 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
Department of Infectious Diseases, Imperial College School of Medicine, Hammersmith Hospital, London, United Kingdom.
Correspondence to Dr. Thomas J. Evans, Department of Infectious Diseases, Imperial College School of Medicine, Hammersmith Hospital, Du Cane Road, London W12 0NN, United Kingdom. Phone: +44-20-8383-8576; Fax: +44-20-8383-3394; E-mail: tom.evans{at}ic.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, interleukin-1, and interferon-
, the actin
cytoskeleton was disrupted and cells became elongated, with extension of long
filopodial processes. Cytokines induced shedding of viable, apoptotic, and
necrotic PTEC, which was dependent on NO synthesized by inducible NO synthase
(iNOS) produced as a result of cytokine actions on PTEC. Basolateral exposure
of polarized PTEC monolayers to cytokines induced maximal NO-dependent cell
shedding, mediated in part through NO effects on cGMP. Cell shedding was
accompanied by dispersal of basolateral ß1 integrins and
E-cadherin, with corresponding upregulation of integrin expression in clusters
of cells elevated above the epithelial monolayer. These cells demonstrated
coexpression of iNOS and apically redistributed ß1 integrins.
Attachment studies demonstrated that the major ligand involved in cell
anchorage was laminin, probably through interactions with the integrin
3ß1. This interaction was downregulated by
cytokines but was not dependent on NO. These studies provide a mechanism by
which inflammatory cytokines induce PTEC damage in sepsis, in the absence of
hypotension and ischemia. Future therapeutic strategies aimed at specific iNOS
inhibition might inhibit PTEC shedding after cytokine-induced injury and delay
the onset of acute renal failure in sepsis. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3ß1 integrins to the apical cell compartment
(12). Dissolution of
ß1 integrincell matrix adhesion results in tubule cell
detachment from the basement membrane into the lumen
(13) and may lead to tubule
cell apoptosis (14). Shed
tubule cells aggregate into casts, which is possibly promoted by
RGD-containing ligands, such as urinary fibronectin (Fn), that bridge
integrins expressed on detached and in situ cells
(13,15).
Tubule obstruction by shed cells and casts is a major factor leading to GFR
reduction in ARF (16). This is
supported by studies demonstrating the presence of tubular casts in tissue
sections and the recovery of viable and nonviable tubule cells from the urine
of patients with ARF
(5,17).
However, renal function can deteriorate without a reduction in renal blood
flow
(18,19),
demonstrating that nonhemodynamic factors are also important. A large body of
data has demonstrated that the inflammatory cytokines tumor necrosis
factor- (TNF-), interleukin-1 (IL-1), and interferon-
(IFN-) are of central importance in septic shock
(2,20).
High levels of these cytokines are produced during sepsis and have been
demonstrated to mediate many of the pathophysiologic changes that occur among
patients with sepsis (20).
Purified cytokines, such as TNF-, can produce acute tubular necrosis
when administered to experimental animals
(21). However, because
TNF- can also induce hypotension, it is impossible to distinguish
direct cytokine effects on tubule cells from indirect effects of hypotension.
Indeed, little information exists regarding the specific effects of
inflammatory cytokines on PTEC.
One of the known effects of inflammatory cytokines on PTEC is the
generation of nitric oxide (NO) through induction of the enzyme inducible NO
synthase (iNOS)
(22,23).
Maximal production of iNOS requires stimulation of cells with a combination of
inflammatory cytokines, such as TNF-, IL-1, and IFN-
(24). The exact role of NO in
these cells is not clear, but NO may produce cytotoxic effects
(25). In vitro and
in vivo studies demonstrated that NO produced via iNOS contributes to
ischemic tubular injury
(26,27,28,29),
but there are no data on the role of endogenous tubule cell NO production
after cytokine stimulation.
The aim of this study was to determine the effects of the combination of inflammatory cytokines observed in sepsis on human PTEC in the absence of complicating hemodynamic factors. In particular, we wished to determine whether these cytokine effects could induce the cytoskeletal changes and cell shedding that have been noted in ischemic damage and to assess the contribution of NO to these processes.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, human TNF-, and human IFN-
(all from Peprotech EC).
Phase-Contrast Microscopy
PTEC grown on "slideflasks" (Life Technologies) were observed
by using a Nikon Diaphot 300 inverted phase-contrast microscope (Nikon,
Surrey, UK).
Transepithelial Resistance Measurements
Transepithelial resistance was measured across confluent PTEC monolayers
grown on 24-mm, Transwell-COL, collagen-coated, cell culture inserts (no.
