Click to edit Master title style
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
5/14/2014
#
Cathelicidin augments epithelial receptivity and pathogenesis in experimental E. coli cystitis
Elizabeth S. Danka
Graduate Student
Molecular Cell Biology Program
Division of Biology and Biological Sciences
Washington University in St. Louis
[email protected]
Campus Box 8208
Funding sources
National Science Foundation Graduate Research Fellowship
DGE-1143954
National Institutes of Health grant R01-DK080752 and P50-DK064540
Abstract
UTI and cathelicidins
Elizabeth S. Danka* and David A. Hunstad
Departments of Pediatrics and Molecular Microbiology, Washington University
Background: Cathelicidin antimicrobial peptides (AMPs) are traditionally viewed as important for the defense of epithelial surfaces, including that of the urinary tract, and are active against many bacteria, including the uropathogenic E. coli (UPEC) used in this study. In addition to their bactericidal activities, AMPs have been shown to be immunomodulators that recruit immune cells to further promote clearance of bacteria.
Methods: C57Bl/6 mice deficient in the cathelicidin CRAMP (cathelin-related antimicrobial peptide; ortholog of human LL-37) and wild-type C57Bl/6 mice were infected with UPEC strain UTI89 and the severity of the infection was measured by bladder titer quantification, intracellular bacterial community (IBC) formation, urine analysis, and histology. The immune response was characterized through quantification of invading neutrophils and cytokine analysis.
Results: Using microscopy and histology, we were able to detect bacteria forming IBCs within the bladder epithelial cells (BECs) of both wt and CRAMP KO (knock-out) mice at multiple time points, indicating an active infection. Surprisingly, KO mice had significantly lower bladder titers at 1, 6, 16, 24 and 48 hours post-infection (hpi), and resolve infection faster than wt mice. KO mice had less edema and fewer infiltrating neutrophils at 6 and 24 hpi, reflected in lower bladder cytokine levels in KO mice at multiple time points. KO mice liberated significantly more bacteria in the urine at 1 hpi, although there were no differences in initial BEC exfoliation between the two strains of mice. Though there were no apparent alterations in BEC structure in KO mice, epithelial binding and invasion by UPEC was less efficient in KO mice.
Conclusions: In our murine UTI model, we confirmed that cathelicidin is an important driver of the immune response to infection. However, our data indicate that cathelicidin is not required to clear bacteria from the urinary tract, and that mice that are deficient in CRAMP have less severe infections than wild-type mice that express this AMP.
UPIIIA
Conclusions
Using a murine model of UTI, we characterized the progression of cystitis in the absence of the murine cathelicidin CRAMP.
CRAMP is not essential for the defense of the urinary tract, as CRAMP-deficient mice displayed lower bladder bacterial titers and resolved cystitis more rapidly than wild-type mice. Neutrophils derived from the bone-marrow of CRAMP-deficient mice killed UPEC as well as those from wild-type mice (data not shown).
UPEC binding and invasion was impaired in CRAMP-deficient mice within the first hour of infection. This defect was not due to changes in the piliation state of UPEC, the rate of exfoliation of bladder epithelial cells, or cAMP-driven expulsion of bacteria-containing vesicles.
Lack of CRAMP did not affect intracellular growth or IBC formation by UPEC in bladder epithelial cells, although fewer IBCs formed in the bladders of mice that do not express CRAMP.
CRAMP-deficient mice exhibited less robust immune responses to infection, as determined by cytokine expression, edema, and neutrophil influx to bladder tissue. We attribute this difference to the role of CRAMP as a driver of the inflammatory response.
Future experiments will determine how early pathogenic events are disrupted in CRAMP-deficient mice.
We will continue to investigate the role of CRAMP as an immune stimulant in this model, and characterize the extent to which the difference in inflammatory response is due to altered bladder titers.
UPEC must bind to mannosylated residues called uroplakins (UPs) on the surface of bladder epithelial cells in order to be internalized and grow intracellularly. CRAMP-deficient mice expressed both UPIa and UPIIIa at similar levels (Figure 6A).
CRAMP-deficient mice did not exfoliate more of their bladder epithelial cells (along with attached/invaded bacteria) at 1 h post-infection (hpi) than wild-type mice (Figure 6B).
Activation of TLR4 by bacterial LPS can increase intracellular concentration of cAMP, triggering exocytosis of internalized bacteria. CRAMP has been shown to mask LPS recognition by TLR4. However, the bladder epithelial cells of CRAMP-deficient mice did not have the significantly higher intracellular cAMP levels that would drive more frequent expulsion of internalized UPEC (Figure 6C).
