Article Text

Flagellin administration protects gut mucosal tissue from irradiation-induced apoptosis via MKP-7 activity
  1. Rheinallt M Jones,
  2. Valerie M Sloane,
  3. Huixia Wu,
  4. Liping Luo,
  5. Amrita Kumar,
  6. Matam Vijay Kumar,
  7. Andrew T Gewirtz,
  8. Andrew S Neish
  1. Epithelial Pathobiology Unit, Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, Georgia, USA
  1. Correspondence to Andrew S Neish, Department of Pathology, Emory University School of Medicine, Room 105F, Whitehead Bldg, 615 Michael Street, Atlanta, GA 30322, USA; aneish{at}emory.edu

Abstract

Background and aims Radiotherapy for neoplastic disease is associated with significant adverse enteric effects associated with excessive cell death. Ionising radiation induces cell death by a mechanism that is dependent on JNK (c-jun N-terminal kinase) pathway signalling. Additionally, it is known that cells exposed to extracellular bacterial products such as flagellin, pleiotropically activate a number of innate immune pathways, including that of JNK. The JNK pathway controls its own activity by inducing the transcription of mitogen-activated protein kinase phosphatase-7 (MKP-7) which directly targets phosphorylated JNK, thus functioning as a negative feedback loop. Previously, it has been shown that flagellin limits ionising radiation-induced mortality in mice, but the cellular mechanism of protection remained unknown.

Methods Wild-type C57BL/6 or tlr5−/− C57BL/6 were injected with flagellin 2 h before exposure to irradiation, and their intestines were examined for apoptosis. Candidate proteins mediating cytoprotection from irradiation were identified by expression profiling. One of these candidates, MKP-7, was cloned and packaged into adenovirus particles, used to infect cultured cells, and examined for the extent to which its activity reduced cellular apoptosis by flow cytometry or immunoblot analysis.

Results Flagellin pretreatment protected mice from radiation-induced intestinal mucosal injury and apoptosis via a Toll-like receptor 5 (TLR5)-dependent mechanism. Expression profiling of flagellin-treated mice showed upregulation of MKP-7, an inducible repressor of the JNK pathway. MKP-7 expression reached a maximum at 2 h after flagellin treatment, coinciding with suppression of phosphorylated JNK and JNK pathway inhibition. Furthermore, constitutive MKP-7 expression protected cultured cells from radiation-induced apoptosis.

Conclusions Flagellin is a promising adjuvant for suppressing ionising radiation-induced injury. MKP-7 activity exhibits cytoprotective effects, and is thus a candidate cellular molecule for limiting the damaging effect of radiotherapy on the gastreointestinal system.

  • Radiation
  • flagellin
  • apoptosis
  • JNK
  • mucosal injury
  • radiation therapy
View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known about this subject?

  • Cancer radiotherapy results in harmful side effects secondary to gut apoptosis.

  • Flagellin preadministration to mice has been reported to have protective effects against irradiation-induced mortality.

  • Irradiation-induced apoptosis in known to be JNK pathway dependent.

What are the new findings?

  • Here, we found that flagellin pre-treatment protects the intestinal mucosa from injury resulting from irradiation-induced apoptosis.

  • Flagellin treatment induced transcription of mkp-7 in the intestinal mucosa to levels that reached a maximum at 2 h after flagellin treatment.

  • Flagellin is most effective at protecting against irradiation-induced apoptosis when administered 2 h pre-exposure, coinciding with the period when levels of MKP-7 are highest.

  • Constitutive expression of MKP-7 is a potent inhibitor of JNK pathway signalling and irradiation-induced apoptosis.

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

  • Our results suggest the specialised use of flagellin as potential adjuvant therapy for limiting irradiation-induced intestinal injury.

  • MKP-7 is a candidate molecule for the modulation of irradiation-induced apoptosis.

Introduction

Ionising radiation is used as adjuvant or curative treatment in cancer patients to control malignant cell growth. Despite clinical efficacy, irradiation treatment also results in harmful side effects in radiosensitive tissues and cell types with high rates of cell division, including, but not limited to, lymphoid organs, bone marrow, testis, ovaries and the intestine. In the intestine, radiation is known to induce programmed cell death in the mucosa, particularly in proliferating cells at the base of the villous crypts. Within days of exposure, damage to the epithelial mucosa culminates in gastrointestinal syndrome, in which extensive infection and electrolyte imbalance may cause significant morbidity and ultimately result in death.1–3 One aim of modern radiotherapy is to reduce the harmful side effects to a minimum. Recent reports have shown that bacterial products and/or proteins as well as cytokines decrease the sensitivity of cells to irradiation-induced apoptosis.4–6 Previous reports have described the phenomenon that pretreatment of mice with the Toll-like receptor 5 (TLR5) agonist flagellin before insult by sublethal doses of ionising radiation results in improved murine survival rates.7–9 However, the molecular mechanism of flagellin (or any bacterially derived compound used as a therapeutic) - mediated protection against ionising radiation injury is unknown, and the pathology of radiosensitive tissues in experimental subjects remain unexamined.

