* Safety Assessment US, AstraZeneca Pharmaceuticals, Wilmington, Delaware, 19850; DMPK, AstraZeneca Pharmaceuticals, Wilmington, Delaware, 19850;
Neuroscience, AstraZeneca Pharmaceuticals, Wilmington, Delaware, 19850;
Target Biology, AstraZeneca Pharmaceuticals, Wilmington, Delaware, 19850; ||| Chemistry, AstraZeneca Pharmaceuticals, Wilmington, Delaware, 19850; and || Safety Assessment UK, AstraZeneca Pharmaceuticals, Alderley Park, Macclesfield, Cheshire SK10 4TG, United Kingdom
Received May 26, 2004; accepted August 11, 2004
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: -secretase; Notch receptor; Notch intracellular domain; NICD; intestinal goblet metaplasia; Rath1; Hes1; adipsin; atonal.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Recently Searfoss et al. (2003) reported that a potent Notch modulating
-Sec-I caused intestinal goblet cell metaplasia in vivo in Fisher 344 rats and induced a putative Hes-1 regulated factor called adipsin, which may serve as a toxicity biomarker for the lesion. Similar lesions caused by a potent Notch modulator, but not by a weaker Notch modulator, were observed in mice (Wong et al., 2004
). We confirm and extend these findings by demonstrating that the goblet cell metaplasia first observed in intestinal crypts of Han Wistar rats is preceded by the mRNA induction of the Hes-1 regulated Rath1. It is also accompanied by increased levels of intestinal mucus and an mRNA expression profile representing secretory lineage differentiation, and is followed by crypt epithelial regenerative hyperplasia. While other putative substrates for
-secretase are described in the literature, only a subset of them have been demonstrated to be true substrates in nonrecombinant systems. Furthermore, the early and sustained upregulation of Rath1 is suggestive of a causative link between Notch modulation and the metaplasia. Moreover, the absence of such toxicity in animals dosed with a weak Notch modulating
-Sec-I suggests that it may be possible to inactivate the processing of
-secretase substrates selectively and thereby avoid such pharmacology related toxicity.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
SUP-T1 NICD assay. Human T-cell lymphoma SUP-T1 cells cultured in RPMI1640 media supplemented with 1 mM sodium pyruvate, 2 mM Glutamax I, 1x penicillin/streptomycin and 10% fetal calf serum, exposed to -Sec-I for 16 h. Harvested cells were washed with 200 µl of Hanks buffer once and resuspended in 13 µl lysis buffer. Cell lysates were cleared by a 5 min spin at 2800 rpm. Protein concentrations were determined using the BCA kit from Pierce (Rockford, IL). Western blotting was performed by standard approaches. Twenty µg of total protein was resolved electrophoretically on a NuPAGE 7% Tris-Acetate gel in SDS-Tris acetate buffer. Protein was transferred onto a 0.2 µm nitrocellulose membrane, blocked with 5% non-fat dry milk in Tris-Buffered Saline Tween-20 (TBS-T), and incubated overnight at 4°C with an anti-Notch rabbit polyclonal antibody (Val1744) (Cell Signaling Tech.) at a dilution of 1:1000 in the same buffer. Between washes contained TBS-T buffer. The membrane was then probed with goat-anti rabbit horesradish peroxidase (HRP) and ECL substrate. Chemiluminescence signals were then detected by exposing x-Ray films for 210 min. Signals were digitized on a flatbed scanner and quantified with ONE-Dscan 2.03 (Scanalytics, Inc.).