Type 1 pili are the fimbriae that bind to uroplakins on epithelial cells. There was no difference in the presence of type 1 pili on the surface of UPEC that were or were not exposed to CRAMP in vitro (Figure 6D).
Figure 6: (A) CRAMP-deficient mice express similar levels of UPIa and UPIIIa on the surfaces of their bladder epithelial cells, as quantified by Western blot and visualized using immunofluorescence microscopy. (B) Exfoliation of bladder epithelial cells at 1 hpi as quantified by UPIIIa expression in bladder homogenates. (C) cAMP expression in uninfected bladder epithelial cells and at 30 min post infection. Forskolin is a positive control that induces cAMP production, and DMSO is a vehicle control. (D) Piliation state of bacteria that have been exposed to CRAMP as determined by TEM (84.8% vs 78.2%).
C57Bl/6
CRAMP KO
Nuclei UPIIIa
Factors affecting early infection
A
B
C
D
Urinary tract infections (UTIs) are caused by a variety of bacteria and can affect different parts of the urinary tract separately or together. Infection of the bladder is called cystitis, and infection of the kidneys is called pyelonephritis.
Women are most susceptible to UTI. Over 50% of women will have at least one UTI in their lifetime, and 25% of the women that experience a UTI will have a recurrence within 6 months.
UTIs are responsible for >10 million physician visits each year and are associated with >$4 billion in annual health care costs.
Identification of bacterial components that allow uropathogens to subvert the immune system will allow for development of more effective therapies.
Uropathogenic Escherichia coli (UPEC) are the most common cause of UTIs. UPEC express a variety of virulence factors, including type 1 pili (which mediate attachment to bladder epithelium), P pili (which mediate attachment to kidney epithelium), LPS, and other components.
UPEC bind the superficial epithelial cells lining the urinary tract using pili, are internalized into these cells, and continue to survive and multiply (Figure 1A-B). Bacteria are protected from components of the immune system and antibiotics within this niche.
Using a murine model of cystitis, we can investigate how various bacterial factors and host factors contribute to the pathogenesis of UTI. In this model, female, 8-10 week old C57BL/6J mice are infected transurethrally with a suspension of approximately 107 bacteria in sterile saline.
AMPs are short, naturally-occurring, 12-40aa peptides that cause bacterial cell death through a variety of mechanisms. These peptides comprise part of the innate immune system that is important for counteracting the initial stages of infection.
In humans, the most important classes of AMPs are the cathelicidins and the defensins. Humans and mice both express a single cathelicidin: LL-37 and CRAMP (cathelin-related antimicrobial peptide), respectively.
CRAMP is expressed by the uroepithelial cells that line the urinary tract, and is also produced by neutrophils that are recruited to the site of infection.
UPEC are susceptible to the murine ortholog of LL-37, known as CRAMP, in in vitro survival assays (Figure 2).
Figure 1: UPEC express type 1 pili (A) that help them bind and invade into superficial bladder epithelial cells (B) (unpublished data and Mulvey et al, 1998).
A
B
Figure 2: Patient-derived UPEC strain UTI89 shows decreasing survival after exposure to increasing concentrations of the murine cathelicidin CRAMP.
Using a murine model of cystitis, we have characterized the contribution of CRAMP with respect to the development and pathogenesis of UTI, including intracellular bacterial growth and the immune response to infection.
Figure 3: (A) Bladder titers are lower in CRAMP-deficient mice than in C57Bl/6 mice. (B) The urines of CRAMP-deficient mice contain more bacteria at 1 hpi. (C) Using an in vivo binding and invasion assay, CRAMP-deficient mice are found to have fewer bacteria in the lumen of the bladder and within bladder epithelial cells at 1 hpi. (D) Confocal microscopy shows that bacteria form early and mature IBCs and filaments in bladder epithelial cells, regardless of host strain, and PMNs are recruited to IBC-bearing cells. (E) At 16 hpi, lacZ staining shows that CRAMP-deficient mice have fewer IBCs than wild-type mice.
Figure 4: There are no detectable differences in cytokine levels (A) or MPO expression (B) in the bladder tissue of uninfected mice. At 6 hpi, C57Bl/6 mice have significantly higher bladder cytokine levels (B), as well as more neutrophils present in bladder tissue (as measured by MPO)(D). By 24 hpi, there are no longer any significant differences in the bladder cytokine levels (E) or neutrophil presence in the bladder tissue (F).