Exposure of cells to ionising radiation initiates the activation of cellular signalling pathways that result in apoptosis. The cellular mechanisms of ionising radiation-induced apoptosis have recently been characterised and show an essential role for c-jun N-terminal kinase (JNK) in activating the intrinsic/mitochondrial apoptotic pathway.10–13 Ionising radiation-induced JNK phosphorylation activates the proapoptotic members of the Bcl-2 family of intracellular proteins, including Bax and Bak, which cause mitochondrial dysfunction resulting in the release of proapoptotic factors (eg, cytochrome c) that activate the initiator caspase-9 and terminate in apoptosis.11 13 14 Upregulation of the JNK signalling pathway can be also initiated by a variety of extracellular signals including bacterial products and proteins such as flagellin.15 Similar to other mitogen-activated protein (MAP) kinases, JNK is activated by a phosphorylation cascade module comprised of a MAP kinase kinase and a MAP kinase kinase kinase. Once activated, JNK phosphorylates a number of transcription factors which results in the increased expression of a battery of innate immune- and apoptosis-related genes.14 JNK pathway activity is negatively regulated by dual specificity phosphatases (DUSPs) which are transcriptionaly activated in response to JNK activation and serve as a negative feedback loop to return JNK signalling to basal levels.16 One member of this family, MAP kinase phosphatase-7 (MKP-7), also known as DUSP16, has activity and high specificity against phosphorylated JNK.17–20

In this study, we show that systemic (intraperitoneal) administration of purified flagellin 2 h prior to radiation insult significantly reduced the number of apoptotic cells in the gut mucosa. Importantly, we also show that flagellin pretreatment mediates protection against ionising radiation via the transient upregulation of MKP-7 in the cytoplasm, which dephosphorylates JNK, thus inhibiting additional JNK pathway activation that would ensue following ionising radiation insult. Indeed, transfection of cells with small interfering RNA (siRNA) against MKP-7 resulted in elevated levels of phosphorylated JNK and increased apoptosis even in the absence of an external stimulus. Furthermore, constitutive cellular expression of MKP-7 within cells significantly decreased irradiation-induced apoptosis. Together, these data show that the mechanism of flagellin-elicited cytoprotection against irradiation-induced injury to the gut mucosa is by MKP-7 inhibition of JNK pathway activity. This investigation highlights the cytoprotective activity of MKP-7 and also suggests its potential suitability as a therapeutic molecule.

Materials and methods

Murine subjects and γ-irradiation

All experiments were done using 12-week-old C57BL/6 background mice that had been purchased from The Jackson Laboratory (Bar Harbor, Maine, USA), except for TLR5−/− mice, which were generated/maintained as previously described.21 Mice whole bodies were exposed to 8 Gy of γ-radiation using a γ-Cell 40 137Cs irradiator at a dose rate of 75 rads/min. Flagellin (50 μg) (or the indicated concentration) was administered intraperitoneally 2 h (or the indicated time) preceding (or following) irradiation, and body weights and mortality were monitored. Animal experiments were approved by the Emory University institutional ethical committee and performed according to the legal requirements. Cultured mammalian cells were exposed to 12 Gy of γ-radiation using a γ-Cell 40 137Cs irradiator at a dose rate of 75 rads/min. Histological sections of small intestine were prepared from three irradiated animals per treatment. Sections were assessed by TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay (Roche, Indianapolis, Indiana, USA) and by immunohistochemistry using antibodies against cleaved caspase-3. TUNEL- or active caspase-3-positive cells were counted and the average number of positive cells in forty 200× fields per treated animal was determined.

Microarray analysis

Details of cDNA microarray fabrication, hybridisation, scanning and labelling of RNA samples have been described previously.15 Total RNA from murine subjects was prepared using TRIzol reagent (Invitrogen, Carlsbad, California, USA). With the use of an oligo(dT) primer and Superscript II Reverse Transcriptase (Invitrogen), labelled cDNA was synthesised from 40 μg of total RNA. Reference and experimental RNA samples were labelled with Cy3- and Cy5-coupled dCTP (Amersham Biosciences; Piscataway, New Jersey, USA), respectively, and hybridised to microarrays. Three repeats for each experimental condition were performed. Following normalisation, genes with intensity >4 times the background mean were selected. To minimise variability, the mean of three independent experiments for each gene was calculated and used for final data clustering. Only genes that showed significant change (over twofold difference) were selected for report. Cluster/Tree view (Michael Eisen, Stanford University, Stanford, California, USA) analytic software packages were used for hierarchical clustering.