Animals and treatments. All animal treatment and anesthesia procedures were carried out in accordance with recommended guidelines set forth by the AstraZeneca (Wilmington, DE) Animal Care and Use Committee. Male Han:Wistar substrain Crl:WI(Glx/BRL/Han)IGSBR rats were acquired from Charles River (Cambridge, MA). All rats were approximately 78 weeks of age at the start of the study. Animals were fed ad libitum with a standard certified pelleted diet of Purina Mills International (PMI) Rodent Laboratory Chow Item # 5002. Water was provided ad libitum. Animals were housed individually and randomly assigned to test groups of n = 3. The doses of 0, 10, 30, and 100 µmol/kg selected for the three test items were based on approximate efficacy values determined from an in vivo APP transgenic mouse model used for screening Aß40 lowering in brain and plasma. The high dose of 100 µmol/kg b.i.d. was found to be sufficient to lower plasma Aß40 in this model for all three compounds tested (see below and Table 1). The AS was solubilized in 0.1 M Meglumine, 2% (w/v) dextrose, pH 9 (solution); the DBZ was suspended finely in 0.5% (w/v) hydroxypropyl methylcellulose (Methocel E4M) and 0.1% (w/v) Tween 80 in water, and the BZ was suspended finely in 6% (v/v) ethanol/94% (v/v) Labrafil M 1944 CS (for chemical structures, see Fig. 1). Rats were dosed for up to 5 days i.p. b.i.d. using dosing volumes of 5 ml/kg for the AS, and 2.5 ml/kg for the BZ and DBZ. Control animals received corresponding vehicles and volumes. In addition, in time-course studies, three control groups were employed and sacrificed after days 1, 3, and 5 (n = 3 each, totaling n = 9), and 5 groups each of n = 3 treated with the mid-dose 30 µmol/kg b.i.d. of BZ were employed and sacrificed after 15 days of dosing. Tissue sections were examined on each of days 15, after dosing, in order to define temporally the histopathological changes. At the end of the dosing regimens, animals were fasted overnight. The following morning, animals were anesthetized with carbon dioxide and euthanized by exsanguination from the abdominal aorta. Tissue sections were either snap frozen in liquid nitrogen for total RNA isolation or fixed in 10% neutral buffered formalin for histopathological assessment.
|
|
In vitro MDR1-MDCK cell monolayer transport experiments. The transport of -Sec I compounds across MDCK cells transfected with the human P-glycoprotein gene MDR1 was assessed in triplicate at a single time point of 60 min, at 1 and/or 10 µM, and in the presence or absence of a potent and specific inhibitor (GF120918, 2 µM; gift from Glaxo SmithKline, Research Triangle Park, North Carolina) as described elsewhere (Polli et al., 2001
).
Determination of the plasma concentration of -Sec-I and pharmacokinetic analyses. Plasma was analyzed for test compounds by LC-MS-MS (MicroMass Quattro Ultima, Manchester, UK) by direct injection of supernatant following precipitation of proteins with 2 volumes of acetonitrile. Plasma protein binding was determined in duplicate at a concentration of 1 µM by equilibrium dialysis (37°C, overnight). Non-compartmental estimation of pharmacokinetic parameters was carried out with WinNonlin Enterprise 3.1 (Pharsight Corp., Mountain View, CA).
Histology and immunohistochemistry. Sections were dehydrated in graded alcohols, embedded in paraffin wax, sectioned at 5 µm, stained with hematoxylin-eosin, and examined by light microscopy. In addition, Alcian/blue periodic acid-Schiff (PAS) and Alcian blue/high iron diamine (HID) staining were carried out to ascertain the acidic versus basic and sulphomucin versus sialomucin content of goblet cells, respectively.
Cleaved caspase 3 (apoptosis marker). Sections of intestine from all control and treated animals per time point were pretreated for 10 min with 3% hydrogen peroxide. Antigen retrieval was performed using 10 min microwave pretreatment with 1 mM citrate solution. After a TBS-T wash, nonspecific protein labeling was blocked for 15 min (Protein Serum-free Block, Dako). The primary antibody (rabbit polyclonal anti-cleaved caspase-3; Cell Signaling Technology, 9661L) was diluted 1:300 and applied to sections for 60 min at room temperature. Slides were washed in TBS-T, and detection of antibody staining was then carried out using a rabbit labeled polymer method (Rabbit Envision, Dako) for 30 min and chromogenic development for 10 min with diaminobenzidine (DAB; A. Menarini). Negative controls of buffer with omitted primary and a rabbit isotype immunoglobulin (IgG) control (Dako, X0903) were used. Cleaved caspase-3 immunopositivity was scored semiquantitatively in each section by microscopically assessing the numbers of positive cells present per section. The scoring range was 0 (no positive cells) to 5 (highest numbers of positive cells).
Chromogranin A and B (pan-neuroendocrine cell marker). Sections of distal jejunum from two control and two treated animals per time point were pretreated for 10 min with 3% hydrogen peroxide. The primary antibody (rabbit polyclonal anti-Chromogranin A + B; Maine Biotechnology, PAB925NS) was diluted 1:30 and applied to sections for 60 min at room temperature. All other methodology is as described for cleaved caspase-3. Positive cells were scored semiquantitatively in the section. The score ranged from 1 (minimal positive cells) to 3 (maximal positive cells).
