Discussion
The primary goal of the study was to assess the potential of the oral intestine-specific NHE3 inhibitor tenapanor to prevent intestinal obstruction in CFTR null mice. Several major obstacles were noticed during the study preparation and had to be addressed: Tenapanor quickly precipitates from aqueous solutions at neutral or alkaline pH. Since microencapsulated tenapanor for use in mice is not available, it had to be applied by gavage, and dose finding studies were necessary. We aimed at finding a tenapanor dose and application frequency that resulted in a stable but mild increase in stool water, because diarrhea is quickly lethal for mice (Barone et al., 2009). 30 mgkg-1tenapanor, applied twice daily by gavage, resulted in a mild increase in stool water, and it did not significantly influence the body weight of the mice. Surprisingly, the required dose was the same inCftr -/- and Cftr -/-mice, suggesting relative independence of the CFTR mediated route of fluid secretion and the NHE3 mediated route of fluid absorption. It is likely that the microencapsulation of tenapanor available for use in humans will permit lower doses in the CF population. An optimal dosing regimen to prevent obstructive episodes in CF patients without causing diarrhea or other unwanted side effects needs to be established.
The Cftr -/- mice experienced a mild loss of body weight in both the tenapanor and vehicle group, which may be related to an increased stress level due to the twice daily gavage. The lack of effect of tenapanor on body weight is an important finding, because it is known that NHE3 function is required for PEPT1-mediated dipeptide absorption (Thwaites & Anderson, 2007). It is in agreement with a lack of reduction of PEPT1-mediated substrate absorption during short term tenapanor ingestion in healthy volunteers (Johansson et al., 2017). However, NHE3 has also been shown to be involved in the absorption of amino acids (Anderson & Thwaites, 2005) and micronutrients (Shawki et al., 2016). Therefore, careful observation is necessary during long term studies in CF patients.
Replacing the drinking fluid containing the osmotic laxative with tap water may result in an initiation of an obstructive episode within the first day (which becomes evident only several days later); we therefore started the tenapanor gavage on the last day of the oral laxative administration. It is well known that the incidence of obstructive episodes (always lethal) in Cftr -/- mice is very high during weaning (Snouwaert et al., 1992), and 90%Cftr -/- mice are dead by day 30 unless preventive measures are taken (Clarke, Gawenis, Franklin & Harline, 1996), and decreases in adulthood. Previous studies that aimed at pharmacological prevention of obstructive episodes inCftr -/- mice have therefore administered the pharmaceutical agent to mice aged 3-6 weeks (Walker, Simpson, Levitt, Boyle & Clarke, 2006), and 20 days, respectively (Lord et al., 2018). For our study, the mice had to tolerate twice daily gavage and preferably not loose body weight due to this procedure; we therefore chose the age range of 9-13 weeks. This explains why the percentage of obstructions is lower in our study compared to the studies mentioned above.
The primary goal of our study was the questions whether intestinal obstructive episodes can be prevented by tenapanor in a test group at high risk for obstructions, which is the CFTR null mouse (no CFTR protein present). Although our study was conducted in an age group in which the mice have a decreased risk compared to younger age, the results showed that the risk was still high (46% within the observation time) and was reduced to 8% with tenapanor treatment.
Apart from obtaining a positive result of the study, we wanted to understand how tenapanor conveyed this reduction in obstructive episodes at a cellular and molecular level. As previously studied in intestinally perfused mice, the direct effect of tenapanor will be a reduction of NHE3-mediated sodium and fluid absorption and an increase in the luminal alkalinity in both the small and large intestine (Tan et al., 2021), which was also seen in the stool of the mice in this study (Figure 1g). A normalization of the delayed GTT in the tenapanor-treatedCftr -/- mice is likely another key effect protecting the mice against obstructions.
Somewhat surprisingly, tenapanor treatment resulted in the reversal of other features of the intestinal abnormalities inCftr -/- mice. Of particular interest to us was the finding that tenapanor treatment reversed the cryptal hyperproliferation in the ileum and proximal colon. A hyperproliferative response has been observed previously inCftr -/- small (Gallagher & Gottlieb, 2001) and large intestine (Tan et al., 2021). The intracellular pH ofCftr -/- deficient enterocytes is more alkaline than that of corresponding Cftr +/+ cells (Simpson, Gawenis, Walker, Boyle & Clarke, 2005). This feature is preserved in the crypts including the stem cells of intestinal organoids (Liu, Walker, Cook, Ootani & Clarke, 2012) and is causally related to the increased proliferative rate, which is preserved inCftr -/- intestinal organoids (Strubberg et al., 2018). The presence of NHE3 has been reported both in undifferentiated and differentiated intestinal organoids (Foulke-Abel et al., 2016). Immunohistochemical staining of NHE3 overlaps with that of CFTR in the villus region and cryptal mouth region (Jakab, Collaco & Ameen, 2011). Therefore, it is possible that chronic tenapanor application reverses in part some of the pathological effects of the high intracellular pH on proliferative activity. A similar observation was made by Bradfordet al . in their studies of mice which were homozygous negative for CFTR, for NHE3, or for both genes. While bothCftr -/- and Nhe3 -/-mice displayed an increased cell proliferation in duodenal crypts, this increase was significantly reduced in the crypts ofCftr-/-/Nhe3-/- mice (Bradford, Sartor, Gawenis, Clarke & Shull, 2009). Thus, long term tenapanor treatment, if proven safe in CF patients, may even have to potential to lower the risk for GI malignancy.