Results

Flagellin pretreatment reduces irradiation-induced apoptosis in murine intestinal mucosa by a TLR5-dependent mechanism

Radiation insult is known to induce premature apoptosis in the mammalian intestinal mucosa, particularly in proliferating cells in villous crypts. Previous reports by our research group and others showed that intraperitoneal administration of flagellin in mice mediates protection from radiation-induced mortality.9 In order to examine the extent to which intraperitoneal flagellin administration prior to radiation insult has protective activity against injury to the intestinal mucosa, we injected flagellin purified from Salmonella typhimurium SL3201 into wild-type C57BL/6 mice and TLR5−/− C57BL/6 mice at 2 h before whole-body insult with 8 Gy of irradiation. Animals were sacrificed at 6 h postirradiation and transverse sections of ileal tissue examined for crypt apoptosis by morphological analysis. Irradiated C57BL/6 mice pretreated with phosphate-buffered saline (PBS) exhibited increased numbers of pycnotic cells (condensed nuclei) and karyorrhexis (fragmented nuclei) within the crypt base compared with untreated mice (figure 1A). Strikingly, C57BL/6 mice pretreated with flagellin before irradiation had significantly reduced numbers of pycnotic or karyorrhexis cells compared with PBS-treated irradiated mice (figure 1A,B), indicating that flagellin administration reduced the number of apoptotic cells in wild-type ilea. However, irradiated TLR5−/− C57BL/6 mice treated with flagellin showed no decrease in the number of pycnotic or karyorrhexis cells compared with the irradiated mice treated with PBS. In fact, irradiated TLR5−/− C57BL/6 mice exhibited elevated crypt apoptosis when pretreated with flagellin, compared with mice pretreated with PBS, indicating that flagellin potentially has a radiosensitising effect on mice devoid of TLR5-mediated signalling (figure 1A,B). We also examined radiation-induced apoptosis in ilea of C57BL/6 or TLR5−/− C57BL/6 mice using TUNEL assay, which directly detects nuclear DNA fragmentation in late stage apoptotic cells. Consistent with data described above, C57BL/6 wild-type mice pretreated with flagellin before irradiation had markedly decreased numbers of TUNEL-positive cells in ileal crypts, compared with control mice pretreated with PBS. However, flagellin pretreatment before irradiation did not reduce the number of TUNEL-positive cells in the ileal cryps of TLR5−/− C57BL/6 mice (figure 1C,D). Finally, ileal crypts of irradiated mice were examined by immunohistochemistry using an antibody against cleaved caspase-3. Again, wild-type C57BL/6 mice pretreated with flagellin 2 h before irradiation had decreased numbers of active caspase-3-positive cells compared with isogenic controls treated with PBS, whereas TLR5−/− C57BL/6 mice showed no evidence of decreased numbers of apoptotic cells following flagellin pretreatment (figure 1E,F). Together, these data indicate that pretreatment of mice with flagellin reduces radiation-induced apoptosis in the murine intestinal crypt cells compartment by a TLR5-dependent mechanism.

Figure 1

Flagellin pretreatment reduces ionising radiation-induced apoptosis in the intestinal mucosa by a Toll-like receptor 5 (TLR5)-mediated mechanism. (A) H&E stain of caeca from C57BL/6 and C57BL/6 (tlr5−/−) mice administered 50 μg flagellin (intraperitoneally) 2 h before insult by 8 Gy of γ-radiation. Arrows indicate pycnotic cells. (B) Quantitative representation of pycnotic cells per crypt in genotypes and treatments described in (A). Student t test (n=40) (*p<0.05) (C) TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay analysis of caecal sections from C57BL/6 and C57BL/6 (tlr5−/−) mice treated as described in (A). Arrows indicate TUNEL-positive cells. (D) Quantitative representation of TUNEL-positive cells in thge genotypes and treatments described in (C). Student t test (n=40) (*p<0.05). (E) Immunostain analysis of caecal sections from C57BL/6 and C57BL/6 (tlr5−/−) mice treated as described in (A) with an antibody against cleaved caspase-3. Arrows indicate cleaved caspase-3-positive cells. (F) Quantitative representation of caspase-3-positive cells in the genotypes and treatments described in (E). Student t test (n=40) (*p<0.05).