3491; Corning Costar, High Wycombe, Bucking-hamshire, UK), using an EVOM
epithelial voltohmmeter (World Precision Instruments, Sarasota, FL).
Analysis of NO Production
NO production was measured as either nitrite or nitrate plus nitrite
(NOx) production, using the Griess assay
(31), after reduction of
nitrate to nitrite using a Nitralyzer (World Precision Instruments). The
effects of NO production were determined by using the selective iNOS inhibitor
L-N6-(1-iminoethyl)lysine (L-NIL) (CN Biosciences,
Nottingham, UK) or the specific NO synthase (NOS) inhibitor
L-NG-monomethyl arginine (L-NMMA) (Alexis, Nottingham,
UK). The effects of L-NMMA were compared with those of the inactive
stereoisomer D-NMMA (Alexis).
Immunofluorescence Assays
The following reagents were used: anti-ß1 integrin
monoclonal antibody (mAb) (Autogen Bioclear UK, Wiltshire, UK);
anti-3 integrin mAb (P1B5; Chemicon International, Harrow,
UK); anti-E-cadherin mAb (C20820; Transduction Laboratories, Lexington, KY);
anti-endothelial NOS mAb (N30030; Transduction Laboratories); anti-neuronal
NOS mAb (N31020; Transduction Laboratories); tetrarhodamine
isothiocyanate-conjugated phalloidin (Sigma Chemical Co., Poole, UK);
anti-pan-cytokeratin mAb (clone C-11; Sigma); Cy3-conjugated, AffiniPure,
donkey anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA);
FITC-conjugated antimouse IgG (Sigma); anti-iNOS mAb raised to residues 961 to
1144 of human iNOS (N39120; Transduction Laboratories), used at a
concentration of 5 µg/ml; and rabbit anti-iNOS polyclonal antibody raised
to residues 54 to 76 of human iNOS and affinity-purified with the immunizing
peptide (32), used at 0.5
µg/ml. Relevant isotype-specific control antibodies and preimmune sera were
obtained from Serotec (Oxford, UK).
Incubations were performed at room temperature. Cells were fixed for 1 h in 1% paraformaldehyde in phosphate-buffered saline (PBS) and were then incubated for 1 h with 10% normal horse (for primary antibodies raised in mice) or goat (for primary antibodies raised in rabbits) serum in PBS. Cells were sequentially treated with primary antibody (1 h) and biotinylated secondary antibody (1:200 dilution, 1 h) (Vector Laboratories, Peterborough, UK). iNOS staining was observed by using fluorescein-conjugated avidin DCS (Vector Laboratories), and cells were mounted in Vectashield (Vector Laboratories). In experiments using other primary mAb, staining was observed by using a secondary, fluorochrome-conjugated, anti-mouse Ig. Conventional fluorescence microscopy was performed by using a Nikon Eclipse E600 microscope. Confocal microscopy was performed with the BioRad MRC 1024 System (Hercules, CA), using a Zeiss Axiovert microscope (Zeiss, Welwyn Garden City, UK) and LaserSharp software.
Immunofluorescence staining of cadherin and integrins was quantified as the mean peak pixel intensity per cell, using NIH Image 1.62 software. iNOS and apical integrin expression was calculated as a percentage of the area of the epithelial cell sheet in a given micrograph, using NIH Image 1.62 software to measure the area of stained cells with pixel intensities above the threshold level.
Western Blotting of iNOS Protein
Cell lysates were prepared by scraping confluent cells directly into
boiling Laemmli buffer, and proteins were separated by 8% sodium dodecyl
sulfate-polyacrylamide gel electrophoresis. Proteins were transferred to
polyvinylidene difluoride membranes (Amersham Pharmacia Biotech,
Buckinghamshire, UK), and iNOS was detected by using an anti-iNOS mAb and the
ECL Plus system (Amersham Pharmacia Biotech).
iNOS Activity Assays
iNOS activity was assayed as arginine/citrulline conversion measured in
duplicate for each culture sample, using a NOS activity assay kit (NOSdetect
assay kit; Stratagene, Cambridge, UK).
