In this manuscript we present evidence that mice treated with an active synthetic APF derivative [Galβ1-3GalNAcα-O
-TVAAVVVA] have decreased bladder epithelial repair following intravesical acetic acid instillation. This finding mimics the bladder epithelial thinning/ulceration found in biopsies from IC/PBS patients [1
] as well as the decreased proliferation seen in both bladder epithelial cells explanted from IC/PBS patients and as
-APF-treated primary normal bladder epithelial cells [13
]. In addition, our data indicate that as
-APF treatment of these mice also resulted in decreased expression of UPIII and ZO-1 in the regenerating epithelium, similar to abnormalities in expression of these proteins found in bladder biopsy specimens from IC/PBS patients in vivo
], as well as explanted epithelial cells from IC/PBS patients and as
-APF-treated primary normal bladder epithelial cells in vitro
Finding a reliably effective treatment for IC/PBS has been hampered for many years by the lack of a well-understood inducible animal model with histologic/biochemical features shared by the disease [22
]. Many putative animal models for IC/PBS have been developed to date, but evidence for abnormal expression of the same bladder epithelial proteins as documented in human biopsy specimens is scant for these models. For example, although the feline IC model and the protamine sulfate rat model share evidence for decreased bladder epithelial tight junctions [51
], only the feline IC model has been documented to have decreased ZO-1 expression and increased epithelial permeability [51
]. Like bladder tissue from IC/PBS patients, the feline IC model also has been shown to have increased iNOS [23
]. However, unlike patients with the human illness (who exhibit in
creased P2X2 or P2X3 urothelial receptors) cats with feline IC have de
creased bladder epithelial P2X1 receptor expression without abnormalities in P2X2 or P2X3 expression [53
]. In addition, because the etiology of feline IC remains unknown it is not an inducible model, and because of its sporadic and spontaneous nature cats with feline IC can be difficult for many researchers to obtain.
Similarly, the relationship of the antigen-induced, virally-induced, chemical toxin-induced, or autoimmune models of cystitis to the human disease are also not clear, as there is no definite evidence for a similar pathogenesis (with a defined antigen, virus, toxin, or autoantigen) for the human disease [22
]; therefore, although these models are more readily available, their utility for testing therapeutic or preventive agents for this syndrome also remains unknown. Abnormally expressed bladder epithelial cell genes similar to those found in bladder biopsies of IC/PBS patients have been limited to UPIII [decreased in acrolein-induced cystitis [54
] and cyclophosphamide (CYP)-induced cystitis [55
]], and iNOS [increased in the CYP model] [56
]. However, bladder epithelial cell UPIII expression rapidly normalizes (within 24 hours) in the CYP-induced cystitis model, and bladder hyperreflexia in these models can similarly resolve within days [57
] indicating the transient nature of some of these changes in these models. Therefore, additional studies are also required to compare the durability of abnormal bladder epithelial gene expression in all of these models for us to understand the relative utility of each model for studies of IC/PBS pathogenesis and treatment.
The model described in this preliminary report has certain advantages over the other animal models that have been described to date. It is an inducible rodent model, making it potentially more readily available than the spontaneous feline model. This mouse model also has changes in expression of two proteins similar to changes found consistently in patients with IC/PBS – UPIII and ZO-1. Unlike some protein expression abnormalities found in the other models to date, these changes appear to be durable (sustainable) for up to 3
weeks. And whereas most of the other models are induced by factors that are almost certainly not causative in the pathogenesis of IC/PBS, this model is based on a potential etiology of IC/PBS (APF) that was first discovered in urine from patients with this illness, subsequently shown to be made by bladder epithelial cells in these patients and to induce abnormal expression of at least six epithelial proteins known to be abnormally expressed in bladder tissue from IC/PBS patients [14
]. Because APF inhibits bladder epithelial cell replication but does not by itself induce cell death in vitro
] we previously hypothesized that APF may inhibit bladder epithelial repair following injury caused by other inciting factors such as a urinary tract infection [16
]. However, the use of acetic acid to induce such bladder epithelial injury prior to APF instillation in the current model is also by itself not related to the pathogenesis of IC; future refinements to the model may therefore include induction of epithelial shedding by bacteria known to cause cystitis [48
] prior to APF instillation.
Disadvantages of this model, however, include its requirement for daily intravesical instillations of active synthetic APF (or its control peptide), making it relatively labor-intensive and expensive for a rodent model. In addition, some of the variability seen in this model regarding the effects of APF on epithelial repair or gene expression may result from limited (only 3 hours daily) exposure to the instilled APF. However, based on the preliminary results from this model, development of a continuous APF release model (via an implanted pump), or a transgenic mouse model that would express active APF either constitutively or inducibly, appear to be warranted.