Total RNA isolation and gene expression profiling. To identify statistically significant changes in gene expression associated with -Sec-I treatment, a transcriptional profiling strategy was employed using Affymetrix high-density oligo microarrays (Santa Clara, CA). Starting material for these experiments was individual Han Wistar rat duodenum total RNA isolated from three animals per treatment group (n are listed later for each group in the captions for Figures 6 and 7, and in 4 to identify those instances where n < 3) Expression profiling of total RNA samples was performed using the Affymetrix rat array U34A according to the manufacturer's instructions. Briefly, total RNA was isolated using the RNeasy purification system with DNAse treatment according to the manufacturer's instructions (Qiagen, Valencia, CA). RNA concentration was determined spectrophotometrically (260/280 nm); RNA integrity was assessed by an Agilent Bioanalyzer. Samples were aliquoted to 10 µg and used as a template for double-stranded cDNA synthesis. Reverse transcription was primed using a T7-modified oligo-dT primer (5'GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGGT24-3'). In vitro transcription was performed on double-stranded cDNA synthesis, using a kit from Enzo Diagnostics. The resultant cRNA products were purified using RNeasy columns (Qiagen) and quantitated spectrophotometrically. For each sample, 20 µg of in vitro transcribed cRNA was fragmented by heating the sample to 94°C in the presence of 40 mM Tris-acetate, pH 8.1, 100 mM potassium acetate, and 30 mM magnesium acetate. Purity and quality of nucleic acids were evaluated at every step by Bioanalyzer. Hybridization cocktails contained 0.05 µg/µl of fragmented cRNA; 50 pM control B2 oligonucleotide; 1.5, 5, 25, and 100 pM of BioB, BioC, BioD, and Cre spiked cRNAs, respectively; 0.1 mg/ml herring sperm DNA; 0.5 mg/ml acetylated bovine serum albumin (BSA); 100 mM 2-(N-morpholino)ethanesulfonic acid (MES); 1 M NaCl, 20 mM ethylene diamine tetraacetic acid (EDTA); and 0.01% Tween-20. Hybridization, washing, and scanning were performed according to Affymetrix's recommendations. Image acquisition and segmentation were performed using a GeneArray Scanner (Agilent and MicroArray Suite 5.0 (Affymetrix)). Tab-delimited text file outputs containing raw fluorescence intensity data were exported for downstream analysis to Excel or GeneSpring 5.0. Gene lists were generated by comparing raw data values of each treated group to controls using one-way analysis of variance (ANOVA) p < 0.05 with a Benjamini-Hochberg False Discovery correction, Venn analysis, and a 1.5-fold cut-off for two of the five treatment groups in the 5-day time course.
|
|
Adipsin
Rath1
Rath1 and adipsin Western blotting of feces extract. Stool samples were placed in freshly prepared TBS-T containing 2x protease inhibitor cocktail sets I & III (Calbiochem, San Diego, CA) and homogenized on ice at full speed for 30 s using a Power Gen 1800G homogenizer with a 7 mm x 95 mm probe. The samples were centrifuged at 500 x g for 10 min at 4°C in a Hereaus table-top centrifuge, and the supernatant was collected. The supernatant was then centrifuged at 10, 000 x g for 5 min and again collected. Protein was determined by the method of Lowry using a kit from Pierce (Rockford, IL). Protein was resolved on 412% NuPage Novex Bis-Tris Polyacrylamide gradient gels (in MES buffer) and electrophoretically transferred to PVDF membranes. Membranes were blocked in TBS-T containing 1% BSA and probed with adipsin antibodies in the same buffer at 1:10, 000 dilution, and with Math1 antibodies in the same buffer at 1: 1,000 dilution. The membranes were probed with secondary rabbit anti-sheep HRP or goat anti-rabbit HRP for 1 h at room temperature. After they were washed, the blots were incubated in ChemiGlow West ECL substrate for 5 min (Alpha Innotech). Chemiluminescence signals were then detected with an Alpha Innotech Corp digital camera fitted with appropriate filters.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Pathology. After administration of DBZ and BZ at 10, 30, or 100 µmol/kg b.i.d., there was distension of the stomach and the small and large intestines, as well as an accumulation of mucus within the intestines. The severity of these findings was dose-dependent. In the time-course study, similar intestinal changes were noted from day 1 onward after treatment with 30 µmol/kg b.i.d. There were no gross findings after AS administration, even when the dosing was extended for 30 days and despite clear in vivo activity of this -Sec-I (Table 1) at similar plasma concentrations in the APP mouse model (data not shown).
