− mice as compared with wild-type mice (fig 5).

Figure 4

Effects of CD14 deficiency on the severity of caerulein-induced pancreatitis. There was no significant difference in serum amylase (A). Pancreatic water content (B), pancreatic MPO activity (C), and acinar cell necrosis (D) were all significantly reduced in CD14−/− mice, as compared with wild-type mice. Results shown are means with the SEM of at least three separate experiments with statistical significance at *p<0.05. MPO, myeloperoxidase.

Figure 5

Morphological changes of pancreatitis in CD14−/− mice. There was a marked reduction in acinar cell injury or necrosis in pancreatic tissue from CD14−/− mice treated with caerulein (D), as compared with wild-type mice treated with caerulein (B). Controls were injected with saline alone for wild-type mice (A) and CD14−/− (C).

Effects of l-arginine-induced pancreatitis in TLR4/ mice

To verify that the reduced effects of caerulein-induced acute pancreatitis seen in TLR4−/− mice were not acute pancreatitis model-specific, we examined the severity of pancreatitis induced in wild-type and TLR4−/− mice by l-arginine as previously described by our laboratory.18 Results show that there was significantly less serum amylase (fig 6A), pancreatic MPO activity (fig 6C), and acinar cell injury or necrosis in TLR4−/− mice treated with l-arginine (fig 6D) as compared with wild-type mice. Morphological changes associated with acute pancreatitis were less severe in TLR4−/− mice as compared with wild-type mice, as well (fig 7).

Figure 6

Effects of TLR4 deficiency on the severity of l-arginine-induced pancreatitis. Serum amylase (A), pancreatic myeloperoxidase (MPO) activity (C), and acinar cell necrosis or injury (D) were all significantly decreased in TLR4−/− mice as compared with wild-type mice. Results shown are means with the SEM of at least three separate experiments with statistical significance at *p<0.05. TLR, toll-like receptor.

Figure 7

Morphological effects of pancreatitis in TLR4−/− mice. There was a marked reduction in acinar cell injury or necrosis in pancreatic tissue from TLR4−/− mice treated with l-arginine (D), as compared with wild-type mice treated with l-arginine (B). Controls were injected with saline alone for wild-type mice (A) and TLR4−/− (C).

In vitro effects of caerulein-induced secretion in pancreatic acini from TLR4−/−, CD14−/− and wild-type mice

Intact CCK receptor function is critical for the induction of secretagogue-induced acute pancreatitis. In order to determine if CCK receptor function is intact in TLR4−/− and CD14−/− mice, the secretion from pancreatic acinar cells prepared from these and wild-type mice was examined. No significant difference in amylase secretion in response to the various caerulein concentrations in acini isolated from either TLR4−/− mice (A) or CD14−/− mice (B) as compared with acini isolated from wild-type mice was observed (fig 8). Our finding indicates that deleting either TLR4 or CD14 does not affect the CCK receptor function in pancreatic acini.

Figure 8

Effects of caerulein-induced amylase secretion from pancreatic acini prepared from wild-type mice, TLR4−/− mice and CD14−/− mice. There was no difference in amylase secretion in the TLR4−/− mice, except at the highest concentration of caerulein tested (0.1 μmol/l) (A). We found no difference in amylase secretion in CD14−/− mice at all concentrations (B). Results represent mean with the SEM of at least three separate experiments. TLR, toll-like receptor.

Quantification of lipopolysaccharide and bacteria

No bacterial growth was observed on any of the blood agar plates inoculated with dilutions of all samples harvested. In addition, no LPS was detected in any tissue sample higher than 1 unit of LPS/ml, using the chromogenic Limulus amebocyte lysate assay. Our findings indicate that LPS is not present in significant enough concentrations to contribute substantially to the development of acute pancreatitis, thus TLR4 and CD14 must function independently of LPS.


The innate immune system that is activated in response to a variety of microbial PAMPs has been evolutionarily conserved throughout plants, insects and vertebrates. TLRs, named for their amino acid sequence similarity to the Drosophila toll protein, function as the first line of defence against invading microorganisms.23 All TLRs initiate transmembrane signalling through a region of amino acid sequence similarity called the toll/interleukin-1 receptor (TIR) domain. This signalling results in the upregulation of NF-κB and in the eventual production of proinflammatory cytokines.24 25 Positional cloning and targeted deletion have demonstrated unequivocally that TLR4 is the specific receptor for recognition of LPS and that it requires at least two accessory proteins, CD14 and MD-2, for optimal signalling.

In our study, a model of caerulein-induced acute pancreatitis was used in strains of mice that were genetically deficient in TLR4. Results demonstrate less severe acute pancreatitis in TLR4−/− mice, as measured by decreased serum amylase, pancreatic oedema, pancreatic MPO activity; and pancreatic acinar injury or necrosis, as compared with wild-type mice. In addition, there was less severe lung injury, in TLR4−/− mice, as compared with wild-type mice.

To confirm that our findings were not unique to TLR4−/− mice, we elicited caerulein-induced acute pancreatitis in CD14−/− mice and again found less pancreatic oedema, less pancreatic MPO activity, and less pancreatic acinar injury or necrosis as compared with wild-type mice. However, we did not observe any significant change in serum amylase levels in the CD14-deleted animals. While hyperamylasaemia is associated with pancreatitis, several studies have indicated that it is not a predictor of severity. In clinical pancreatitis, for example, patients with severe disease often have lower levels of serum amylase as compared with patients with a milder form of the disease.

