Sage Journals: Discover world-class research

Abstract

AZD7969 is a potent inhibitor of glycogen synthase kinase 3 (GSK3β), which is a multifunctional serine/threonine kinase that negatively regulates the Wnt/β-catenin signaling pathway. Treatment of rats and dogs with AZD7969 for periods of up to 4 weeks resulted in a number of changes, the most significant of which was a dose-dependent, and treatment-related, increase in proliferation in a number of tissues that was thought to arise from derepression of Wnt/β-catenin signaling in the stem cell compartment. Phenotypically, this resulted in hyperplasia that either maintained normal tissue architecture in the gastrointestinal tract, liver, kidney, and adrenals or effaced normal tissue architecture within the bones, incisor teeth, and femorotibial joint. In addition to these changes, we noted a treatment-related increase in iron loading in the liver and proximal small intestines. This off-target effect was robust, potent, and occurred in both dogs and rats suggesting that AZD7969 might be a useful tool compound to study iron storage disorders in the laboratory.

Keywords

AZD7969 GSK3b glycogen synthase kinase 3β stem cell safety assessment.

Introduction

GSK3 is a widespread multifunctional serine/threonine kinase that exists in 2 isoforms, α and β (Stambolic and Woodgett 1994), with 98% sequence homology (Kannoji et al. 2008). It was initially identified in rabbit skeletal muscle as an enzyme that phosphorylated and inactivated its target substrate, glycogen synthase (Embi, Rylatt, and Cohen 1980; Frame, Cohen, and Biondi 2001), but it is now known to also regulate a number of other cellular processes such as tau protein (Eldar-Finkelman 2002) and the Wnt/β-catenin pathway (Jope and Johnson 2004).

GSK3β is an unusual serine/threonine kinase in that it is active under resting conditions and inactivated upon cell stimulation. It exists as a multimeric complex of adenomatosis polyposis coli (APC), axin, and casein kinase 1 (CK1)—the “destruction complex”—where one of its primary functions under resting conditions is to phosphorylate β-catenin (CTNNB1 catenin [cadherin-associated protein] β 1), leading to its ubiquitination and proteasomal degradation (Rubinfeld et al. 1996; Chen, Ding, and McCormick 2000). Upon cell stimulation, Wnt glycoprotein ligand activates the canonical β-catenin pathway by signaling through Frizzled and Low-density lipoprotein receptor–related protein 5 (LRP5) and 6 (Cadigan and Liu 2006), which inhibits the destruction complex allowing β-catenin to stabilize in the cytoplasm and translocate to the nucleus. Accumulation of β-catenin in the nucleus leads to binding to T cell factor/lymphocyte enhancing factor (Tcf/Lef) and transactivation of target genes such as c-Myc (He et al. 1998), cyclin D1 (Tetsu and McCormick 1999), and PPARdelta (He et al. 1999).

The role of GSK3β in tumorigenesis is complex and often paradoxical. It was initially thought to act as a tumor suppressor in an analogous manner to that of the “gatekeeper” gene APC where inactivation of APC (present in 80% of sporadic cases) through mutation ultimately results in colorectal cancer (Kinzler and Vogelstein 1996; Wang et al. 2011). To this effect, loss of tumor suppressor function through inactivation of GSK3β has been found in 30% of human cancers (Kang et al. 2008), particularly cancers of epithelial origin, such as skin, breast, and oral cancers (Mishra 2010). However, it is now becoming clear that GSK3β may also act as an oncogene in other types of cancer such as bone (Tang et al. 2012), liver, stomach, colorectal, and kidney tumors (Mishra 2010). The mechanism of this context-dependent role is still poorly understood, but it is thought that in some cell types, GSK3β acts as a positive regulator of NF-κB to promote cell proliferation and survival (Hoeflich et al. 2000; Tang et al. 2012).

In addition to its role in cancer, the Wnt/β-catenin pathway regulates stem cell fate determination in both development and the adult animal. In the adult, Wnt/β-catenin signaling serves to regulate adult tissue renewal in stem cell niches such as the intestinal crypt, hair follicle, and bone marrow (Clevers 2006). Experimental modulation of the Wnt/β-catenin signaling axis in these compartments leads to profound changes in tissue development and renewal such as intestinal crypt loss (Kuhnert et al. 2004), hair loss (Huelsken et al. 2001), and changes in bone mass (Kato et al. 2002). In the nervous system, GSK3 inhibition increases Notch intracellular domain and Gli proteins (Kim and Snider 2011) and has been shown to promote proliferation of neural progenitor cells in the presence of bFGF and EGF (Esfandiari et al. 2012). It is not surprising therefore that GSK3β inhibitors have attracted considerable attention. This has not gone unnoticed by the pharmaceutical industry, as they potentially offer a novel target to treat a number of diseases such as type 2 diabetes, obesity, Alzheimer’s disease, mood disorders, osteoporosis, atherosclerosis, and certain types of cancer (Kannoji et al. 2008). We therefore generated AZD7969 (AR79), a novel small molecule inhibitor of GSK3β which inhibited GSKα/β enzymes with equivalent potency (IC₅₀ 3–5 nM), increased β-catenin stabilization, and stimulated bone mineralization in vitro (Gilmour et al. 2013). Daily oral administration of AZD7969 increased bone mass following 1 to 2 weeks of dosing in rats (Gilmour et al. 2013). These pharmacological responses were commensurate with modulation of GSK3β and similar to effects seen with other small molecule GSK3 inhibitors, for example, AZD2858 (Marsell et al. 2012), lithium chloride (Clement-Lacroix et al. 2005), and LY603281-31-8 (Kulkarni et al. 2006). Additional pharmacological screening for off-target effects with AZD7969 has shown inhibition at several different kinases (Table 1) with equipotent inhibition of HIPK2, activity at PKD1, PIM-3, and DYRK-1A. In a range of G-protein-coupled receptor and enzyme assays, AZD7969 was >40× selective for 5-HT1A/B (Table 1).

Table 1.

Potency determination for AZD7969 in kinase, GPCR, and enzyme assays.

Kinase assay	IC₅₀ (nM)	Fold selectivity versus GSK-3
GSK3	2.5
GSK3	3.2
HIPK2	5	1
PKD1	28	9
PIM-3	39	12
DYRK1A	79	25
80 other assays		>100
GPCR/Enzyme assays
5HT1A/B		40
100 other assays		>100

Note: PKD1 = protein kinase D1.

In this report, we describe the finding resulting from daily oral administration of AZD7969 to rats and dogs in 3-week and 1-month pivotal safety studies. We discuss the range of changes induced by treatment with AZD7969 and describe how they might be reconciled with the known functions of GSK3β inhibition within the stem cell compartment. We also describe an unexpected off-target effect of AZD7969 treatment that resulted in iron accumulation within the liver and proximal small intestine that may act as a useful model to investigate other types of iron storage disease.