Flagellin pretreatment or constitutive TLR5 activation reduces ionising radiation-induced apoptosis in cultured intestinal epithelial cells (IEC-6)

In order to determine the extent to which flagellin mediates cytoprotective effects on epithelial cells, we subjected cultured epithelial cells to radiation insult following flagellin pretreatment. Radiosensitive cultured IEC-6 were treated with flagellin or PBS for 2 h before insult with 12 Gy of radiation—the minimum radiation dose required to induce apoptosis in cultured IEC-6. Apoptotic cells entering late-phase programmed cell death were detected by TUNEL assay, and counted per 20× microscopy field. Consistent with our previous observations in mouse epithelial mucosa, cultured cells pretreated with flagellin had visibly fewer TUNEL-positive cells following irradiation compared with cells pretreated with PBS (figure 2A,B). We also stained these cell populations with annexin V and propodium iodide before analysis by flow cytometry. Cell populations pretreated with flagellin exhibited a marked decrease in annexin V-positive cells compared with irradiated cells or unirradiated cells treated with flagellin alone (figure 2C,D). These results also indicate that the cytoprotective effects of flagellin can be modelled in cultured IEC-6, and that the observed cytoprotective effect is epithelial, and not due to a paracrine effects.

Figure 2

Flagellin pretreatment or constitutive Toll-like receptor 5 (TLR5) activation reduces ionizing radiation-induced apoptosis in cultured intestinal epithelial cells. (A) TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay analysis of IEC-6 cultured cells pretreated with flagellin for 2 h before insult by 12 Gy of γ-radiation. (B) TUNEL-positive cells per 20× microscopy field in the assay described in (A). Student t test (n=40) (*p<0.05). (C) Flow cytometry analysis for annexin V-positive cells in IEC-6 cultured cells pretreated with flagellin for 2 h before insult by 12 Gy of γ-radiation. (D) Quantitative representation of annexin V-positive cells described in (C). Green fluorescent protein (GFP)-positive cells were gated and the percentage of annexin V/GFP-positive cells was quantified. Student t test (n=40) (*p<0.05).

Flagellin induces upregulation of cytoprotective genes in murine small intestine

Following our observations of flagellin-mediated protection from radiation-induced injury in murine radiosensitive intestinal tissues and cultured IEC-6, we undertook an expression profiling study of genes upregulated in the small intestine in response to systemic flagellin. C57BL/6 mice were treated intraperitoneally with flagellin or PBS for 2 h before preparation of total RNA from murine ilea. The abundance of mRNA species in the intestinal tissue was measured by cDNA microarray according to microarray system and data management methodology reported previously.15 Gene expression levels from experimental flagellin-treated animals were normalised against the levels for untreated mice. Flagellin treatment resulted in the upregulation of genes which have been shown to function as proinflammatory mediators, microbicidals, antiapoptotic agents, antioxidants and stress-response genes (table 1). Particularly notable is the robust (16-fold) upregulation of dual-specificity phosphatase 16 (DUSP16) (figure 3A). Also known as MKP-7, this protein functions to dephosphorylate phospho-JNK, thus acting as a negative feedback loop to restore JNK pathway signalling to tonic levels. Because radiation-induced apoptosis is mediated via JNK pathway activation,10–12 we hypothesised that flagellin-mediated induction of cellular MKP-7 levels (which leads to the dephosphorylation of cytoplasmic JNK) suppresses radiation-induced apoptosis. In order to recapitulate these data in a cultured cell model, we treated IEC-6 with flagellin and then measured MKP-7 transcript levels at intervals up to 6 h using quantitative PCR. In cultured cells, MKP-7 transcript levels increased up to 3 h after flagellin stimulation and then decreased to levels similar to those of unstimulated cells at 4 h after flagellin stimulation (figure 3B). In a parallel experiment, we analysed JNK pathway activity in IEC-6 treated with flagellin at identical time points. Levels of phosphorylated JNK increased to a maximum at 15 min after flagellin stimulation, but were undetectable at 2 h (figure 3C). The reduction of phosphorylated JNK species directly coincides with elevated MKP-7 levels up to 2 h poststimulation. Furthermore, elevated levels of MKP-7 were detected in cultured IEC-6 by immunoblot analysis up to 3 h after flagellin stimulation (figure 3D). To confirm that repression of MKP-7 functions in JNK dephosphorylation, siRNA against MKP-7 was transfected into IEC-6 for 48 h. A marked dose-dependent increase in phosphorylated JNK was detected in these cells, showing that inhibition of MKP-7 translation results in constitutive activation of JNK signalling (figure 3E). Cells transfected with siRNA against MKP-7 also exhibited elevated levels of apoptosis even in the absence of external stimulation; levels that were further increased following irradiation (figure 3F). Together, these data show that MKP-7 is an inhibitor of JNK pathway signalling, and MKP-7 loss of function induces cellular apoptosis.