Cell Detachment Assays
Confluent PTEC monolayers were cultured in fetal calf
serum/collagen-coated, 5-cm, tissue culture Petri dishes. PTEC detachment was
determined by removing supernatants at 24 h and counting cell numbers with a
hemocytometer. In experiments using polarized PTEC cultured on Transwell-COL
inserts, serum-free medium containing cytokines was added to either the apical
(total volume, 0.5 ml) or basolateral (total volume, 1 ml) compartment. At 24
h, shed cells were counted after removal of apical supernatants,
centrifugation at 500 x g for 5 min, and resuspension of the
cell pellet in 50 µl of Hanks' balanced salt solution. Data are presented
as the number of cells per milliliter, taken as the mean of eight cell counts
from two 20-µl aliquots of cell suspension. In some experiments, cells were
treated with the exogenous NO donor dipropylenetriamine nonoate (DPTA)
(Alexis), in the presence or absence of the cGMP inhibitor
1H-[1,2,4]oxadiazole[4,3-]quinoxalin-1-one (ODQ) (Alexis).
Cell Adhesion Assays
Assays of cell adhesion to matrix proteins were adapted from a previously
described method (33). Human
collagen type IV, human Fn, and human laminin (all from Becton Dickinson,
Oxford, UK) were coated on 96-well cell culture plates. Fifty microliters of
matrix proteins, either used at stock concentrations (1 mg/ml Fn, 1.25 mg/ml
laminin, and 100 µg/ml collagen type IV) or diluted in PBS to a final
concentration of 20 µg/ml, were added to each well. After coating for 16 h
at 4°C, the protein solution was aspirated and wells were washed twice
with calcium-free PBS. Nonspecific adhesion was blocked by incubation for 2 h
with 2% bovine serum albumin in PBS, and the wells were washed twice with PBS
before use. Confluent PTEC monolayers were trypsinized with 0.1% trypsin/0.04%
ethylenedia-minetetraacetate in Hanks' balanced salt solution and were allowed
to recover for 30 min in serum-free DMEM/F-12 medium. Aliquots (100 µl) of
the PTEC suspension (105 cells/ml in DMEM/F-12 medium with 0.1%
bovine serum albumin) were then added to the wells, and the cells were allowed
to adhere for 90 min at 37°C. Supernatants were then removed, the wells
were washed three times with PBS, and cells were fixed with 1%
paraformaldehyde for 1 h. Attached cells were stained with 0.25% aqueous
crystal violet for 10 min, washed with tap water, and allowed to dry at
37°C. One hundred microliters of 33% glacial acetic acid were added, and
the OD at 570 nm was recorded.
Monolayer Cell Viability Studies
Necrosis was assessed by using a live/dead reduced-biohazard
viability/cytotoxicity kit (Molecular Probes, Eugene, OR). Apoptosis was
determined by assessment of the nuclear morphologic features of cells stained
with hematoxylin and eosin and by terminal deoxynucleotidyl
transferase-mediated dUTP nick-end labeling using a commercially available kit
(Promega Corp., Southampton, UK).
Shed Cell Viability Studies
Necrosis and apoptosis were examined by assessment of the nuclear
morphologic features of cells stained with acridine orange
(34) and by flow cytometric
analyses using propidium iodide and annexin V labeling.
Proliferation Assays
Confluent PTEC were pulsed with [3H]thymidine (2 µCi/ml;
Amersham Pharmacia Biotech) at 10 h after the start of each experiment.
Supernatants were removed, and the monolayers were washed three times with
PBS. Cellular protein was precipitated by the addition of 5% TCA, washed three
times, and dissolved in 0.2 M sodium hydroxide. Cells in the supernatant were
centrifuged, and protein was precipitated with TCA as described above. DNA
concentrations were measured by recording the OD at 260 nm, and incorporated
radioactivity was determined by scintillation counting.
Data Analyses
Data are reported as the mean ± SEM. The level of significance,
P, for differences from the control experiments was calculated by
using the t test (two-tailed). A value of <0.05 was taken to
indicate statistical significance.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
·cm2, n = 5).
Cytokeratin intermediate filaments were present within the cells in an ordered
pattern (Figure 1C).
Fluorescence staining of these cells for polymerized actin demonstrated basal
actin stress fibers (Figure
1E). E-cadherin was peripherally distributed at the cell-cell
junctions of unstimulated PTEC (Figure
1G).
|
To reproduce the mixture of inflammatory cytokines observed in sepsis, we
stimulated cells with 10 ng/ml IL-1, 200 U/ml IFN-, and 10 ng/ml
TNF-. These concentrations reflect circulating cytokine levels observed
in sepsis in vivo
(35), and this combination of
cytokines consistently produces maximal iNOS production in PTEC cultures
(Table 1)
(23,36).