Materials and Methods

Test Material

For the in vivo studies, test compound was supplied by the formulation and analytical support group, AstraZeneca, Alderley Park. For the 3-week studies, AZD7969 was formulated as a solution in 0.5% hydroxypropyl methylcellulose/0.1% polysorbate 80/phosphate-buffered saline. For the 4-week studies, AZD7969 was formulated as a solution in 0.5% w/v HPMC and 0.1% w/v polysorbate 80 in 0.01 M phosphate buffer. For all studies, compound or vehicle was administered once daily by oral gavage.

Animals and Husbandry

All the in vivo studies were conducted in accordance with the standards of animal care and ethics described in “Guidance on the Operation of the Animals (Scientific Procedures) Act 1986 (HC321)” issued by the U.K. Home Office. They were conducted so that any clinical expression of toxicity remained within a moderate severity limit as described in guidelines agreed with the U.K. Home Office inspector.

Male and female Wistar Hannover rats (HsdHan:WIST) aged 10 weeks were obtained from Harlan Laboratories UK Ltd (Blackthorn, Bicester, Oxfordshire, OX25 1TP) and were group housed (3–5 per cage) in transparent solid-bottom plastic cages and acclimatized for at least 2 weeks before the first dose. The animals had free access to pelleted RM1 (E) special quality service (SQC) diet and to water from the site drinking water supply.

Beagle dogs, between 12.5 and 15.5 months of age, were obtained from the Dog Breeding Unit, AstraZeneca, Alderley Park. They were single penned and acclimatized for at least 4 weeks before the first dose. Water from the site drinking water supply was freely available and a set quantity of a dry pelleted dog diet (Dog-D3 [E] SQC) was provided daily in the afternoon, access being allowed for a minimum of 2 hr. The dogs were not fed on the day of scheduled necropsy.

Temperature and relative humidity ranges were maintained at 19°C to 23°C and 40% to 70%, respectively, for the rats, whereas for the dogs, temperature was set at 21°C and was controlled within limits of 15°C and 24°C. The animal rooms and exercise areas were illuminated by artificial light from fluorescent tubes on a 12 hr light–dark cycle.

Study Design

Rat 3-week dose range finding (DRF) study

This study comprised a 5-day repeat phase study (data not shown) to assess tolerability and determine the dose levels for the 21-day repeat dose study. For the 3-week repeat dose phase, 4 groups of Wistar Hannover rats (HsdHan:WIST), each consisting of 3 males and 3 females, were treated with AZD7969 at 0, 10, 30, and 70 mg/kg/day. At 70 mg/kg/day, 3 animals (F22, M36, and M35) were euthanized for welfare reasons on days 12, 14, and 15, respectively, due to weight loss. All remaining animals dosed at 70 mg/kg/day were terminated on day 15 (main study animals) or on day 16 (satellite animals) after their last toxicokinetic bleed. As a consequence of findings at 70 mg/kg/day, an additional group of animals was included in this study at a reduced dose of 50 mg/kg/day for 21 days.

Rat 1-month study

Four groups of Wistar Hannover rats (HsdHan:WIST), each consisting of 10 males and 10 females, were treated with AZD7969 at 0, 10, 30, and 50 mg/kg/day for 28 or 29 days (with the exception of 3 males at the high dose which were euthanized early for welfare reasons after either 22 days [M62] or 26 days [M61 and M63] of dosing) due to weight loss. Additional groups (5 males and 5 females) were included to assess recovery after a treatment-free period of 4 weeks. Two recovery females (F96 and F97) were euthanized early for welfare reasons during the recovery period (day 33) due to weight loss.

Dog 3-week study

This study comprised an initial maximum tolerated dose (MTD) phase (data not shown) that was used to establish tolerability and dose levels for the main repeat phase 21-day study. For the 3-week repeat dose phase, 2 groups of Beagle dogs, each consisting of one male and one female were treated with AZD7969 at 5 and 10 mg/kg/day.

Dog 1-month study

Four groups of Beagle dogs, each consisting of 3 males and 3 females, were treated with AZD7969 at 0, 1, 3, and 10 mg/kg/day for 28, 29, or 30 days. An additional high-dose group (3 males and 3 females) was included to assess recovery at 4 weeks after cessation of dosing.

For both dogs and rats, on the scheduled day of necropsy, animals were sacrificed and necropsied according to standard operating procedures. Body weight, organ weights, and macroscopic abnormalities were recorded. Tissues were fixed and processed to H&E slides and pathology data were recorded in Pathdata V6.2d. Group mean organ weight data where appropriate was analyzed using the PathData weighing program, which applied Dunnett’s test based on pooled variance and Dunn’s test. For further investigatory work, sections of liver from both the rat and dog studies were stained with Perls’ Prussian Blue. Additional sections of gastrointestinal (GI) tract and heart from the 1-month rat study and heart from the 1-month dog study were also stained with Perls’ Prussian Blue.

Clinical Signs and Measurements

Clinical signs were recorded once or twice daily, and a physical examination was performed at least once weekly for the 1-month studies. Food consumption was measured over weekly intervals for each cage of rats or daily for dogs. Body weight was recorded during the pretest period and either daily (21-day studies) or twice weekly (1-month studies) until necropsy. For the 1-month studies, an ophthalmoscopic examination was performed prestudy and toward the end of the repeat dose phase and recovery period. Blood samples for toxicokinetics, clinical hematology, and plasma chemistry were collected from the tail vein or terminally from the vena cava (rat studies) or the jugular vein or another peripheral vein (dog studies) both predose and at the end of the dosing period and recovery period. Urine samples were collected once pretest and toward the end of the repeated dose phase and recovery period. Toxicokinetic samples were taken prior to dosing and then at various time points within the first 24-hr postdose at each dose level on day 1 and on the final day of the study as well as 24 hr after the final dose. Finally, for the dog studies, an electrocardiogram was recorded twice prestudy and then prior to dosing and at approximately 1 hr and 3 hr after dosing on day 1 and toward the end of the dosing periods.

Results

Tolerability and Clinical Signs

Rat studies

In the repeat phase of the DRF study, there was no consistent gender difference in the exposure. At steady state, C _max and AUC₍ _0-last) increased more than in proportion to dose in the dose range 10 to 50 mg/kg/day, but at 70 mg/kg/day on day 15, exposure decreased to approximately the exposure observed at 30 mg/kg/day. Mean t _max increased with increasing dose, occurring at 3, 6, and 12 hr after administration at doses of 10, 30, and 50 mg/kg/day, respectively. Treatment with AZD7969 at 70 mg/kg/day was not tolerated, and all remaining animals were sacrificed early on day 15. Treatment at 50 mg/kg/day with AZD7969 resulted in a 32% to 51% decrease in body weight gain, which was accompanied by reduced food intake of up to 33% (Table 2). Termination of the high-dose group and the adverse effect on body weight gain/food consumption were considered to be due to the clinically observed sign of abnormal growth of the incisor teeth (which was confirmed to be dental dysplasia at histological examination). Other clinical signs consisted of increased salivation, ploughing (a sign of abdominal discomfort where the animal pushes through the bedding material head first), and alopecia (primarily ventral abdomen, dorsum, and head).