Table 1

Flagellin induces expression of cytoprotective molecules in the murine intestine: microarray analysis of mRNA isolated from murine intestine 2 h after intraperitoneal administration of 50 μg of flagellin

Figure 3

Flagellin induces the expression of c-Jun N-terminal kinase (JNK)-specific phosphatase mitogen-activated protein kinase phosphatase 7 (MKP-7; DUSP16) in cultured IEC-6. (A) Quantitative PCR analysis of MKP-7 gene expression in murine ileum preparations at 2 h following intraperitoneal administration of 50 μg of flagellin. (B) Quantitative PCR analysis of MKP-7 gene expression in cultured IEC-6 after treatment with flagellin. (C) Immunoblot analysis of cultured IEC-6 treated with flagellin using an antibody against phosphorylated JNK. (D) Immunoblot analysis of lysates from cultured IEC-6 treated with flagellin for the indicated periods, using an antibody against MKP-7. MKP-7 was detected at 73 kDa. A non-specific (N.S.) band at 95 kDa was used as loading control. (E) Immunoblot analysis of lysates of IEC-6 either untreated (U) or transfected with non-targeting (NT) or MKP-7 small interfering RNA (siRNA), using an antibody against phosphorylated JNK. (F) Flow cytometry analysis for annexin V-positive cultured IEC-6 transfected with 10 nM non-targeting (NT) or MKP-7 siRNA. Some transfected cell populations were analysed before irradiation (Un) and others at 24 h following insult by 12 Gy of γ-radiation (Irr.). Error bars indicate the SEM. GADPH, glyceraldehyde phosphate dehydrogenase.

Constitutive MKP7 expression potently inhibits JNK phosphorylation and reduces radiation-induced apoptosis in cultured cells

To determine the extent to which the cytoprotective effects of flagellin pretreatment can be mediated by constitutive MKP-7 expression, we cloned the MKP-7 gene sequence into the mammalian expression vector pCMV-myc. Additionally, we created a control construct harbouring a catalytically inactive form of MKP-7 (mMKP-7), where a critical residue known to be required for MKP-7 catalytic activity was mutated to an alanine residue. To test whether the catalytic activity was preserved in MKP-7, and to test whether the catalytic activity had been abolished in mMKP-7, we co-transfected these constructs, along with plasmids harbouring the JNK pathway intermediates JNK1 and MKK4 to initiate pathway activation, into cultured HEK293 cells. As expected, pCMV-myc-MKP-7 potently abrogate induced levels of phosphorylated JNK1, whereas pCMV-myc-mMKP-7 could not (figure 4A). Bacterial flagellin is a known ligand for several pattern recognition receptors, including IPAF and TLR5. In order to determine the extent to which recombinant MKP-7 could abrogate phosphorylation of JNK in response to TLR5 activation, we investigated the effects of transfecting cultured cells with a plasmid harbouring a constitutively active form of TLR5 transcriptionally fused to the surface receptor cluster of differentiation 4 (CD4). We created the pEF6-V5-CD4TLR5 construct, which includes a CD4 extracellular domain on the N-terminus, and a TLR5 intracellular (including the TIR domain) domain on the C-terminus. Similar to the data described in figure 4A, pCMV-myc-MKP-7 potently abrogated induced levels of phosphorylated JNK1 initiated by pEF6-V5-CD4TLR5, whereas pCMV-myc-mMKP-7 could not (figure 4B). In order to determine the extent to which constitutive MKP-7 expression abrogates apoptosis resulting from irradiation insult, we co-transfected cultured IEC-6 with pcDNA-EGFP (enhanced green fluorescent protein) and pCMV-myc-MKP7 or pCMV-myc-mMKP7, respectively. After 24 h cells were subjected to 12 Gy irradiation and incubated for another 24 h. Cells were then stained with annexin V and propodium iodide and analysed by flow cytometry. Data analysis where GFP-positive populations were gated and the percentage of annexin V-positive cells determined showed a decrease in the number of irradiation-induced apoptosis-positive cells in the population expressing MKP-7, but not in populations expressing mMKP-7 or vector alone (Supplementary figure 1). However, a caveat of using cultured IEC-6 in transfection studies is the relatively low transfection efficiency of this cell line; in the order of 5–10% efficiency. To achieve high efficiency expression of MKP-7 in radiosensitive cells, we subcloned myc-MKP7 and myc-mMKP7 into pShuttle-IERS-hrGFP-1 vector. These two constructs or the pShuttle-IERS-hrGFP-1 vector alone was then recombined with pAdEasy-1 adenoviral vector to make pAdEasy-myc-MKP7-GFP, pAdEasy-myc-mMKP7-GFP and pAdEasy-GFP. The three viral constructs were packaged, amplified and purified to a titre of 107 pfu/ml. Packaged adenovirus were then infected into cultured IEC-6 with an estimated efficiency of 80% as determined by counting GFP-positive cells (Supplementary figure 2). Importantly, and consistent with data presented in figure 3E using siRNA against MKP-7, infection with mutant and catalytically inactive pAdEasy-myc-mMKP7-GFP resulted in elevated levels of phosphorylated JNK at a multiplicity of infection (MOI) of >25, whereas no elevation in phosphorylation of JNK was detected at an MOI of 10 (figure 4C). We then confirmed that the activity of MKP-7 expressed from the adenovirus at an MOI of 10 was sufficient to inhibit flagellin-induced JNK phosphorylation (figure 4D). A population of IEC-6 infected at an MOI of 10 with the aforementioned adenoviruses were subjected to 12 Gy irradiation, incubated for 24 h and stained with annexin V and propodium iodide before analysis by flow cytometry. Consistent with the data described above, adenovirus-mediated expression of MKP-7 reduced the number of annexin V-positive cells within the GFP-positive IEC-6 population, showing that blockade of JNK pathway signalling during irradiation insult inhibits apoptosis and diverts cellular fate (figure 4E,F). Together, these data show that irradiation insult-induced apoptosis can be inhibited by constitutive MKP-7 expression.