In addition to using two different genetically deficient strains of mice, we sought to confirm our finding that acute pancreatitis was less severe in the absence of TLR4 with a different model of acute pancreatitis, one using l-arginine. Hyperstimulation with caerulein induces relatively mild pancreatitis in mice, while the l-arginine model is more severe and the succession of events associated with the pancreatitis induced by it occurs over a longer time span (peak injury occurred in 12 h as compared with the 72 h for caerulein and l-arginine-induced pancreatitis, respectively). As shown in figs 4D and 6D, necrosis is significantly greater in l-arginine-induced pancreatitis as compared with caerulein-induced pancreatitis. Deleting TLR4 significantly protects against injury in both models, though it is a little more protective in the severe form of the disease, ie, the one induced by l-arginine. These results establish that TLR4 is a significant determinant of severity in pancreatitis, and its effect is not model-specific but common to all models tested.

Pastor et al26 used mice deficient in TLR4 expression to examine the role of TLR4 on pancreatic and pulmonary injury associated with endotoxaemia. They concluded that TLR4 may not play a role in pancreatitis-associated lung injury, but participates in the associated pulmonary injury mediated by endotoxaemia.26 Another study by Wang et al27 used pretreatment with either CD14 or TLR4 antibodies to delineate the role of endotoxin-related signalling in progression of acute pancreatitis. Our study is the first to examine a role for both TLR4 and CD14 using the appropriate knockout mice in caerulein-induced pancreatitis without LPS. Therefore, deleting either of these receptors decreases the severity of pancreatitis, no matter which model of pancreatitis is used.

For caerulein to induce acute pancreatitis, the CCK response must be intact. The data verified that deleting either TLR4 or CD14 did not alter the CCK response of pancreatic acinar cells from genetically deficient mice. In vitro secretion of amylase by acinar cells from TLR4−/− mice and CD14−/− mice was not significantly different from that of acinar cells from wild-type mice, indicating that the CCK response was intact. Our findings clearly indicate that the proinflammatory effects are mediated by TLR4 and CD14 in caerulein-induced acute pancreatitis, yet the underlying mechanisms have not been defined. We are currently designing experiments to do so.

Gram-negative bacteria contain LPS in their outer membrane, and LPS plays a key role in the initiation of the innate immune response by its specific interaction with TLR4 in septic shock and other inflammatory diseases associated with infection. We considered the possibility that LPS contributes to the severity of the caerulein-induced acute pancreatitis in our experiments, and concluded the relationship to be unlikely, because bacteria are not present in the early stages of animal models of induced pancreatitis or in clinical pancreatitis. In caerulein-induced pancreatitis, morphological changes in acinar cells appear as early as 4 h, far earlier than the appearance of bacteria due to translocation from the gut, which has been determined by others to be 24 or more hours.28 Nevertheless, to rule out the involvement of LPS in our model, tissue samples were examined for the presence of both LPS (using the Limulus amebocyte lysate assay) and bacteria (by plating blood and pancreatic tissue homogenates directly on growth media). The data show no evidence of either LPS or bacteria in the blood or the pancreatic tissue samples from any of the mice used in our experiments at 12 h.

Pancreatic acinar cell injury is considered to be a primary event in the development of acute pancreatitis. The severity of injury is the cumulative result of events originating in acinar cells and advancing into a systemic response, including the sequestration and activation of neutrophils, and the elaboration of proinflammatory cytokines. Factors that affect any of these events also influence the severity of acute pancreatitis. Administration of LPS with caerulein leads to increased severity in the caerulein-induced model of pancreatitis,29 which implicates the role of LPS-responsive receptors and their downstream signalling in pancreatitis and associated lung injury. These results do not, however, unambiguously establish the presence of TLR4 on pancreatic acinar cells. There is also some evidence to suggest that LPS can stimulate NF-κB activation by itself in isolated rat pancreatic acinar cells.30 Exposure of AR4-2J cells to LPS caused pancreatitis-associated proteins to be expressed, again indicating the presence of TLR4 on acinar cells.31 In addition, other cells like macrophages, neutrophils and monocytes are known to have TLR4 receptors and participate in pancreatic injury. The absence of microbes or endotoxins during the initial stages argues against the possibility of LPS as a TLR4 agonist during the early stages of pancreatitis, although its role is clearly established by our studies. Our data suggest that a hitherto unknown agonist for TLR4 released from acinar cells during initial injury might act via an autocrine or paracrine mode, and that the progression of inflammatory response is disrupted in the absence of receptors on acinar and other inflammatory cells. Therefore, the observed protective effect of genetic deletion of either TLR4 or CD14 in our study could be due to the decreased injury to acinar cells and/or the subsequent progression of injury mediated through macrophages and neutrophils.

Given the results of our study, we conclude that deleting TLR4 attenuates the severity of acute pancreatitis, indicating that TLR4 is pro-inflammatory during acute pancreatitis. Furthermore, our results also indicate that TLR4 in pancreatitis is also activated in the absence of LPS or bacteria. In view of these findings, we hypothesise that a novel endogenous agonist generated during injury to the pancreatic acinar cells, is involved in activation of TLR4 during acute pancreatitis.


View Abstract


  • Competing interests: None.

  • Funding: This study was supported from National Institute of Diabetes and Digestive and Kidney Diseases Grants DK058694 and DK072439 to AKS.

  • Ethics approval: All experiments with animals were performed according to protocols approved by the Institutional Animal Care and Use Committee of the University of Massachusetts Medical School at Worcester, Massachusetts.

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.

Linked Articles

  • Digest
    Robin Spiller Magnus Simren