Table 2.

Body weight changes and food consumption.

	Rat			Dog
	3-week DRF	1-month pivotal	Recovery	3-week DRF	1-month pivotal	Recovery
Male animals—body weight changes	32% ↓ in body weight gain (at 50 mpk)	12% ↓ in relative body weight gain 50 mpk; 8% ↓ in relative body weight gain 30 mpk; 14–20% body weight loss in preterminal mortalities 50 mpk	Yes	No significant changes	No significant changes	n/a
Female animals—body weight changes	51% ↓ in body weight gain (at 50 mpk)	No significant effect; 12–17% body weight loss in preterminal mortalities 50 mpk	Yes	No significant changes	No significant changes	n/a
Male animals—food consumption changes	Up to 33% ↓ food consumption (at 50 mpk)	Up to 35% ↓ food consumption	Yes	Slight but inconsistent reduction	Slight decrease but animals received supplementary feeding	Yes
Female animals—food consumption changes	Up to 33% ↓ food consumption (at 50 mpk)	Up to 29% ↓ food consumption	Yes	Slight reduction		Yes

Note: ↓ decrease; n/a = not applicable. DRF = dose range finding. MPK = mg/kg.

In the 1-month study, there was no consistent gender difference in the exposure and, at steady state, exposure increased more than in proportion to dose. Treatment with AZD7969 resulted in a significant treatment-related decrease in food consumption and body weight in male animals dosed at 30 to 50 mg/kg/day (Table 2). This change had partially recovered by the end of the treatment-free period but was still significantly reduced compared to the concurrent control group. As a result of body weight loss and adverse clinical signs, 3 animals (M61, M62, and M63) treated with 50 mg/kg/day were sacrificed for welfare reasons (14–20% body weight loss) on days 21–26 of study. A further 2 female animals (F96 and F97) treated with 50 mg/kg/day were sacrificed for welfare reasons (12–17% body weight loss) on day 33 of study. The cause of morbidity for all animals was established as incisor tooth dental dysplasia, which was considered to be the dose limiting toxicity. In surviving animals, the most consistent clinical signs were increased salivation, patchy hair loss from around the muzzle, ploughing, and abnormal growth of incisor teeth and red discharge from the eyes. The observation of tooth abnormalities continued into the recovery period.

Dog studies

During the repeat dose phase of the 3-week DRF study, treatment with AZD7969 at 10 mg/kg/day was associated with vomiting once or twice a day. As a result of this, it was not possible to determine the relationship between dose and exposure. Clinical signs consisted of increased salivation in both the male and female dog together with subdued behavior and soft/fluid faeces in the female dog. No significant effects were noted in body weight.

In the 1-month study, at steady state, there was a more than dose proportional increase in exposure with dose. Treatment was well tolerated but induced clinical signs of vomiting, increased salivation, soft/fluid feces, ataxia, trembling, and a slight decrease in motor activity.

Clinical Pathology

Rat studies

In both the 3-week and 1-month study, a number of relatively minor changes were noted during clinical pathology assessment (summarized in Table 3). Hematological changes consisted of decreases in red cell indices (hemoglobin [Hgb], hematocrit [HCT], and red blood cells [RBC]), increases in reticulocytes and red cell distribution width (RDW) and increases in white blood cell (WBC) indices of up to 2-fold. This was accompanied by small changes in clinical chemistry, the most significant of which were slight to moderate increases in urea and creatinine, calcium, and phosphate. In the 1-month study only, minor changes were also seen in urine analysis in male and female animals treated with 50 mg/kg/day characterized by 2 to 2.5 fold increases in urinary N-acetyl-β-D-glucosaminidase (NAG)/creatinine and 3-fold increases in glucose/creatinine values. All the changes noted in the 1-month study showed complete recovery after the end of the 1-month treatment-free period except for the decreases in red cell parameters (Hb, HCT, and RBC), which showed partial recovery by the end of the treatment-free period (summarized in Table 3).

Table 3.

Summary of clinical pathology changes in 3-week and 1-month rat and dog studies.

Hematology	Rat					Dog
Hematology	DRF study		1-month study			DRF study		1-month study
Parameter	Dose (mpk)	Change	Dose (mpk)	Change	Recovery	Dose (mpk)	Change	Dose (mpk)	Change	Recovery
Hgb, HCT, RBC	≥50	10–15% ↓	50 ♀ and ♂	11–17% ↓	No
RDW			50 ♀ and ♂	17–21% ↑	No
Reticulocytes	70	35–40% ↓	50 ♀ and ♂	26–35% ↑	No
WBC	≥50	1.5–2 fold ↑	50 ♀ and ♂	30–40% ↑	Yes

Plasma chemistry	DRF Study		1-month study			DRF Study		1-month study
Parameter	Dose (mpk)	Change	Dose (mpk)	Change	Recovery	Dose (mpk)	Change	Dose (mpk)	Change	Recovery
Increased urea and creatinine	≥50 ♂; 70 ♀	approx. 15–50%	50 ♂	41%/12%	Yes
Increased calcium	≥30	5–15%	≥10 ♀; ≥30–50 ♂	6–14%	Yes			10 ♀	∼10% ↓	Yes
Increased phosphate	≥30	up to 2 fold	30–50 ♀ and ♂	27–42%	Yes

Urine analysis	DRF Study		1-month study			DRF Study		1-month study
Parameter	Dose (mpk)	Change	Dose (mpk)	Change	Recovery	Dose (mpk)	Change	Dose (mpk)	Change	Recovery
NAG/creatinine			50 ♀ and ♂	2–2.5 fold ↑	Yes
Glucose/creatinine			50 ♀	3 fold ↑	Yes
Urine contents						10 mpk	Urinary crystals	10	Urinary crystals	Yes

Note: Blank boxes represent no change; ↑ = increase; ↓ = decrease; mpk = mg/kg/day. DRF = dose range finding; HCT = hematocrit; Hgb = hemoglobin; NAG = N-acetyl-β-D-glucosaminidase; RBC = red blood cells; RDW = red cell distribution width; WBC = white blood cells.

Dog studies

No significant changes were noted in hematological assessment, and only minor changes were noted in all other clinical pathology assessments (summarized in Table 3), the most significant of which were mild increases in total calcium in male dogs noted in the 1-month study and the presence of urinary crystals in the bladder of dogs treated with 10 mg/kg/day in both the 3-week and 1-month studies. All changes noted in the 1-month study showed complete recovery by the end of the treatment-free period.