Figure 4

Constitutive mitogen-activated protein kinase phosphatase 7 (MKP-7) expression potently inhibits c-Jun N-terminal kinase (JNK) phosphorylation and reduces radiation-induced apoptosis in cultured cells. (A) Immunoblot analysis of lystaes from 293HEK cultured cells co-transfected with plasmids harbouring myc-JNK1 and flag-MKK7 (in combination they constitutively activate the JNK pathway) and with plasmids harbouring myc-MKP7 and the catalytically inactive myc-mMKP7. Blots were probed using antibodies against phosphorylated JNK, against c-Myc for detection of JNK1, MKP-7 and mMKP-7 expression, against Flag for detection of MKK7 expression, and against β-actin as loading control. (B) Immunoblot analysis of lysates from 293HEK cultured cells co-transfected with plasmids harbouring V5-CD4-TLR5 and with plasmids harbouring myc-MKP7 and myc-mMKP7. Blots were probed with the same antibodies used in (A) except for anti-V5 which was used to detect expression of CD4-TLR5. (C) Immunoblot analysis of cultured IEC-6 infected with adenovirus harbouring pAdEasy-GFP at a multiplicity of infection (MOI) of 100, of pAdEasy-myc-tMKP-7-GFP at an MOI of 100 or pAdEasy-myc-mMKP-7-GFP at an MOI of 10, 25, 50 or 100, using an antibody against phosphorylated JNK. (D) Immunoblot analysis of cultured IEC-6 transduced with adenovirus harbouring pAdEasy-GFP, pAdEasy-myc-tMKP-7-GFP or pAdEasy-myc-mMKP-7-GFP at an MOI of 10 and treated with flagellin for 15 min. (E) Flow cytometry analysis of cultured IEC-6 infected by adenovirus harbouring pAdEasy-GFP, pAdEasy-myc-tMKP-7-GFP or pAdEasy-myc-mMKP-7-GFP at an MOI of 10, before insult by 12 Gy of γ-radiation. (F) Quantitative representation of annexin V-positive cells detected in (E). Student t test (n=9) (*p<0.05). GFP, green fluorescent protein.

Flagellin pretreatment inhibits ionising radiation-induced JNK phosphorylation and activation of apoptotic protein markers