Within 2 h after cytokine treatment, the morphologic features of the cells
changed dramatically (Figure
1). The cells lost their cobblestone appearance and aggregated
into heaps scattered over the monolayer
(Figure 1B, arrowheads). Cells
became elongated and extended multiple long filopodial processes, some of
which were longer than 200 µm (Figure 1,
B, D, and F). These cells maintained their epithelial phenotype,
as evidenced by staining for cytokeratin
(Figure 1D). Immunofluorescence
staining demonstrated marked disruption of basal actin stress fibers
(Figure 1F), with a significant
reduction in the mean number of actin bundles (from 9.6 ± 1.36 bundles
cell for control cells to 1.0 ± 0.548 bundles/cell for cytokine-treated
cells, n = 5, P < 0.002). Cadherin staining was markedly
reduced after cytokine treatment [mean peak fluorescence intensity, 134.8
± 20.8 (control) versus 50.76 ± 15.9
(cytokine-treated); n = 5; P < 0.02]
(Figure 1, G and H). However,
actin and cytokeratin filaments extended into the long cellular processes
(Figure 1, D and F).
|
Inflammatory Cytokines Stimulate NO Production by iNOS Expressed in
PTEC Displaying Altered Morphologic Features
Stimulation with the combination of the cytokines IL-1 (10 ng/ml),
IFN- (200 U/ml), and TNF- (10 ng/ml) for 24 h produced maximal
NOx release, compared with unstimulated cells
(Table 1). After cytokine
treatment in the presence of 1 mM L-NMMA or 1 mM L-NIL, there were
statistically significant reductions in NO output by PTEC, compared with
cytokine treatment alone (88.5 ± 0.55 and 91.63 ± 0.39%
reductions, respectively; n = 4; P < 0.0001). Addition of
1 mM D-NMMA to cytokine-treated cells did not alter NO production. These
results demonstrated that NO production by cytokine-stimulated human PTEC was
NOS-dependent.
We also assessed the expression of iNOS in PTEC by using immunofluorescence and Western blotting assays. Unstimulated cells did not express iNOS (Figure 2A). After cytokine stimulation, there was marked induction of iNOS (Figure 2, B and C). The typical pattern of staining indicated that iNOS was expressed in only a proportion of the epithelial cells, with a median value of 7.75% iNOS-positive cells/high-power field (range, 2.59 to 16.9% of cells/high-power field in eight separate experiments). These cells were epithelial in nature (positive staining for cytokeratin and negative staining for vimentin; data not shown). iNOS-positive cells often appeared in aggregates (Figure 2C, arrowheads), as can be better observed in the lower-power view in Figure 3B. The specificity of staining with the anti-peptide serum was confirmed by absorption with the specific iNOS peptide, and isotype-specific control antibodies yielded negative results (data not shown). There was no change in the distribution of iNOS staining after NOS inhibition. Western blotting analysis of cytokine-treated PTEC homogenates with anti-iNOS revealed a 135-kD protein band, corresponding to human iNOS (Figure 2D); there was no iNOS band in unstimulated cell lysates. Endothelial NOS and neuronal NOS were not detectable in these cells by Western blotting (data not shown). Measurements of PTEC NOS activity demonstrated that the conversion of arginine to citrulline was significantly increased after cytokine treatment, compared with control values (73.19 ± 11.1 versus 0.025 ± 0.002 pmol/mg per min, respectively; n = 2; P < 0.05), with no significant change after the removal of calcium from the reaction mixture (66.08 ± 0.09 pmol/mg per min). Taken together, these data demonstrated that NO release was attributable to iNOS.
|
|
Inflammatory Cytokines Induce Detachment of Necrotic and Apoptotic
Cells from PTEC Monolayers, through Production of NO
To determine the extent of cellular injury induced by cytokine treatment,
we examined monolayer viability. Cytokine-stimulated monolayers maintained
their viability (92.4 ± 7.6% viable, n = 4) with no
significant cell necrosis, compared with unstimulated control cells (99
± 0.5% viable). Morphologic examination and terminal deoxynucleotidyl
transferase-mediated dUTP nick-end labeling assays demonstrated no evidence of
apoptosis in control or cytokine-stimulated cells. There was no loss of
viability after NO inhibition (96 ± 2.3% viable) and no appearance of
apoptotic cells.