Necropsy Findings

Rat studies

In both the 3-week and 1-month study, treatment-related macroscopic findings were noted in the lower incisors (malformation/elongated) and small intestines (thickening). During the in-life phase of the study, alopecia or patchy hair loss of the skin was also noted. A number of other less significant changes were noted and are summarized in Table 4. All the changes noted in the 1-month study had recovered at the end of the treatment-free period, except incisor tooth malformation/elongation that persisted throughout the treatment-free period.

Table 4.

Macroscopic changes noted in 3-week and pivotal 1-month rat study.

Necropsy	Rat					Dog
	DRF Study		1-month study			DRF Study
Organ	Necropsy Finding	Dose (mpk)	Necropsy Finding	Dose (mpk)	Recovery	Necropsy finding	Dose (mpk)
Lower incisors	Malformation	70	Elongated	50	No	Reddening	≥5
Gastrointestinal tract	Thickening (duodenum, jejunum, ileum)	70	Thickening (duodenum, jejunum, ileum)	50	Yes	Fluid contents and raised areas	30
Adrenal glands			Discoloration: pale/faded	50 ♂	Yes
Skin	Alopecia	70
Seminal vesicles			Small	50 ♀	Yes
Mesenteric lymph node	Discoloration: red	70	Discoloration: red	50	Yes	Red discoloration	≥ 5
Thymus	Reduced size	70

Note: Blank boxes represent no change; mpk mg/kg/day. DRF = dose range finding.

Dog studies

In the 3-week-DRF study, reddening and fluid contents with raised areas (likely correlates with inflammation) were noted in the gastrointestinal tract as well as red discoloration of the mesenteric lymph node. In the 1-month study, no treatment-related macroscopic changes were noted.

Organ Weights

Rat studies

Significant treatment-related increases in organ weights were noted in the liver, kidneys, adrenal glands, spleen, and heart. Significant decreases were noted in the prostate gland with a nonsignificant trend toward decreased weight in the thymus (Table 5). These changes tended to be dose-related and of greatest magnitude in animals treated with 50 mg/kg/day. The adrenal and prostate gland had fully recovered by the end of the treatment-free period. Partial recovery, indicated by a decrease in the magnitude of the organ weight change, was also noted in the liver, kidneys, spleen, and heart, whereas the thymus showed no evidence of recovery. In this organ, a larger (and statistically significant) decrease in organ weight was noted in male animals at the end of the treatment-free period compared to recovery control animals.

Table 5.

Summary of organ weight changes in 1-month rat and dog studies.

	1-month rat		1-month dog
Organ	Weight change/dose (mpk)	Recovery	Weight change/dose (mpk)	Recovery
Liver	↑ ≥ 30	No ♀	↓ ≥ 3	Yes
Kidneys	↑ 50	No ♀
Adrenal gland	↑ 50 ♀	Yes
Testes			↓ ≥ 1	No
Prostate gland	↓ 50	Yes
Spleen	↑ 50	No
Thymus	Trend toward ↓ (not statistically sig.)	No ♀	↑ ≥ 1 ♀; 1 ♂	No
Heart	↑ ≥ 30 ♂; 50 ♀	No ♀

Note: Blank boxes represent no change; ↑ = increase; ↓ = decrease; mpk = mg/kg/day.

Dog studies

Due to the small group size, organ weights were not analyzed for statistical significance. However, treatment-related changes in absolute and relative (to body weight) organ weights that correlated with microscopic observations were noted in the liver (decreased), thymus (increased), and testes (decreased) in the 1-month study. By the end of the treatment-free period, the liver organ weight change had recovered whereas the testes and thymus were still altered compared to concurrent controls.

Major Histological Findings

A number of treatment and dose-related findings were noted in the dog and rat (summarized in Table 6). Several of these toxicities were common to the dog and rat and could be broadly characterized into proliferative changes (increased mitotic figures or hypercellularity), degenerative changes (single cell necrosis, hypocellularity, and atrophy), and miscellaneous changes (secondary to inanition, stress, or of unknown aetiology).

Table 6.

Summary of histopathology in 3-week and 1-month rat and dog studies.

		Rat			Dog
		3-week DRF	1-month pivotal		3-week DRF	1-month pivotal	Recovery
Organ	Toxicity	Dose (mpk)	Dose (mpk)	Recovery	Dose (mpk)	Dose (mpk)
Femorotibial Joint	Synovial hyperplasia and fibroplasia	≥30	≥10	No
Bone	Hyperostosis	≥30	≥30	No		≥3 ♂; ≥10 ♀	Yes
Bone	Endosteal osteoblast proliferation				30
Incisors	Dental dysplasia	≥30	≥10 ♀; 30 ♂	No
Liver	Increased hepatic glycogen		≥30 ♀	Yes
	Increased mitotic figures	≥50	≥30	Yes
	Increased EMH	≥30
	Hepatocellular degeneration/necrosis	≥50
	PPB Iron accumulation	≥10 ♂; ≥30 ♀	≥ 10	No		≥3	No
	Bile duct hyperplasia				≥5
	Reduced glycogenation				30°	≥3	Yes
Stomach	Fundic foveolar epithelial hypertrophy/hyperplasia	≥30 ♀; ≥50 ♂	≥30 ♀; ≥50 ♂	Yes	≥5^a
Stomach	Increased foveolar goblet cell prominence					10	Yes
Small Intestines	Villous hypertrophy/hyperplasia	≥30^b	≥10^c	Yes	≥5^d
Small Intestines	Increased goblet cell prominence (ileum and/or colon)					≥3	Yes
Kidneys	Increased mitotic figures	≥50	50 ♀	Yes
Kidneys	Diffuse corticomedullary tubular basophilia	50
Adrenal gland	Increased cortical cell vacuolation	≥30 mpk ♀	≥10	Yes
	Increased mitotic figures	≥30 ♀; 70^e ♂	≥30 ♀; ≥50 ♂	Yes
	Cortical hypertrophy/hyperplasia				10
Skin	Follicular single cell necrosis/degeneration	≥50 ♀; 70 ♂	≥30	Yes
Harderian gland	Acinar degeneration and necrosis		≥30 ♂; 50 ♀	No^e ♂	n/a	n/a
Testes	Seminiferous tubular dilatation				30
Testes	Seminiferous tubular degeneration and atrophy				10	≥1	No
Epididymides	Exfoliated spermatogenic giant cells/germinal cells					≥1	Yes
Spleen	Marginal zone hypocellularity	≥50	50 ♀	Yes
Spleen	Increased EMH	≥30	≥ 10	No
Mesenteric lymph node	Increased sinus erythocytes	Increased incidence of uncertain significance	50 ♀		Increased incidence of uncertain significance
Mesenteric lymph node	Hypocellularity		50	Yes	30^f
Thymus	Hypocellularity		50 ♀	Yes	30^f	≥1	Yes

Note: Blank boxes represent no change. n/a = not applicable; DRF = dose range finding; PPB = Perls’ Prussian blue. ^aChanges present in fundus and pylorus with increased mucus @30 mpk. ^bOnly noted in duodenum @30–50 mpk. ^cOnly noted in duodenum @10 mpk but throughout SI @ high-dose levels. ^dMucosal hypertrophy/hyperplasia was noted in both small and large intestines. ^eSingle animal. ^fPremature decedent.