After administration of DBZ or BZ, there were significant dose-dependent compound-related histopathologic changes throughout the intestine; after DBZ administration only there was clear dose-dependent splenic marginal zone lymphoid tissue atrophy. The intestinal histopathology comprised apoptosis of small intestinal crypt epithelial cells, goblet cell metaplasia, villous stunting, villous epithelial vacuolation, and intraluminal mucus. The term goblet cell metaplasia is used to describe an intestinal change that comprises goblet cell hyperplasia, metaplasia (based on mucin staining property alteration), and dysplasia. There were no histopathological changes after AS administration.
Apoptosis. The earliest change observed at day 1 was apoptosis of small intestinal crypt cell epithelial cells and large intestinal glands (Fig. 2a). This was confirmed with cleaved caspase 3 immunohistochemistry (Fig. 2c). In animals given vehicle control, there was minimal cleaved caspase-3 positive staining in surface epithelial cells of villus tips and rare positive cells in small intestinal crypts (Fig. 2b). After administration of BZ, numbers of immunopositive cells increased in small intestinal crypt epithelium and colonic glands at days 1 and 2. Increased numbers of apoptotic cells occurred for the longest periodi.e., 45 daysin duodenum and jejunum, and numbers were greater on days 1 and 2 than on subsequent days (data not shown). Increased numbers of apoptotic cells were present in the colon and ileum on days 1 and 2, respectively, but the increase was less marked than that occurring in the duodenum and jejunum.
|
|
|
|
|
We compared the time-course gene list shown in Table 4 to that of one generated from a 5-day end point doseresponse experiment (samples collected on day 5 only) that included 10, 30, and 100 µmol/kg b.i.d. BZ and AS treatments (see Supplementary Material). There was considerable overlap between the gene list from the 30 µmol/kg b.i.d. BZ time course and the BZ doseresponse experiment (102 probe sets of 160 on the list for the BZ time course). Despite a small number of genes dysregulated apparently in common between one of three high-dose AS individual animal samples (also no lesion exhibited) in the 5-day doseresponse experiment and that of the BZ lists, the AS changes were not statistically significant.
Rath1 and adipsin mRNA expression. In vivo administration in rats of a potent -Sec-I was recently reported to be associated with increased mRNA expression of the Hes-1 regulated gene adipsin in the gut, with a concomitant increase in the corresponding protein in fecal extracts (Searfoss et al., 2003
). We attempted to confirm these findings in duodenum and feces matrices. In addition, because Rath1 is intimately linked to the Notch pathway, is regulated by Hes1 and we assayed the production of Rath1 and adipsin transcript in the duodenum. In an attempt to identify their earliest appearance, we also assayed Rath1 and adipsin gene products in fecal extracts and related these findings to the temporal development of the metaplastic lesion. We also sought to confirm whether expression changes of these two genes would mirror the relative Notch-modulating potency of the three test compounds, thereby strengthening the Notch culpability hypothesis. The mRNA expression of both adipsin and Rath1 was increased with BZ treatment to a maximum fold-change of 5.6 at 100 µmoles/kg b.i.d. and 4.7 at 30 µmoles/kg b.i.d., respectively. Rath1 expression fell to 2.3-fold that of vehicle control levels at the highest dose (Fig. 6A). Treatment with the most potent Notch modulator, DBZ, caused a greater increase in Rath1 mRNA than BZ. There was a 6.7-fold increase in the expression of Rath1 at the DBZ low dose and a 9.3-fold increase at the DBZ high dose (Fig. 6B). DBZ caused a 3.5-fold increase of adipsin mRNA compared to controls at the mid-dose, with more modest increases at the low and high doses. The response of Rath1 and adipsin to treatment with the AS was near control values at all doses for Rath1; it was also near control values at the low and mid-doses for adipsin, increasing to 2.6-fold at the high dose (Fig. 6C). This change was attributed to an apparent response in a single animal in the group of three.
Two independent 5-day in vivo time-course studies were conducted using the mid-dose 30 µmol/kg b.i.d. BZ treatment (Fig. 7A and 7B). Experiment A exhibited a temporally distinct response for Rath1 and adipsin. Rath1 mRNA increased 2.5-fold by day 1 and remained sustained throughout the time course, whereas adipsin mRNA did not increase on day 1, increased 1.5-fold by day 2, and increased 3.6-fold by day 3, falling to approximately two-fold that of control values by day 5. The mRNA expression pattern for Rath1 was similar in the two experiments. The magnitude of adipsin response was approximately similar in experiments A and B, but the time of induction differed by a day. In experiment B the increase in adipsin mRNA expression was detected on day 1 but with marked variability within the day 1 treatment group.