We observed that the numbers of cells in the supernatants of cytokine-treated PTEC monolayers were much greater than those in control samples (Figure 4A). Treatment with 1 mM L-NMMA reduced the numbers of shed cells back to values similar to those recorded for control samples, whereas D-NMMA had no significant effect (Figure 4A). Therefore, cytokine-induced shedding of these cells is NO-dependent.
|
We examined the shed PTEC population, to assess cell viability and to determine whether NO affected the fate of detached cells. Flow cytometric analysis of shed cells stained with propidium iodide and annexin V demonstrated that the percentage of viable shed PTEC decreased significantly after cytokine treatment (Figure 4B), with significant increases in the percentages of necrotic and apoptotic PTEC. Acridine orange staining confirmed these results (Figure 4C). NOS inhibition did not significantly alter the fate of the shed cell population, demonstrating that NO does not alter the fate of cells shed into the medium (Figure 4, B and C).
PTEC Turnover Is Reduced after Cytokine Stimulation
To determine whether the observed increase in cell shedding could be
attributed to increased cell turnover, we investigated changes in PTEC
proliferation after cytokine treatment. In monolayers,
[3H]thymidine incorporation decreased from 6630 ± 1457
cpm/µg DNA in untreated cells to 1863 ± 90 cpm/µg DNA after
cytokine treatment (n = 5, P < 0.02). In the
supernatants, [3H]thymidine incorporation decreased from 7219
± 812 cpm/µg DNA (control) to 3983 ± 599 cpm/µg DNA (with
cytokines, P < 0.02). Addition of 1 mM L-NIL did not alter the
effects of treatment with cytokines alone. Therefore, cytokine treatment
decreased cell proliferation among PTEC.
Basolateral Cytokine Stimulation Induces Maximal NO-Dependent Cell
Shedding from Polarized PTEC Monolayers, in a Dose-Dependent Manner
Monolayer cultures are a limited model of polarized PTEC, and cytokines are
more likely to interact with the basolateral cell aspect in vivo
(18). Therefore, we compared
the effects of cytokines on cell detachment added at either the apical or
basolateral cell surface to polarized PTEC grown on inserts. Exposure of
polarized PTEC to cytokines on the basolateral aspect resulted in
significantly greater numbers of shed cells, compared with PTEC exposed to
cytokines on the apical surface (Figure
5A). This increase in cell shedding was correlated with increased
PTEC nitrite production, which was significantly greater when cells were
treated with cytokines on the basolateral aspect, compared with the apical
aspect (Figure 5B). Inhibition
of nitrite production with 1 mM L-NIL
(Figure 5B) was associated with
significant reductions in the numbers of cells shed from both apically and
basolaterally treated monolayers (Figure
5A).
|
We determined dose-response curves for shedding and nitrite production by
polarized PTEC stimulated basolaterally with increasing cytokine
concentrations (Figure 6). In
the presence of 10 ng/ml IL-1, increasing concentrations of IFN-
induced highly significant, dose-dependent increases in cell shedding
(Figure 6A) and NO production
(Figure 6B). Maximal cell
shedding was observed at IFN- concentrations of > 100 U/ml
(Figure 6A). In the presence of
200 U/ml IFN-, increasing IL-1 levels of >0.01 ng/ml induced
dose-dependent increases in cell shedding
(Figure 6C) and nitrite
production (Figure 6D). As
assessed using polarized PTEC, TNF- did not alter shed cell numbers or
NO production induced by the combination of IL-1 and IFN-
(Figure 6, E and F). The lack
of effect of TNF- on the level of NO production was observed in
experiments using polarized PTEC cultures. This contrasts with findings for
monolayer cultures, in which the addition of TNF- to IL-1 and
IFN- consistently increased NO output
(Table 1). Taken together,
these data demonstrate that exposure of the basolateral aspect of polarized
PTEC monolayers to IL-1 and IFN- induces maximal NO-dependent
cell shedding.
|
Proinflammatory Cytokines Disrupt ß1
Integrin-Mediated Cell Adhesion to Laminin, Independent of NO Production
To investigate the mechanism underlying cytokine-induced cell shedding, we
examined cytokine effects on cell adhesion to different basement membrane
proteins. With the use of high concentrations of matrix proteins, unstimulated
PTEC bound to laminin with significantly greater affinity, compared with Fn or
collagen type IV (Figure 7A).