Proliferative Changes

Femorotibial joint, sternum, and incisor teeth

In the femorotibial joint, bones, and incisor teeth, treatment with AZD7969 resulted in a number of changes in both the rat and dog that were proliferative in nature. These changes had generally not fully reversed by the end of the treatment-free period.

In the femorotibial joint of the rat, treatment with AZD7969 resulted in synovial hyperplasia and fibroplasia. These changes were characterized by disorganized proliferation of the synovium and fibroblasts associated with the joint capsule and quadriceps femoris tendon, together with areas of prominent and sometimes mature blood vessels indicating prior angiogenesis (Figure 1A). Inflammation was not a feature of the hyperplastic change. In addition, in a proportion of animals, these changes were also associated with a periosteal reaction characterized by degeneration, necrosis, and hemorrhage.

Figure 1.

(A) Femorotibial Joint H&E higher power view of synovial membrane and quadriceps femoris tendon from a rat treated with AZD7969 for 1 month showing hyperplasia of fibrous and epitheliod cells together with prominent blood vessels (angiogenesis; arrows). (B) H&E Aperio whole slide scan of femorotibial joint from a control (upper panel) and a rat treated with 50-mg/kg/day AZD7969 for 1 month (lower panel) showing increased bone mass within the epiphysis and metaphysis together with an area of fibrodysplasia within the quadriceps femoris tendon (arrows). (C) Sternum 20× from an untreated dog showing inactive looking osteoblasts and (D) H&E high power view of sternum from a dog treated with AZD7969 for 3 weeks at 10 mg/kg/day showing increased numbers of activated osteoblasts (arrows) attached to endosteum of trabecular bone.

In the bones of both the rat and dog, treatment with AZD7969 resulted in a minimal (dog) to marked (rat) increase in bone mass (hyperostosis). In the rat, after 1 month of treatment, this change presented subgrossly as a loss of normal epiphyseal and metaphyseal architecture and was characterized by marked thickening of trabecular bone within the metaphysis and epiphysis of the femorotibial joint and sternum (Figure 1B). In the most severe cases, bone marrow space was significantly effaced. In the dog that showed less severe changes, hyperostosis was characterized by minimal cortical and trabecular hyperostosis within the sternum. In the dog 3-week study, precursor lesions were seen, characterized by minimal endosteal osteoblast proliferation (Figure 1D).

Remarkably similar morphological changes were also noted in the pulp cavity of the rat incisor, which showed changes disorganized proliferation of odontoblasts resulting in dental dysplasia. After 1 month of treatment, this change was characterized by severe changes within the odontogenic epithelium characterized by proliferation with inappropriate inward migration of odontoblasts into the pulp cavity resulting in effacement of the pulp cavity space (Figure 2B). Microscopically, this presented as hypercellularity with ectopic deposition of abnormal eosinophilic material (presumptive dentine) within the pulp cavity accompanied by large numbers of odontoblasts attached to and surrounding spicules of dentine (epithelial “rimming”; Figure 2C). These changes correlated with the macroscopic finding of malformation and elongation of the incisors and the clinical observation of abnormal tooth growth. In rats treated with 50 mg/kg/day, variable amounts of pulp necrosis was also observed. This latter change was thought to be caused by the in-life procedure of teeth trimming, as these changes were not present in the treatment groups in which this procedure had not been performed.

Figure 2.

(A) Incisor tooth H&E high power image from an untreated rat showing normal layers of tooth (A = ameloblasts, S = enamel space, D = dentine, O = odontoblasts, P = pulp cavity). (B) H+E Aperio whole slide scan of upper incisor from a rat treated with AZD7969 for 1 month at 50 mg/kg/day showing abnormal growth and disorganization of odontogenic epithelium and dentine. The pulp cavity is enlarged and hypercellular with prominent areas of eosinophilic hyaline material (arrows) that appears morphologically similar to dentine. (C) H&E showing a higher power view of disorganized odontogenic epithelium with abnormal dentine, odontogenic epithelial hyperplasia and migration of odontoblasts into the pulp cavity (asterisks) with cuffing around areas of abnormal dentine-like material (arrowheads).

Liver and Gastrointestinal System

In the liver and throughout the gastrointestinal system, treatment with AZD7969 in both the rat and dog resulted in hyperplasia that maintained normal tissue architecture. There was no evidence of prior tissue damage or inflammation to indicate regeneration (except in the intestines of one dog killed prematurely). In the liver of the rat in both the 3-week and 1-month studies, proliferation was characterized by a reversible increase in the numbers of hepatocyte mitotic figures, whereas in the dog, increased size and numbers of bile ducts were noted in the 3-week study.

Treatment with AZD7969 also resulted in a nonreversible increase in hepatocellular iron in the liver of both rats and dogs (Table 7). This change was first noticed in the rat studies, which presented as a minimal increase in brown pigment in periportal hepatocytes during routine H&E examination. Staining with Perls’ Prussian Blue (PPB; Figure 3A; Hall et al. 2013) revealed a treatment and dose-related increase in hepatocellular iron which showed no no observed effect level (NOEL). This change was characterized by a predominantly periportal increase in fine pericanalicular intracytoplasmic inclusions that spread toward the centrilobular region with increasing dose of AZD7969 (Hall et al. 2013). In addition, increased iron was noted in the proximal small intestines (duodenum). PPB staining was localized predominantly within lamina propria macrophages toward the tip of the villi, but occasionally, fine intracellular granules were visible within villous enterocytes (Figure 3B). In the dog, H&E-stained sections showed no histologically visible brown pigment within hepatocytes; however, PPB staining revealed a treatment-related increase in hepatocellular iron in the 4-week study that appeared morphologically identical to that seen in the rat studies.

Figure 3.

(A) Liver Prussian Blue 40× high power image from a rat treated with AZD7969 for 1 month at 50 mg/kg/day showing accumulation of pericanalicular Perls’ Prussian blue (PPB) positive iron in the periportal region of the liver. No significant Perls’ staining was noted in controls. (B) Duodenum Perls’ Prussian Blue High power image of duodenal villi from rats treated with AZD7969 for 1 month at 50 mg/kg/day showing absorption of PPB positive staining material (left panel) in gut enterocytes and accumulation of PPB stained material (right panel) in lamina propria. No significant Perls’ staining was noted in control sections.