A
B
C
6 hpi
24 hpi
C57Bl/6
CRAMP
Figure 5: (A) By H&E-stained morphological analysis, bladders of C57Bl/6 mice have worse edema than CRAMP-deficient mice at 6 and 24 hpi. (B) The bladder uroepithelium of CRAMP-deficient mice recovers faster from exfoliation than that of C57Bl/6 mice, as quantified by UPIIIa Western blotting.
A
B
A
B
C
D
E
CRAMP-deficient mice have lower bladder bacterial loads than C57Bl/6 (wild-type) mice at all time points examined when infected in parallel with 107 UPEC (Figure 3A).
Bacteria must bind to the epithelial cells that line the lumen of the bladder to avoid being lost by the flushing action of urine. Significantly higher UPEC titers were recovered from the urines of CRAMP-deficient mice at 1 hpi, which may indicate a binding defect in these mice (Figure 3B).
An in vivo binding and invasion assay was used to quantify the early stages of UPEC infection in both CRAMP-deficient and C57Bl/6 mice. CRAMP-deficient mice have significantly fewer bacteria within the lumenal compartment of the bladder at 6 hpi and also have significantly lower numbers of bacteria within a gentamicin-protected, intracellular niche in the bladder (Figure 3C).
UPEC that invade into bladder epithelial cells go through a defined growth pathway called the intracellular bacterial community (IBC) pathway that is characterized by small collections of bacteria at early time points, large, biofilm-like collections by 6 hpi and filamentous bacteria that flux out of the epithelial cells. In addition, neutrophils are recruited to host cells that contain IBCs, in an effort to fight the infection. All of these characteristics can be identified in the bacteria that grow within CRAMP-deficient cells, indicating that a defect in intracellular growth is not likely the cause of the lower titers throughout infection (Figure 3D).
However, significantly fewer IBCs are found within the bladders of CRAMP-deficient mice, again indicating that the infectious process is disrupted early during invasion, before UPEC begin to form these collections (Figure 3E).
Both wild-type and CRAMP-deficient bladders show signs of edema at 6 hpi, but by 24 hpi, CRAMP-deficient bladders have no edema, while C57Bl/5 bladders are still very enlarged and edematous (Figure 5A).
Wild-type bladders appear to have more robust immune cellular infiltrates than CRAMP-deficient mice at 6 hpi and 24 hpi (data not shown).
Uroepithelial exfoliation and recovery were tracked by quantification of UPIIIa (normalized to a total protein loading control) within bladder homogenates. Although the initial rate of exfoliation (at 1 hpi) was equivalent, wild-type mice had shed more of their uroepithelium by 6 hpi. Furthermore, CRAMP-deficient mice completely restored their uroepithelium by 48 hpi, while C57Bl/6 bladders still had only ~ 30% of their baseline levels of UPIIIa (Figure 5B).
Baseline differences in immune activation due to CRAMP deficiency could explain the early differences seen during infection. A bead-based cytokine array of 23 common cytokines did not reveal any differences in cytokine level in uninfected bladders of wild-type or CRAMP-deficient mice (6 representative cytokines are shown, Figure 4A).
Furthermore, there did not appear to be more resting neutrophils present in CRAMP-deficient bladders, as measured by myeloperoxidase (MPO) quantification (Figure 4B). These data indicate that baseline immune activation differences likely do not account for the lower titers seen as early as 1 hpi in CRAMP-deficient mice.
Analysis of cytokine expression in bladder tissue at 6 hpi reveals that C57Bl/6 mice have significantly higher levels of 16 of 23 cytokines tested (Figure 4C). The same representative cytokines are shown; IL-5, IL-10, IL-12 (p40), IL-12 (p70), IL-13, eotaxin, GM-CSF, IFN-γ, MIP-1a and MIP-1b were also significantly higher in wild-type mice.
In addition, MPO analysis of bladder homogenates reveals that significantly more neutrophils are recruited to the bladders of wild-type mice in response to infection than in CRAMP-deficient mice (Figure 4D). This is likely driven by the differences in cytokine expression.
By 24 hpi, there are no longer statistically significant differences in bladder cytokine levels in C57Bl/6 or CRAMP-deficient mice (Figure 4E).
Likewise, differences in the neutrophil content of bladder tissue, as measured by MPO analysis, have resolved by 24 hpi (Figure 4F).
D
E
F
CRAMP is not essential for defense of the urinary tract
Click to edit Master title style
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
5/14/2014
#
Click to edit Master title style
Click to edit Master subtitle style
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
5/14/2014
#
Click to edit Master title style
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
Click to edit Master text styles
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
5/14/2014
#
5/14/2014
#
Click to edit Master title style
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
Click to edit Master text styles
5/14/2014
#