Another striking observation in tenapanor treated mice was a reduction in intestinal mucus accumulation. We previously reported an increase in the goblet cell counts, as well as enlarged goblet cell theca, in the mid-distal colon of Cftr -/- mice that did not display evidence of inflammation (Kini et al., 2020; Tan et al., 2021). Because of the well documented crypt-villus hyperproliferation inCftr -/- intestine (Gallagher & Gottlieb, 2001; Tan et al., 2021), we used a fluorometric method of quantifying the Muc2 immunofluorescence in relation to that of the Dapi stained nuclei in this project. The total as well as the extracellular Muc2/Dapi immunofluorescence was significantly reduced in tenapanor vs vehicle-treated mice in the ileum and proximal colon, while it was very high when an obstruction was present.
We also noted a shift in the expression of the Muc genes, with a non-significantly lower expression of Muc2 mRNA and a significantly higher Muc1 mRNA. The 21 day treatment with tenapanor reduced Muc1 mRNA expression compared to vehicle treatment. The somewhat lower Muc2 mRNA expression was also seen in the present study, with no effect of tenapanor. It has been shown that Muc1 deletion, although not a major component of intestinal mucin, reduces the thickness of the firmly adherent mucus layer (Malmberg et al., 2006). Muc1 deleted CF mice have reduced amounts of intestinal mucus and a better survival compared to Muc1-expressing CF mice (Parmley & Gendler, 1998). It is therefore likely that the observed tenapanor induced changes in Muc1 expression and mucus accumulation play a major role in reducing obstructive episodes.
The intestinal tract of patients with cystic fibrosis display an inflammatory phenotype in the absence of bacterial infection or enzyme induced fibrosing colitis. Multiple explanations have been provided, including bacterial overgrowth due to dysmotility, oxidative stress, intracellular accumulation of mutated CFTR protein, and lack of pancreatic and/or intestinal defensins (Bruzzese et al., 2004; Crites et al., 2015; Galli et al., 2012; Lisowska, Mdry, Pogorzelski, Szydłowski, Radzikowski & Walkowiak, 2010; Raia et al., 2000; Werlin et al., 2010). In mice, chronic exposure to PEG containing laxative in the drinking fluid both increased survival and reduced the inflammatory signature in the intestine of these mice (Bradford, Sartor, Gawenis, Clarke & Shull, 2009; Clarke, Gawenis, Franklin & Harline, 1996). We had previously reported no difference in proinflammatory cytokines in the different segments of Cftr -/- andCftr +/+ littermates while on the PEG containing laxative, yet several features that had been correlated with an absent CFTR function (in addition to the reduced anion and fluid secretory response) were present: increased enterocyte pHi, an increased mucus content in the goblet cell theca, crypt-villus elongation, decreased surface pH with fluid hyper-absorption (Kini et al., 2020; Tan et al., 2021). In this study, a discreet but significant increase in the number of mast cells was observed in the lamina propria of the intestinal tract of the vehicle treatedCftr -/- mice compared toCftr +/+ mice, which was completely or partially prevented by tenapanor treatment. A very discreet increase in neutrophil infiltration was seen in Cftr -/- ileum and mid-distal colon, not accompanied by an increase in any of the proinflammatory cytokines. In contrast, a strong increase in neutrophil infiltration and in the expression of proinflammatory cytokines was only seen in mice with an obstruction, and was restricted to the intestinal segments proximal to the obstruction that were filled with fecal matter. These segments also exhibited severe alterations in the mucosal architecture. Thus, the intestinal inflammation was a sequelae of the obstruction, and not its cause.
A relative shortcoming in our study is that the application mode of the drug is relatively invasive, and therefore the study duration is fairly short. The number of mice permitted for the study by the animal committee was sufficient to reach the primary endpoint, but was not enough to determine significance for the reversal of every studied CF-associated intestinal abnormalities by tenapanor treatment in every intestinal segment. When we performed the GTT experiments, we noticed that the stool water content was still elevated in theCftr +/+ mice 11 hours after the last dosing, but not in the Cftr -/- mice, suggesting that more frequent dosing, not feasible with this form of application, may have been optimal. Obviously, this problem will not exist in CF patient trials. For a tenapanor trial in CF patients, it will be necessary to define the patient population at high risk for intestinal complications, and preferably test its prophylactic potential. Patients with meconium ileus at birth and prior DIOS episodes, and lung or liver transplanted patients are at particularly high risk, and CF diabetes, pancreatic insufficiency and female gender also carry a higher risk (Houwen et al., 2010; Lavie et al., 2015; Munck et al., 2016). While corrector and potentiator therapy is beneficial for CFTR-dependent functional outcome parameters (Graeber et al., 2022; Ooi et al., 2018), adjunctive therapy specifically targeting the constipation and obstructive episodes appears necessary in many CF patient subgroups.
Tenapanor treatment is considered safe for humans (Block et al., 2021; Sinagra et al., 2020). It has not been tested in children so far, and the longest observation time in a published trial is 52 weeks. Given the good safety profile, clinical trials testing the effect of tenapanor on the intestinal symptoms of CF patients seem warranted.