Table 7.

PPB staining in 3-week and 1-month studies.

Organ	Rat			Dog
	3-week DRF	1-month pivotal		3-week DRF	1-month pivotal
	Dose (mpk)	Dose (mpk)	Recovery	Dose (mpk)	Dose (mpk)	Recovery
Liver	≥10 ♂; ≥30 ♀	≥10	No		≥3 mpk	No
Heart	n/a			n/a
Duodenum	n/a	≥10	No	n/a	n/a
Remainder of SI and LI	n/a			n/a	n/a

Note: Blank boxes represent no change; n/a = not applicable; DRF = dose range finding; PPB = Perls’ Prussian blue; mpk = mg/kg.

In addition to these changes, increased hepatocellular glycogen vacuolation was noted in the 1-month rat study, whereas in the 3-week rat study, increased extramedullary hematopoiesis (EMH) and increased single cell necrosis were noted. The increases in mitotic index, glycogen, EMH, and iron correlated with the increase in organ weight noted at necropsy. The increase in liver organ weight failed to recover after the 4-week treatment-free period.

Changes in the GI tract of both the dog and rat were characterized by diffuse mucosal epithelial hypertrophy/hyperplasia of the fundic and pyloric (dog only) stomach foveolar epithelium and small and large (dog only) intestinal mucosa. Within the intestines, this appeared subgrossly as an enlargement of the intestinal villi/mucosa (Figure 4A) which was characterized by marked intestinal epithelial cell hypertrophy and prominent/enlarged crypts (Figure 4B). In the rat, this change correlated with the macroscopic finding of “thickened” noted throughout the small intestines in both the 3-week and 1-month studies.

Figure 4.

(A) Duodenum H&E low power images of duodenum from a control rat (left panel) and a rat treated with AZD7969 for 1 month at 50 mg/kg/day (right panel) showing diffuse hypertrophy/hyperplasia of the mucosa. Intestinal crypts are prominent and villi are enlarged. Inset—Aperio whole slide scan showing areas where 4× images from control (upper) and treated (lower) images were taken. (B) H&E higher power view of a control (left panel) and treated rat (right panel) showing detail of villi. Enterocytes are hypertrophic but show regular morphology.

Kidneys and Adrenals

In the rat, reversible proliferative changes were noted in the kidney renal cortical tubules and adrenal cortex in both the 3-week and 1-month studies. As was the case in the liver, proliferation consisted of increased numbers of mitotic figures that resulted in hyperplasia with no architectural changes. Again, no prior tissue injury or inflammatory changes were present. In the kidneys from the 3-week DRF study, this change was accompanied by a diffuse corticomedullary tubular basophilia indicative of a more diffuse proliferative change. In the 1-month study, the proliferative changes were accompanied by a correlating increase in organ weight. Histological changes had reversed by the end of the treatment-free period, but organ weights were still significantly increased. Mildly increased urea and creatinine were also noted in the 1-month study, but it is unclear whether this change is related to the increase in cortical tubular proliferation.

In the adrenals, increased mitotic figures (noted at the zona fasiculara/glomerulosa boundary) was accompanied by a diffuse increase in cortical cell vacuolation which correlated in the 1-month study with the gross observation of discoloration (pale/faded). The increase in mitotic figures and cortical cell vacuolation were again accompanied in male animals by a correlating and reversible increase in adrenal weight.

Degenerative Changes

Degenerative changes were noted in the skin and Harderian gland of the rat and the testis of the dog. In the skin, single cell necrosis (apoptosis) of the follicular hair bulb/shaft was seen in both the 3-week and 1-month rat studies. In the most severely affected animals from the 3-week study, abnormal growth and orientation of the hair follicle was observed (Figure 5B) accompanied by degenerative vacuolar changes within the hair bulb and hair shaft with abnormal production of hair (Figure 5C). These changes correlated with alopecia that was noted in various areas of the body at necropsy in the 3-week-DRF study. In addition, patchy hair loss around the muzzle was noted during the in-life phase of the 1-month study which may correlate with this change, but could also be due to rubbing/irritation secondary to dental dysplasia. In the Harderian gland, single cell necrosis was seen in the 1-month study which resulted in acinar degeneration with compensatory regeneration. This change had almost completely recovered by the end of the treatment-free period except for the presence of moderate degeneration and single cell necrosis noted in a single female animal.

Figure 5.

(A) Skin H&E low power and higher power view (inset) from an untreated rat showing normal arrangement and structure of hair follicles. (B) H&E low power image of skin from a rat treated with 70 mg/kg/day of AZD7969 showing degeneration of the hair follicle with vacuolation of the hair bulb/shaft and abnormal orientation of some hair follicles (arrowheads). (C) H&E showing detail of the hair bulb from another rat treated with 70 mg/kg/day of AZD7969 showing single cell necrosis (apoptosis).

In the dog only, treatment with AZD7969 resulted in seminiferous tubular degeneration/atrophy of the testis in both the 3-week and 1-month studies. This change was accompanied in the 1-month study by exfoliation of spermatogenic giant cells and germ cells into the lumen of the epididymides. Seminiferous tubular degeneration correlated with the decrease in organ weight noted at necropsy. Neither seminiferous tubular degeneration nor the organ weight decrease had recovered by the end of the treatment-free period (although exfoliation of germ cells into the epididymidis did recover). Finally, in dogs treated with 30 mg/kg/day in the 3-week study, seminiferous tubular dilatation was also noted.

Miscellaneous Changes

Lymphohematopoietic System

In the rat, increased EMH and hypocellularity of the marginal zone were seen in the spleen in both the 3-week and 1-month studies. Increased EMH correlated with the increase in organ weight noted at necropsy and the changes in RBC indices noted during clinical pathology assessment. Marginal zone hypocellularity showed complete recovery by the end of the treatment-free period whereas increases in EMH and changes in RBC indices, as well as increased organ weight, did not return to normal.

This was accompanied in the mesenteric lymph node by reduced cellularity in a small number of male animals dosed at 50 mg/kg/day and an increased incidence and severity of (agonal) hemorrhage into the medullary sinus. This latter change correlated with the macroscopic observation of “discoloration, red.” In the thymus, decreased cellularity was noted in the rat 1-month pivotal study, which correlated with the decrease in organ weight noted at necropsy in rats treated with 70 mg/kg/day. Histologically, this change was not present at the end of the treatment-free period, although organ weights were still decreased in male animals.

In the dog, hypocellularity was also noted in the thymus in the 1-month study. This change did not correlate with the increase in organ weight noted in this study. At the end of the treatment-free period, a slight reduction in the amount of atrophy was noted in the treated animals compared to concurrent controls.