After cytokine treatment, there was a highly significant reduction in cell
adhesion to laminin, compared with unstimulated cells
(Figure 7A), and a smaller, but
still significant, reduction in the binding of cytokine-treated cells to
collagen type IV. Inhibition of NO production with 1 mM L-NIL did not alter
the cytokine effects on cell adhesion. To further explore possible
NO-dependent effects on PTEC adhesion, we repeated the assays using lower
concentrations of matrix proteins (Figure
7B). At lower concentrations, there was a small, but still
significant, reduction in the binding of cytokine-treated cells to collagen
type IV, compared with unstimulated cells, but no differences in binding to
laminin after cytokine treatment. There were no differences in adhesion after
iNOS inhibition with L-NIL. Therefore, rebinding of the cells to the
extracellular matrix does not seem to be a NO-dependent phenomenon.
|
Immunofluorescence experiments demonstrated that, in control cells, there
was strong ß1 integrin staining at the basolateral surface of
the cells (Figure 8A). After
cytokine treatment, there was marked downregulation of basolateral
ß1 integrin expression throughout the epithelial layer [mean
peak fluorescence intensity, 206 ± 11.0 (control) versus 37.6
± 14.5 (cytokine-treated); n = 8; P < 5 x
10-7] (Figure 8B).
However, in regions of the monolayer where cells had become heaped up into
aggregates, there was strong ß1 integrin staining (median
value, 3.54% ß1 integrin-positive cells/field; range, 2.1 to
5.2% in four separate experiments) (Figures
3A and
8C), which was localized to the
apical compartment of the cells (Figure 3,
D to F). Cell clusters were typically raised 10 µm above the
epithelial monolayer (Figure
3). These PTEC aggregates also demonstrated strong staining for
3 integrins, in a pattern similar to that observed for
ß1 integrins (Figure
8D), suggesting that the 3ß1
integrin heterodimer was the predominant integrin that was apically
redistributed within regions of cellular aggregation. Interestingly,
aggregated cells displaying apically redistributed ß1
integrins demonstrated altered morphologic features, with cell process
formation. When the monolayers were stained for iNOS, the aggregates of
iNOS-positive cells were the same groups of cells with increased apical
ß1 staining (Figure 3, A to
C and G to I). Analysis of these dual-stained PTEC aggregates
demonstrated that 84.3 ± 3.43% of ß1 integrin-positive
cells were also positive for iNOS staining (in four separate experiments).
Although a proportion of ß1 integrin-positive cells were
negative for iNOS staining, these cells were invariably adjacent to
iNOS-positive cells within the aggregates. iNOS did not colocalize with
integrin staining (Figure 3, C and
I), and inhibition of iNOS with L-NIL did not alter the pattern of
integrin staining.
|
Exogenous NO Induces Cell Shedding from Polarized PTEC Monolayers,
Which Occurs in Part via a cGMP-Dependent Mechanism
To further explore the mechanism by which NO induces cell shedding, we used
a NO donor, DPTA, to deliver NO without additional, complicating, inflammatory
cytokine effects. When added to the basolateral compartment of polarized PTEC
monolayers in Transwells, DPTA induced a significant, dose-dependent increase
in cell shedding (half-maximal cell shedding at 46 µM DPTA)
(Figure 9A). Incubation of
cells with the control amine after 24-h decay at 37°C did not induce cell
shedding (data not shown). Therefore, exogenous NO alone is sufficient to
induce cell shedding.
|
To determine the biochemical target of NO, polarized PTEC monolayers were treated with 250 µM DPTA in the presence of increasing concentrations of the cGMP inhibitor ODQ (Figure 9C). ODQ produced a dose-dependent reduction in DPTA-induced cell shedding, reaching statistical significance at 10 µM concentrations (Figure 9B). Furthermore, ODQ significantly reduced the number of cells shed from polarized PTEC monolayers stimulated with inflammatory cytokines, although the decrease was not to the same level observed in control supernatants (Figure 9C). This finding suggests that NO-dependent cell shedding is mediated, in part, through the second messenger cGMP.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
,
IL-1, and IFN- produces a number of injurious changes in human
PTEC. The actin cytoskeleton is disrupted, E-cadherin is dispersed, and cells
develop an elongated cell shape, with extension of long processes. The
cytokines induce shedding of the epithelial cells into the growth medium, a
process that is dependent on NO synthesized by iNOS induced as a result of
cytokine action. The data demonstrate that basolateral exposure of polarized
PTEC to cytokines induces maximal NO-dependent cell shedding, which is
mediated in part through NO effects on cGMP levels. Such shedding of PTEC is a
key process contributing to tubular obstruction in sepsis-related ARF.
Shedding is accompanied by dispersal of basolateral ß1
integrins and E-cadherin, with a corresponding upregulation of integrin
expression in clusters of cells elevated above the epithelial monolayer. These
cells exhibit an apical pattern of ß1 integrin staining, as
well as upregulation of iNOS. Attachment studies demonstrate that the major
ligand involved in cell anchorage is laminin, probably through interaction
with the integrin 3ß1. This interaction is
downregulated by cytokine treatment but is not dependent on NO.