We then investigated the extent to which flagellin pretreatment could inhibit radiation-induced JNK phosphorylation in cultured IEC-6. Cells were treated with flagellin or PBS for 2 h, then washed with Hanks medium before insult with 12 Gy of radiation. Thereafter, lysates of stimulated or control cells were collected at regular time intervals up to 6 h after flagellin treatment and analysed by immunoblot. As expected, flagellin induced JNK phosphorylation up to 1 h poststimulation, whereupon levels of phosphorylated JNK rapidly became undetectable at 2 h after flagellin stimulation, presumably due to the phosphatase activity of MKP-7 (figure 5A). At 3 and 4 h after flagellin treatment (1 and 2 h postradiation insult), we detected the presence of phosphorylated JNK1 (46 kDa) and, to a much lesser extent, phosphorylated JNK2 and JNK3 (both 54 kDa), only in PBS-treated cells but not in the flagellin-treated cells (figure 5A). Thereafter, the elevation of the apoptotic marker cleaved PARP (poly(ADP-ribose) polymerase) was detected in PBS-treated irradiated cells up to 6 h (4 h postradiation insult) (figure 5A). Importantly, detectable levels of cleaved PARP were reduced in cells pretreated with flagellin before radiation insult, although the emergence of a faint cleaved PARP band at 6 h in the flagellin-treated irradiated cells was probably due to incomplete protection of the whole cell population within the cultured cell model system. To confirm that MKP-7 activity reduces radiation-induced apoptosis, cultured IEC-6 were infected with adenovirus particles harbouring pAdEasy-myc-MKP-7-GFP, pAdEasy-myc-mMKP-7-GFP or pAdEasy-GFP. Cells were then incubated for 24 h, before being subjected to 12 Gy radiation insult. Thereafter, lysates were prepared from the treated cells and analysed by immunoblot. Consistent with the above data, phosphorylated JNK1 and, to a lesser extent, phosphorylated JNK2 and JNK3 were detected in pAdEasy-myc-mMKP-7-GFP or pAdEasy-GFP cell lysates at 1 and 2 h postradiation insult, along with cleaved PARP at intervals up to 6 h (figure 5B). No phosphorylated JNK1 or cleaved PARP was detected in lysates from cells harbouring pAdEasy-myc-MKP-7-GFP. Together, these data show that expression of MKP-7 inhibits radiation-induced JNK phosphorylation and PARP cleavage.

Figure 5

Flagellin pretreatment inhibits ionising radiation-induced c-Jun N-terminal kinase (JNK) phosphorylation and activation of apoptotic protein markers (A) Immunoblot analysis of cultured IEC-6 lysates pretreated for 2 h with either phosphate-buffered saline (PBS) (left-hand panels) or flagellin (right hand panels) and then subjected to insult by 12 Gy of γ-radiation. Blots were probed using antibodies against phosphorylated JNK, and cleaved PARP (poly(ADP-ribose) polymerase). (B) Immunoblot analysis of cultured IEC-6 infected by adenovirus harbouring pAdEasy-GFP, pAdEasy-myc-tMKP-7-GFP or pAdEasy-myc-mMKP-7-GFP before insult by 12 Gy of γ-radiation and analysed by immunoblot using the same antibodies as described in (A). GFP, green fluorescent protein.

Flagellin pretreatment inhibits the intrinsic/mitochondrial apoptotic pathway in response to ionising radiation insult

Apoptosis resulting from radiation insult activates the intrinsic/mitochondrial apoptotic pathway.10 11 A number of proteins have been shown to be required for intrinsic/mitochondrial apoptotic pathway signalling, including Bcl-2 which is phosphorylated during intrinsic pathway activation, Bax that becomes a part of the mitochondrial membrane upon pathway activation, and cytochrome c, which is released from the mitochondria during apoptosis. To show that flagellin pretreatment before radiation insult inhibits intrinsic pathway-induced apoptosis, we repeated experimental procedures described in figure 5, except that lysates were prepared from IEC-6 up to 2 days following irradiation. Lysates were analysed by immunoblot using antibodies against Bcl-2. In control cells subjected to radiation insult only, levels of Bcl-2 gradually decreased up to 2 days after irradiation (figure 6A). However, in flagellin-pretreated irradiated cells, no alteration in Bcl-2 levels was detected (figure 6A). Then, we investigated cellular Bax and cytochrome c distribution in flagellin-treated and irradiated IEC-6. Experiential lysates were collected, fractionated by centrifugation and analysed by immunoblot using anti-Bax or anti-cytochrome c antibodies. Consistent with the data presented above, Bax was recruited to the mitochondria up to 48 h postirradiation (figure 6B), whereas increased cytochrome c was detected in cystolic fractions of IEC-6 over the same period in cells subjected to radiation insult alone (figure 6C). No recruitment of Bax to the mitochondria or release of cytochrome c into the cytosol was detected when cells were pretreated with flagellin before radiation insult. Thus, these data demonstrate that flagellin pretreatment prior to irradiation inhibits the intrinsic cell death pathway.

Figure 6

Flagellin pretreatment inhibits the intrinsic/mitochondrial apoptotic pathway in response to ionising radiation insult. (A) Immunoblot analysis of cultured IEC-6 lysates pretreated with flagellin for 2 h before insult by 12 Gy of γ-radiation. Blots were probed using antibodies against Bcl-2. (B) Immunoblot analysis of the mitochondrial fraction of lysates of cultured IEC-6 pretreated with flagellin for 2 h and then subjected to insult by 12 Gy of γ-radiation using an antibody against Bax (C) Immunoblot analysis of the cyotplasmic fraction of lysates of cultured IEC-6 pretreated with flagellin for 2 h and then subjected to insult by 12 Gy of γ-radiation using an antibody against cytochrome c (CytC). Hsp60, heat shock protein 60; PBS, phosphate-buffered saline.