Western blotting of anti-Rath1 and anti-adipsin reactive fecal proteins. Whereas the molecular weight of deglycosylated rat adipsin is approximately 25 kDa, glycosylated forms resolve in the range of 3845 kDa (Searfoss et al., 2003). The anti-human adipsin antibody employed in these studies recognized purified human adipsin (25 kDa) (Fig. 8, blot A, lane 1), variably recognized in fecal extracts a number of number of faint bands, including a 25 kDa protein (Fig. 8, blots AB), and avidly and reproducibly labeled a 6 kDa protein. This protein is likely an autolytic breakdown product or a fragment generated by residual protease in the sample. Samples maintained at 4°C for an extended period exhibit greater levels of this protein (data not shown). BZ and DBZ (blot A, lanes 49; blot C, lanes 25) but not AS (blot D, lanes 48) caused increased levels of this protein at doses that caused intestinal goblet cell metaplasia in a dose-dependent manner. A mid-dose of BZ (30 µmol/kg b.i.d.), which caused a moderate degree of metaplasia, increased anti-adipsin reactive protein to a level in the feces that was detectable by day 4 (blot B, days 13, lanes 68; day 4, lanes 45). The increased abundance of this band is sustained through day 5.
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The magnitude of intestinal pathologic changes was comparable between the two potent Notch-inhibitor compounds, with DBZ being the slightly more potent of the two. However, AS did not cause these lesions, despite higher plasma free exposure than the toxic compounds (Table 2), and plasma concentrations exceeding its in vitro IC50 values for APP (data not shown) and Notch cleavage (Table 2). Although it may be considered that the free plasma concentration of AS was a smaller multiple of its in vitro IC50 than the toxic compounds, potentially causing a less complete inhibition in vivo, it should be noted that its unbound AUC (012 h)/SupT1 IC50 (=5.7) for Notch-cleavage inhibition (i.e., normalized to Notch inhibitory potency) was similar to the same ratios for the most toxic compound, DBZ, at the two lowest doses (unbound AUC/SupT1 IC50 = 2.2 and 4.6). Therefore, three options are still possible: (1) that the differences between SupT1 and gut progenitor cell milieu are different enough to yield a differing doseresponse curve for Notch-processing inhibition (e.g., gut being less sensitive), (2) that the concentration reaching the target cells differed from what the free plasma concentration indicates, or (3) the lesion is caused in part by altered processing of other -secretase substrates that are also expressed in the gut (e.g., ERBb4).
Both BZ and DBZ were shown to be substrates of the efflux transporter P-glycoprotein in MDR1-MDCK cells, which is expressed apically in enterocytes and endothelial cells of the bloodbrain barrier (Ambudkar et al., 2003). This glycoprotein decreases intracellular and brain tissue concentrations, therefore it is likely to modulate beneficial effects as well as some toxic effects of
-Sec-I, potentially compromising safety margins (c.f., higher exposure in target organ systems such as the spleen, where P-glycoprotein is not expressed). It is not clear to what extent this interaction can modulate intracellular exposure in globlet or progenitor cells when high local concentrations following i.p. or oral administration could lead to saturation of transport. The extent of efflux was similar for BZ and DBZ in vitro (intermediate); AS was not a substrate of P-glycoprotein. Although differences in P-glycoprotein efflux do not appear to explain differences in susceptibility to gut toxicity, it would be interesting to establish whether mice that are genetically deficient in mdr1a P-glycoprotein would show greater sensitivity.
In a time-course evaluation, using the mid-dose BZ, the severity of metaplastic changes in the gut progressed temporally throughout the small and large intestines, although this effect was less severe in the large intestine. This may be due to the endogenous population of goblet cells being greater in the more distal parts of the gut, thus masking the effect of goblet cell metaplasia. Alternatively, the compounds may be peculiarly effective in causing goblet cell metaplasia of higher parts of the intestinal tract. The lack of metaplasia in the cecum was considered to be due to the presence of typhlitis from day 3. This inflammatory change was considered to be secondary to gut perturbation in the small intestine rather than a consequence of intraperitoneal injection, and this concept is consistent with the fact that no typhlitis was observed in control animals. Intestinal crypt epithelial and glandular epithelial apoptosis was observed on days 15, followed by metaplasia on days 25, and crypt epithelial and glandular epithelial regeneration on days 45. We did not test whether the effects of these compounds were reversible.