Establishment of epithelial cell polarity depends on cell adhesion to the
basement membrane through specific interactions between basolateral integrins
and matrix protein ligands
(37,38,39).
This depends on cooperation between integrins and the actin cytoskeleton. In
ischemic tubular injury, disruption of PTEC actin microfilaments and
integrin-dependent cell-matrix interactions results in loss of polarity,
apical redistribution of integrins, shedding of PTEC, and tubular obstruction
(8,9,10,12).
We now demonstrate that inflammatory cytokines, in the absence of an hypoxic
insult, induce similar changes in human PTEC in vitro and that
subsequent cell shedding is dependent on NO produced by iNOS. In the context
of sepsis-induced ARF in vivo, these findings suggest that cytokines,
through NO production, play a major role in the pathogenesis of tubular injury
early in the course of sepsis, before the onset of hypotension and renal
hypoperfusion (when circulating levels of inflammatory cytokines are high)
(20,35,40). In support of this concept, our data demonstrate that inflammatory
cytokines induced high NO output within PTEC
(Table 1) and significant
NO-dependent PTEC shedding (Figures
4,5,6).
This was independent of changes in cell turnover, because the cytokine mixture
decreased cell proliferation, as expected on the basis of the
antiproliferative actions of IFN-
(41). Furthermore, exogenous
NO delivered to the basolateral surface of polarized PTEC induced significant
cell detachment (Figure 9).
The significant increase in NO output and NO-dependent cell shedding after
basolateral cytokine stimulation of polarized PTEC
(Figure 5) may reflect greater
sensitivity of the IFN- receptor on the basolateral surface of
polarized epithelial cells
(42). This polarized in
vitro model more closely resembles in vivo sepsis, in which the
highest cytokine concentrations within the kidney are produced by infiltrating
inflammatory cells within peritubular capillaries and the interstitium
(18,43)
and hence are more likely to present to the basolateral aspect of tubule
cells. In addition, the range of cytokine concentrations that induced
dose-dependent increases in NO-dependent cell shedding
(Figure 6) reflect the levels
observed in sepsis in vivo
(35,40,44).
The contribution of NO synthesis to cytokine effects was determined by using two competitive arginine analogs to inhibit NOS, i.e., the specific NOS inhibitor L-NMMA and the highly specific, potent, iNOS antagonist L-NIL. L-NIL and L-NMMA, but not the inactive stereoisomer D-NMMA, inhibited NOx production and significantly reduced cell shedding, confirming that NO output was iNOS-dependent and cytokine-induced cell shedding was NO-dependent. Importantly, previous work investigating the role of NO in hypoxic mouse tubule cell injury demonstrated that the results of experiments using pharmacologic NOS inhibition were identical to the results of experiments with iNOS-knockout animals (27). Therefore, data from our in vitro studies using NOS inhibitors are likely to be of pathophysiologic relevance in vivo.
Despite the high NOx output measured from PTEC monolayers, immunofluorescence staining demonstrated that iNOS was expressed in a relatively small percentage of cells (Figure 2). The reason for this apparent discrepancy is unclear, although findings are similar to the pattern of iNOS immunoreactivity observed in a previous study using lung epithelial cells (36). Intriguingly, in this study, the cells that stained positively for iNOS were less attached to the culture surface, compared with nonstained cells.
How does NO induce cell shedding? Given that iNOS was expressed in the same
PTEC aggregates that displayed apically redistributed ß1
integrins (Figure 3), an
attractive hypothesis would be that NO exerts its effects through alterations
of integrin function. However, in the cell adhesion assays, although cytokines
downregulated binding to the 3ß1 ligand
laminin, we did not observe any NO-dependent changes. This discrepancy may
reflect the different processes underlying cell detachment and reattachment.
For example, cell reattachment involves binding of basolateral cell integrins
to matrix protein ligands, whereas detachment is a more complex process
requiring dissolution of both integrin-dependent cell-matrix adhesion and
cadherin-dependent cell-cell adhesion. Importantly, our data demonstrate that
endogenous and exogenous NO-induced shedding is mediated in part via a
cGMP-dependent mechanism (Figure
9). Previous studies demonstrated that NO interferes with the
establishment and maintenance of mesangial cell adhesion to extracellular
matrix components via a cGMP-dependent pathway
(45). Additionally, NO
inhibited chondrocyte migration, cell-matrix adhesion, and actin
polymerization in an in vitro model of inflammatory arthritis
(46,47).
These effects resulted from NO-mediated disruption of a RhoA-actin-focal
adhesion kinase complex, via a cGMP-dependent pathway, and these findings may
provide an explanation for the mechanism underlying NO-dependent shedding of
human PTEC.