Discussion

The intestinal epithelium is the most dynamically renewing tissue in adult mammals. Stem cells exiting the intestinal crypts towards the villus apex differentiate into enterocytes, goblet cells and enteroendocrine cells which, together with intercellular junctional proteins, form a tight biological barrier. Spontaneous or tonic levels of apoptosis occur in small intestine crypts and the luminal surface of healthy mammalian intestine as part of normal gut maintenance or as part of the homeostatic mechanism regulating stem cell numbers. However, elevation of normal gut apoptosis leads to gastrointestinal disease, a condition characteristic of many enteric infections and commonly manifested in patients with cancer during irradiation treatment. Ionising radiation considerably elevates normal levels of apoptosis, particularly in the small intestine crypts at 2–6 h following exposure. Elevated apoptosis directly leads to reduced enterocyte numbers and weakened barrier function, which potentially allow resident flora to invade intestinal organs and tissue. Strategies to limit ionising radiation-induced gut apoptosis have not been substantially explored. Here, we describe the specialised use of flagellin as potential adjuvant therapy for limiting irradiation-induced gut apoptosis and identify MKP-7 as a potent cytoprotective factor transiently upregulated in response to flagellin. Our investigations revealed that MKP-7 expression is markedly elevated in the murine small intestine in response to flagellin at 2 h post-treatment. MKP-7 functions in the selective inhibition of the JNK pathway by potently dephosphorylating p-JNK, with highly conserved functionality.17 19 22 Consistently, constitutive MKP-7 expression in cultured cells potently inhibited the phosphorylation of JNK, thereby also inhibiting cellular outcomes of JNK pathway activation, including ionising radiation-induced apoptosis. Due to an MKP-7-mediated negative feedback loop, p-JNK is undetectable at 2 h after flagellin treatment, a time point that directly correlates with maximal MKP-7 expression. Hence, the optimal time for flagellin administration is at a time window 2 h before ionising radiation insult. Flagellin pretreatment outside of this time window failed to reduce ionising radiation-induced apoptosis in the murine small intestine.

As well as the radioprotective effects from the modulation of JNK signalling reported in this study, other reports have also demonstrated the radioprotective effects of the nuclear factor-κB (NF-κB) pathway.4 23 24 Interestingly, our expression profiling studies of mouse intestinal tissue only detected the twofold induction of the NF-κB-associated Faim antiapoptotic factor, and the most highly expressed NF-κB-induced genes detected were of reported antimicrobial function. Nevertheless, due to increased reports of cross-talk in the regulation of the NF-κB and JNK pathways,25 it is highly plausible that proteins upregulated by either pathways elicit temporal cytoprotective effects. Here, we report on the potential for utilising purified flagellin as an adjuvant cytoprotective therapy within a distinct time frame (2 h) before irradiation.

Recent reports have shown that recognition of commensal microflora by TLRs is required for intestinal homeostasis.26 However, disease prevention approaches where innate immunity is activated by bacterial products or proteins have been given little consideration due to the potential for severe inflammatory pathology. Nevertheless, not all bacterially derived innate immune activators potentiate equivalent inflammatory effects. For example, flagellin has been shown to elicit comparatively mild pathological outcomes compared with lipolysaccharide administration, which results in severe inflammation and shock. The disparate pathological effect of bacterial products within the host is due to diverse cellular responsiveness to flagellin. Whereas epithelial cells robustly respond to flagellin,27–29 macrophages are devoid of TLR5,30 and thus flagellin is a poor inducer of tumour necrosis factor α in mice.9 30 In fact, most populations of murine macrophages or dendritic cells do not respond to purified soluble flagellin,31 32 and the majority of the flagellin-induced cytokine elevation is mediated by MyD88-dependent signalling in non-haemopoietic cells.33 Another noteworthy observation supporting the suitability of flagellin pretreatment as adjuvant therapy is that flagellin has a relatively short half-life in mice. Flagellin is undetectable in the murine serum 2 h following intraperitoneal administration of 50 μg, suggesting efficient immune mechanisms of proteolytic elimination, and also avoiding the possible determinative effect of prolonged signalling. Together, these data strongly support the use of purified soluble flagellin as adjuvant therapy to reduce the harmful side effects of ionising irradiation-induced gut apoptosis. Strategies to activate antiapoptotic factors transiently with flagellin pretreatment, a comparatively less toxic bacterial product, should be considered for development in specialised use in limiting determinative effects of external insult.

References

View Abstract

Supplementary materials

  • Online only appendix

    Files in this Data Supplement:

Footnotes

  • Funding This work was supported in part by National Institutes of Health grants DK-71604 and AI-64462 (ASN), and DK-64399.

  • Competing interests None.

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

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.