Using expression data for separating chemical toxicity from pharmacology is an important goal in toxicology. To do that with this work, one would compare expression profiles of AS samples to BZ samples. Expression profiling revealed significant mRNA expression changes in samples taken from animals dosed with BZ in the time-course study. We do not have AS array time-course data or DBZ 5-day data. However, we did perform array analysis of samples from animals dosed with 10100 µmol/kg AS and BZ for 5 days. Despite the comparison between a single day's readout to that from a time course, quite a few of the changes observed in the time-course study with mid-dose BZ treatment were recapitulated on day 5 in the BZ doseresponse study. By contrast, we saw no statistically significant changes in mRNA expression with the AS samples when compared to controls. Thus, the lesions caused by BZ and the attendant gene expression profiles described here appear to be pharmacology (mechanism-based) related only. This conclusion is based on the correlation of Notch modulatory potency with lesion severity (DBZ > BZ, no response with AS), on the remarkable qualitative similarity in these unusual lesions observed between the different chemistries, and on the fact that there are no other adverse effects to report than those described here. Because no gene expression changes, marker increases, or lesions were attributed to AS we cannot at this time relate by comparison to AS any expression changes with BZ to chemistry or to other nonspecific effects.
Expression changes with BZ treatment included several deregulated panendocrine, hormonal and transcription factors that are a consequence both of altered Notch processing and altered differentiation of duodenal gastrointestinal secretory lineages. Pathological change (altered differentiation) was marked by upregulation of mucin, somatostatin, secretin, and CCK mRNAs by BZ at higher doses on day 5. Many of these changes with BZ were previously reported by Searfoss et al. (2003) and are supported by the immunohistochemical observation reported here of increased intestinal neuroendocrine cells, particularly at days 35 after compound administration. Unfortunately, the predictive value of these day-5 data are limited. When the changes are greatest they may be relatively late; they appear to be a function of lesion development; certainly, they cannot predict outcome or serve as premonitory markers of toxicity. Many toxic mechanisms that result from multicomponent cascades play out over time, rendering as uninformative snapshot views of transcription at single or late time points. So, to identify earlier, more predictive changes, the time-course evaluation was performed and early changes were identified. Coincident with apoptosis in intestinal crypts and glands on day 1, an mRNA signature of early-predictive changes was identified; it preceded the appearance of goblet cell metaplasia and remained altered throughout the 5-day treatment period (several enteroendocrine genes were in fact upregulated at least twofold as early as day 2). Among the changes noted were the marked induction of the Hes-regulated transcriptional activator of gut secretory lineage differentiation Rath1, as well as induction of Delta1 (Notch) ligand, furin, and NeuroD mRNAs in the Notch pathway.
The mechanistic significance of the Rath1, Delta1, and NeuroD findings in their relationship to regulation of intestinal differentiation and goblet metaplasia is underscored in part by the literature on gene deletion of bHLH proteins and Notch pathway constituents including Hes1 (Jensen et al., 2000), neurogenin 3 (Jenny et al., 2002
), NeuroD (Mutoh et al., 1997
; Naya et al., 1997
), and Math1. The balance between positive regulator bHLH proteins and Hes1 will determine both exocrine and endocrine cell differentiation. Hes1 is a basic helix-loop-helix (bHLH) transcriptional repressor that is expressed in most undifferentiated cells; it is essential for their maintenance and is required for precursor pool expansion. It inhibits differentiation of many cell types by repressing the transcription of other factors, including the bHLH transcription factor Math1 (encoded by Atoh1) (Kageyama et al., 2000
).
In contrast to the robust data set for Rath1, we analyzed the mRNA expression of Hes1 in duodenal samples and achieved mixed results. In the 5-day exposure study, real-time PCR analysis revealed consistent decreases in Hes1 mRNA at all doses in samples from animals dosed with BZ and another potent Notch-modulating -Sec-I (which also caused goblet cell metaplasia but is not described here), more variably but downward with DBZ, and not with AS. Like that reported by Searfoss et al. (2003)
, we also observed in a 5-day time-course study a time-dependent and marked reduction in Hes1 mRNA with mid-dose BZ treatment (data not shown). Yet, despite a downward trend, the magnitude of changes was quite variable in a repeated independent in vivo time-course investigation, even as Rath1 and adipsin transcript changes in the identical RNA samples were reproduced across the two time-course trials (Fig. 7). We cannot at this time offer either a technical or a biological explanation for this discrepancy (e.g., Hes1 mRNA half-life or stability sensitivity to minimal procedural changes).