Cytokines induced other PTEC alterations that resembled ischemic tubule
cell injury. Using a cell adhesion assay to measure integrin function, we
observed that cytokines significantly reduced PTEC adhesion to laminin and, to
a lesser extent, that to collagen type IV
(Figure 7). Differences in
binding were identified only by using high concentrations of coating proteins
(Figure 7), in contrast to
previous work that demonstrated, with immortalized epithelial cell lines, that
differences in adhesion could be resolved by using lower matrix protein
concentrations (33). These
differences may reflect lower binding affinities of primary human cells,
compared with cell lines. Our data suggest specific cytokine-induced
disruption of the interaction between the
3ß1 integrin and its ligand laminin. The
3ß1 integrin subtype is the most abundantly
expressed basolateral PTEC integrin; it binds with high affinity to laminin,
binds with lower affinity to collagen type IV, and does not bind to Fn
(37). We also observed that
ß1 integrin distribution was altered after cytokine treatment
(Figure 8), with downregulation
throughout the monolayer (Figure
8B) except in scattered regions of cell aggregation, which
exhibited high levels of 3ß1 integrin
expression (Figure 8D).
ß1 integrins were redistributed to the apical compartment of
cells within these aggregates (Figure
3). These data are in accord with results from models of ischemic
tubular injury, which demonstrated that tubule cell detachment is associated
with apical redistribution of 3ß1 integrins
(12,48).
In conclusion, we have demonstrated that proinflammatory cytokines cause targeted disruption of ß1 integrin-dependent cell-matrix adhesion and induce tubule cell shedding. Importantly, detachment of cells from PTEC monolayers, which is a major feature of ARF in vivo, is dependent on the production of endogenous NO in our in vitro model. Our work demonstrates that cytokines can mediate pathologic effects on human PTEC independently of ischemia, and it suggests that inflammatory cytokines play a major role in the pathogenesis of tubular injury in sepsis. These studies suggest that future therapeutic strategies aimed at specific iNOS inhibition early in the course of sepsis may protect proximal renal tubules from cytokine-induced injury. Inhibition of tubule cell NO production might inhibit cell shedding and delay the onset of ARF in sepsis.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
interferon and 5-fluorouracil on human colorectal
carcinoma cell lines. Int J Cancer46
: 61-66,1990[Medline]
modulation of
epithelial barrier function: Time course, reversibility, and site of cytokine
binding. J Immunol 150:2356
-2363, 1993[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  V. Cantaluppi, L. Biancone, G. M. Romanazzi, F. Figliolini, S. Beltramo, F. Galimi, M. G. Camboni, E. Deriu, P. Conaldi, A. Bottelli, et al. Macrophage Stimulating Protein May Promote Tubular Regeneration after Acute Injury J. Am. Soc. Nephrol., October 1, 2008; 19(10): 1904 - 1918. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Lindsay, S. Ramsey, M. Aitchison, T. Renne, and T. J. Evans Modulation of lamellipodial structure and dynamics by NO-dependent phosphorylation of VASP Ser239 J. Cell Sci., September 1, 2007; 120(17): 3011 - 3021. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Frangie, W. Zhang, J. Perez, Y.-C. X. Dubois, J.-P. Haymann, and L. Baud Extracellular Calpains Increase Tubular Epithelial Cell Mobility: IMPLICATIONS FOR KIDNEY REPAIR AFTER ISCHEMIA J. Biol. Chem., September 8, 2006; 281(36): 26624 - 26632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Y. Poon, E. Raharjo, K. D. Patel, S. Tavener, and P. Kubes Complexity of Inducible Nitric Oxide Synthase: Cellular Source Determines Benefit Versus Toxicity Circulation, September 2, 2003; 108(9): 1107 - 1112. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
), necrotic ([UNK]), and apoptotic (
) cells shed from PTEC
monolayers treated for 24 h with medium alone (controls), cytokines, or
cytokines plus 1 mM L-NMMA or 1 mM
L-N6-(1-iminoethyl)lysine (L-NIL). (B) Values shown are
means ± SEM (n = 3). Significant differences from control
values are indicated. *P < 0.02, +P < 0.05.
(C) Values shown are means ± SEM (n = 6). Significant
differences from control values are indicated. *P <
0.001, +P < 0.01.
P < 0.05 versus apical cytokines. (B)
Values shown are means ± SEM (n = 6). *P
< 10-7 versus controls, +P < 0.0001
versus cytokines, 