Discussion

Treatment of rats and dogs with AZD7969 at dosages up to MTD resulted in a variety of treatment-related effects, many of which could be attributed to inhibition of GSK3β and downstream effects on β-catenin/Wnt signaling. Tolerability and clinical signs were, in general, dose-dependent and correlated well with histopathological assessment. Treatment-related changes could loosely be clustered into proliferative changes induced by GSK3β inhibition, and degenerative changes that resulted from a variety of mechanisms, the most likely of which are inanition (at or near MTD), stress, and off-target pharmacology.

In the rat, the most significant clinical effects resulting from treatment with AZD7969 were alopecia and decreases in body weight and food consumption. Alopecia correlated with the histological change of single cell necrosis within the hair follicle. Decreases in body weight and food consumption were considered to be secondary to the abnormal growth of the incisor teeth. This change was confirmed histologically as dental dysplasia and was considered the dose limiting toxicity.

In the dog, the most significant clinical signs were alopecia and soft/fluid feces that are common clinical signs seen in animals dosed at or near the MTD. Consequently, these adverse events are often attributed to locally high concentrations of compound producing an irritant effect on the gastric and intestinal mucosa. However, in this study, a generalized hyperplasia of the gastrointestinal tract was also noted, and so it is interesting to speculate whether this change may have altered intestinal transit and contributed to the clinical observations of soft/fluid feces and vomiting noted in this study. In the rat, the most interesting clinical pathology changes consisted of decreases in red cell parameters, increases in plasma urea and creatinine, and urinary NAG/creatinine and glucose/creatinine ratios. The hematological changes were consistent with a slight regenerative anemia secondary to increased demand due to lowered red cell mass. This change correlated with an increase in EMH noted in the spleen in the 3-week rat DRF study. The cause of this effect is not entirely clear, but a rare cause of anemia in man has been attributed to hyperostosis (Sreehari, Naik, and Eapen 2011). In this clinical syndrome, increases in bone mass result in effacement of bone marrow space and, consequently, a decrease in red cell production leading to a secondary regenerative effort in extrahemopoietic tissues. In our studies, this idea is supported by the observation that hyperostosis was particularly marked in the rat, whereas in the dog, where hyperostosis was only minimal, no significant hematological changes were seen. This change also correlates with the increases in serum calcium and phosphate, which were also more severe in the rat compared to the dog suggesting increased bone turnover (data not shown) in this species.

The changes noted in clinical chemistry may reflect damage to or loss of function within proximal tubular cells (Kavukcu, Soylu, and Turkmen 2002). However, elevations in these parameters occurred in the absence of cortical tubular basophilia or any other histological sign of cortical tubular damage (inflammation, necrosis, etc.) and are therefore unlikely to correlate with, or be the cause of, the increase in mitotic figures noted in the kidneys. This latter change is an unusual toxicologic change that is discussed subsequently.

During the development of AZD7969 and subsequent safety testing in the rodent and dog 3-week and 1-month toxicity studies, we noted a number of unexpected changes that were characterized by gross increases in organ size, increases in organ weight, and diffuse histological increases in proliferation, in normal adult tissues. In the GI tract, liver, kidney, and adrenal glands, proliferation maintained normal tissue architecture, whereas in the bones, incisor teeth, and femorotibial joint, proliferation resulted in reduced function or effacement of normal tissue space with significant changes in tissue architecture. These changes were particularly unusual in that they occurred in the absence of any precipitating cause or prior requirement for proliferation. However, their consistency with the known published effects of Wnt/β-catenin signal perturbation (discussed subsequently) together with the known specificity of AZD7969 leads us to conclude that these changes most probably resulted from pharmacological inhibition of GSK3β in the stem cell compartment.

The Wnt/β-catenin signaling pathway plays a fundamental role in the maintenance of adult stem cells by promoting self-renewal and cell survival. Inhibition of GSK3β by AZD7969 would be expected to increase Wnt/β-catenin signaling through decreased targeting of β-catenin for ubiquitin-mediated proteasomal degradation (Rubinfeld et al. 1996; Chen, Ding, and McCormick 2000). Although changes in β-catenin within the stem cell compartment were not directly assessed in these toxicity studies, previous studies in rat and human stem cells have shown that AZD7969 increased β-catenin stabilization and altered stem cell fate (Gilmour et al. 2013). In addition, other GSK3β inhibitors have been shown to alter stem cell responses in vivo (Gambardella et al. 2011), suggesting that AZD7969 may affect stem cell function via GSK3β inhibition in these preclinical toxicity studies. In the gastrointestinal system, treatment with AZD7969 presented grossly as a thickening of the intestinal tract, which reflected a diffuse hypertrophy/hyperplasia of the stomach fundic foveolar epithelium and small intestinal mucosa in both the rat and dog. Histological examination revealed significant hypertrophy of intestinal epithelial cells and an increased prominence/enlargement of the crypts indicating expansion of the stem cell niche. These data correlate with the known function of GSK3β in the intestines since the Wnt/β-catenin signaling axis is critical for the establishment and maintenance of the crypt stem cell niche. Transgenic mice lacking Tcf4 (transcription factor 4) do not develop intestinal crypts (Korinek et al. 1998; Clevers 2006) whereas conditional expression of the Wnt signaling inhibitor, Dkk-1 (Dickkopf-related protein 1), in adult mice results in reversible loss of crypts (Kuhnert et al. 2004). Our results are therefore consistent with these effects and suggest that treatment with AZD7969 derepresses stem cell Wnt/β-catenin signaling to promote crypt proliferation. Although these findings are thought to be linked to the effects of AZD7969 on GSK3 and the Wnt/β-catenin signaling pathway, it is possible that GSK3 interacts with other downstream transcription factors such as Notch and DISC (disrupted in schizophrenia 1), as it does in the nervous system to alter stem cell fate (Kim and Snider 2011). In the skeletal system, a generalized increase in bone mass was observed in both the rat and dog. Early changes were characterized by osteoblast hyperplasia and synthesis of trabecular bone. These changes are known to be accompanied by minimal increases in osteoclasts (Kugimiya et al. 2007), which is indicative of increased bone turnover and correlates with the moderate increase in serum phosphate and calcium seen in both the 3-week and 1-month rat studies. Later changes observed in the rat 1-month study consisted of increases in mature bone mass characterized primarily by hyperplasia of osteoblasts and thickening of epiphyseal and metasphyseal bony trabeculae. Again these results are consistent with inhibition of GSK3β in the stem cell niche since canonical Wnt/ββ-catenin signaling is strongly implicated in bone homeostasis (Clement-Lacroix et al. 2005; Kulkarni et al. 2006; Marsell et al. 2012). Pluripotent mesenchymal stem cells within the bone marrow give rise to osteoblastic, chondrogenic, adipogenic, and myogenic lineages; and Wnt/β-catenin signaling is thought to regulate osteoblast lineage commitment (Si et al. 2006). In experiments where Wnt/β-catenin signaling is disrupted due to a defect in LRP5, abnormalities characterized by low bone mass are seen both in man (osteoporosis-pseudoglioma syndrome; Gong et al. 2001) and in mice (abnormal osteoblast proliferation and maturation; Kato et al. 2002). Conversely, experiments in which there are gain-of-function mutations in LRP5 result in bone abnormalities characterized by increased bone mass with increased numbers of active osteoblasts (Little et al. 2002; Hofbauer, Maisch, and Schaefer 2002). Similarly, mice that are heterozygous GSK3β deficient also show increased bone formation (Kugimiya et al. 2007).