In Hes1-deficient mice the expression of positive bHLH genes such as Delta1 (in an apparent positive feedback response), Neurogenin3, NeuroD, and Atoh1/Math1 are upregulated in the gut, and endocrine cells differentiate prematurely. Neurogenin 3 expression is required further downstream in the secretory lineage and is strictly required for the first steps in enteroendocrine differentiation (Jenny et al., 2002), and NeuroD specifies the terminal differentiation of enteroendocrine precursors to secretin and CCK-synthesizing cell types (Mutoh et al., 1997
; Naya et al., 1997
). Mice deficient in Hes-1 exhibit increased numbers of gastrointestinal secretory lineage goblet cells, Paneth cells, and enteroendocrine cells. The entry of stem cells into the cell cycle is generally thought to result in renewal, differentiation, or apoptosis. Thus, despite the finding that Hes-1 deficiency did not appear to affect precursor pool expansion in the gut, increased apoptosis was observed in intestinal crypts and is consistent with this principle (Jensen et al., 2000
). In contrast to Hes1, loss of Math1 depletes secretory lineage cells (Yang et al., 2001
).
Although rigorous proof is lacking that altered Notch signaling is the primary or sole mechanism producing the intestinal lesions described, the early Rath1, Delta1, and NeuroD upregulation described here provides a compelling case that the Rath1-dependent Notch pathway alteration is the mechanistic link to the observed gut metaplasia. This is further underscored by the report suggesting that Hes1-dependent, Math1-independent cofactors are required for specification of the endocrine lineage in uncommitted progenitors in stomach and pancreas (Yang et al., 2001). Rath1 is not expressed in mouse, human, or rat pancreas, or in stomach tissues, where Hes-1 deficiency is known to cause enteroendocrine architectural changes but where no lesions were detected in this study. It is possible that lesions were not detected in these tissues because the cell turnover is lower and the expression of toxicity requires more exposure time.
The identification of biomarkers of toxicity validated in animal models can mitigate risk of clinical toxicity by permitting careful dose escalation studies where therapeutic margins are expected to be narrow and early withdrawal if clinical toxicity is approached. Biomarkers can also be used in counter-screening of drug project back-ups in the discovery of novel drug series. To these ends, we confirmed that Rath1 mRNA and adipsin mRNA may serve as preclinical premonitory markers and that anti-Rath1reactive and anti-adipsinreactive proteins may serve as early noninvasive toxicity (reporter) biomarkers of intestinal goblet metaplasia. It is presumed that the levels of Rath1 and adipsin in feces are dependent on relative cell turnover, residence time of feces, and expression of these proteins in specific parts of the gut. Which component contributes greatest to the levels of these proteins in feces under metaplastic conditions is not yet known.
Rath1 and adipsin may play important roles in the pathogenesis of the metaplastic phenotype and may potentially be exploited as specific, sensitive, and noninvasive clinical biomarkers of both Notch modulation and early drug-induced goblet cell intestinal metaplasia. The predilection for drug-induced goblet cell metaplasia to affect primarily the gastrointestinal tract, a tissue with high cell turnover, suggests that early upregulation of these proteins may be detectable in the feces of patients with early metaplastic changes. Validation of these potential biomarkers in animal models and man is therefore also warranted.
![]() |
SUPPLEMENTARY DATA |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
1 To whom correspondence should be addressed at Molecular & Investigative Toxicology, Safety Assessment US, CRDL 138, 1800 Concord Pike, AstraZeneca Pharmaceuticals, Wilmington, DE 19850. Fax: (302) 886-2341. E-mail: paul.ciaccio{at}astrazeneca.com.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Baron, M. (2003). An overview of Notch signalling pathway. Cell Dev. Biol. 14, 113119.[CrossRef]
Dovey, H. F., John, V., Anderson, J. P., Chen, L. Z., de Saint Andrieu, P., Fang, Freedman, L. Y., Folmer, S. B., Goldbach, B., Holsztynska, E., et al. (2001) Functional gamma-secretase inhibitors reduce ß-amyloid peptide levels in brain. J. Neurochem. 76, 173181[CrossRef][ISI][Medline]
Holcomb, L. A., Gordon, M. N., Jantzen, P., Hsiao, K., Duff, K., and Morgon, D. (1999) Behavioral changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: Lack of association with amyloid deposits. Behav. Genet. 29, 177185.[CrossRef][ISI][Medline]
Jenny, M., Uhl, C., Roche, C., Duluc, I., Guillermin, V., Guillemot, F., Jensen, J., Kedinger, M. and Gradwohl, G. (2002). Neurogenin3 is differentially required for endocrine cell fate specification in the intestinal and gastric epithelium. EMBO J. 21, 63386347.
Jensen, J., Pedersen, E. E., Galante, P., Hald, J., Heller, R. S., Ishibashi, M., Kageyama, R., Guillemot, F., Serup, P., and Madsen, O. D. (2000). Control of endodermal endocrine development. Nat. Genet. 24, 3644.[CrossRef][ISI][Medline]
Jung, K-M., Tan, S., Landman, N., Petrova, K., Murray, S., Lewis, R., Kim, P. K., Kim, D. S., Ryu, S. H., Chao, M. V., and Kim, T-W. (2003). Regulated intramembrane proteolysis of the p75 neurotrophin receptor modulates its association with the TrkA receptor. J. Biol. Chem. 278, 4216142169.
Kageyama, R., Ohtsuka, T., and Tomita, K. (2000). The bHLH Gene Hes1 regulates differentiation of multiple cell types. Mol. Cell 10, 17.[CrossRef]
LaVoie, M. J., and Selkoe, D. J. (2003). The Notch ligands, Jagged and Delta, are sequentially processed by -secretase and presenilin/
-secretase and release signalling fragments. J. Biol. Chem. 278, 3442734437.
Lee C., Perreault N., Brestelli J., and Kaestner, K. (2002). Neurogenin 3 is essential for the proper specification of gastric enteroendocrine cells and the maintenance of gastric epithelial cell identity. Genes Dev. 16, 14881497.
Marambaud, P., Wen, P. H., Dutt, A., Shioi, A. T., Siman, R., and Robakis, N. K. (2003). A CBP binding transcriptional repressor produced by the PS1/cleavage of N-cadherin is inhibited by PS1 FAD mutations. Cell 114, 635645.[ISI][Medline]
Mutoh, H., Fung, B. P., Naya, F. J., Tsai, M-J., Nishitani, J., and Leiter, A. B. (1997). The basic helixloophelix transcription factor BETA2yNeuroD is expressed in mammalian enteroendocrine cells and activates secretin gene expression. Proc. Natl. Acad. Sci. U.S.A. 94, 35603564.
Naya, F. J., Huang, H. P., Qiu, Y., Mutoh, H., DeMayo, F. J., Leiter, A. B., and Tsai, M. J. (1997). Diabetes, defective pancreatic morphogenesis, and abnormal enteroendocrine differentia tion in BETA2/neuroD-deficient mice. Genes Dev. 11, 23232334.
Okamoto, I., Kawano, Y., Murakami, D., Sasayama, T., Araki, N., Miki, T., Wong, A. J., and Saya, H. (2001). Proteolutic release of CD44 in the intracellular domain and its role in the CD44 signalling pathway. J. Cell. Biol. 155, 755762.
Polli, J. W., Wring, S. A., Humphreys, J. E., Huang, L., Morgan, J. B., Webster, L., and Serabjit-Singh, C. S. (2001). Rational use of in vitro P-glycoprotein assays in drug discovery. J. Pharmacol. Exp. Ther. 299, 620628.
Ray, W., Yao, M., Mumm, J., Schroeter, E. H., Saftig, P., Wolfe, M., Selkoe, D. J., Kopan, R., and Goate, A. M. (1999). Cell surface presenilin-1 participates in the -secretase-like proteolysis of Notch. J. Biol. Chem. 274, 3680136807.
Sasai, Y., Kageyama, R., Tagawa, Y., Shigemoto, R., and Nakanishi, S. (1992). Two mammalian helixloophelix factors structurally related to Drosophila hairy and Enhancer of split. Genes Dev. 6, 26202634.[Abstract]
Searfoss, G. H., Jordan, W. H., Calligaro, D. O., Galbreath, E. J., Schirtzinger, L. M., Berridge, B. R., Gao, H., Higgins, M. A., May, P. C., and Ryan, T. P. (2003). Adipsin: A biomarker of gastrointestinal toxicity mediated by a functional -secretase inhibitor. J. Biol. Chem. 278, 4610746116.
Selkoe, D. J. (2001) Alzheimer's disease: Genes, proteins, and therapy. Physiol. Rev. 81, 741766.
Wong, G. T., Manfra, D., Poulet, F. M., Zhang, Q., Josien, H., Bara, T., Engstrom, L., Pinzon-Ortiz, M., Fine, J. S., Le, H-J. J., Zhang, L., Higgins, G. A., and Parker, E. M. (2004). chronic treatment with the -secretase inhibitor LY-411,575 inhibits ß-amyloid peptide production and alters lymphopoiesis and intestinal cell differentiation. J. Biol. Chem. 279, 1287612882.
Yang, Q., Bermingham, N. A., Finegold, M. J., and Zoghbi, H. Y. (2001). Requirement for Math1 for secretory cell lineage commitment in the mouse intestine. Science 294, 21552158.