The changes noted in the rodent incisor appeared morphologically to be remarkably similar to those noted in the skeleton. Here, disorganization and hyperplasia of the odontogenic epithelium were noted, which resulted in deposition of abnormal dentine-like material within the pulp cavity (morphologically similar to hyperostosis in the bone). Many causes for dental changes in the growing rodent incisor have been documented including angiogenic inhibition (Hall, Westwood, and Wadsworth 2006; Patyna et al. 2008; Fletcher et al. 2010) and antitumor agents that target proliferating cells (Satoh et al. 2001). However, the changes induced by AZD7969 appeared to be unique and distinct from those induced by other types of compounds since the principal change appeared to be a proliferative rather than a degenerative change of the odontoblastic epithelium. In addition, the similarity of changes to those noted in the skeleton suggested to us that this change again is due to enhanced Wnt/β-catenin signaling in the odontogenic stem cell niche due to pharmacological inhibition of GSK3β by AZD7969. It also seems likely to us that the increased proliferation noted elsewhere in these studies ie femorotibial joint, liver, kidneys, and adrenals are also likely to be mediated via the same mechanism, namely increased Wnt/β-catenin signaling in the stem cell niche thorough inhibition of GSK3β.

Slight decreases in red cell mass were noted in the hematopoietic system, which are consistent with loss of marrow space due to hyperostosis. This mechanism is also consistent with the known signaling effects of Wnt/β-catenin within the marrow compartment. While in vitro data suggest that Wnt/β-catenin expression is important in maintaining hemopoietic stem cell self-renewal (Van Den Berg et al. 1998), in vivo results show that conditional deletion of β-catenin in transgenic mice has no significant impact upon bone marrow progenitor self-renewal (Cobas et al. 2004). Our results are consistent with those of Cobas et al. (2004) and suggest that Wnt/β-catenin signaling in the adult does not play a significant role in adult bone marrow hematopoietic stem cell renewal.

Treatment with AZD7969 resulted in single cell necrosis in the hair follicle, Harderian gland, and dog testis. In the murine testis, constitutive β-catenin/Wnt signaling in Sertoli cells results in infertility due to defective germ cell differentiation and apoptosis (Tanwar et al. 2010), whereas administration of lithium carbonate (a well-known inhibitor of GSK3b; Galli et al. 2012) to rats resulted in degeneration of spermatogenic cells and vacuolization of sertoli cells cytoplasm in the testis (Thakur et al. 2003). Taken together, these reports suggest that the loss of germ cells and generalized tubular atrophy noted in the dog testis is due to derepression of β-catenin/Wnt signaling by AZD7969. In the skin, however, these data are more difficult to reconcile with the generally growth promoting effects of β-catenin/Wnt signaling in stem cells, especially given that inhibition of GSK3β by valproic acid has been shown to activate the Wnt/β-catenin pathway and promote hair regrowth in vitro and in vivo in mice (Lee et al. 2012). It is, therefore, likely that this effect is mediated through competing off-target activity of AZD7969. Given the in vitro pharmacology profile for AZD7969, 2 possible candidates might account for this effect, protein kinase D1 (PKD 1) and dual specificity tyrosine-regulated kinase 1 (DYRK1), as both of these kinases are known to be expressed in mouse skin (Ernest Dodd et al. 2005; Rahmani et al. 1998). PKD 1 is expressed in proliferating keratinocytes (and a number of other cell types) and is thought to be pro-proliferative and/or anti-differentiative (Ernest Dodd et al. 2005) whereas DYRK1 has sequence homology to cyclin-dependent protein kinase (cdk) kinases (Rahmani et al. 1998) that may regulate cell proliferation and apoptosis (Yoshida 2008). Inhibition of either of these proteins might potentially disrupt the cell cycle control within the hair follicle resulting in promotion of apoptosis. The roles of GSK3β, PKD 1, and DYRK1 (as well as the other off-target hits) in the Harderian gland are, as far as the authors are aware, unknown.

In the 1-month rat study, treatment with AZD7969 resulted in increased hepatic glycogen vacuolation in male animals treated with 30 and 50 mg/kg/day despite an 8% to 12% loss in relative body weight gain in these animals compared to concurrent controls. This result was therefore considered significant and consistent with the expected effect of GSK3β inhibition in the liver. In the steady state condition, GSK3β is constitutively active and, in the resting cell, represses glycogen synthase (GS) activity, a key enzyme regulating the conversion of glucose to glycogen, resulting in the accumulation and storage of glycogen in the liver (Embi, Rylatt, and Cohen 1980; Frame, Cohen, and Biondi 2001). Derepression of GS is normally achieved through insulin which inactivates GSK3β and promotes dephosphorylation of GS via protein phosphatase 1 (Patel, Doble, and Woodgett 2004). In the liver, GSK3β inhibitors have been shown to promote increased GS and glucose uptake (Lochhead et al. 2001; Nikoulina et al. 2002; Cline et al. 2002). Our results are therefore consistent with these data and support the mechanistic rationale that GSK3β inhibitors might act as a therapeutic target for the treatment of type 2 diabetes (Patel, Doble, and Woodgett 2004).

Treatment with AZD7969 and staining with PPB revealed increased hepatic iron accumulation in both the rat and dog (Hall et al. 2013). This change was accompanied by increased iron deposition (PPB staining) in the duodenal mucosa indicating increased uptake from the diet (primary iron overloading). In order to quantify the amount of iron accumulation, and determine the source of the iron loading in the liver, as well as understand the possible risk of iron deposition in other organs such as the heart, sections of liver, intestines, and heart were cut and stained with PPB. The results of the liver investigation have been reported separately (Hall et al. 2013) and summarized in Table 7. No increase in PPB staining was visible in the heart, distal small intestines, or large intestines. However, significantly increased PPB staining was present in the proximal small intestines (duodenum), indicating that the source of the iron loading in the liver was most likely derived from increased absorption of dietary iron from within the proximal small intestines.

Genetic hemochromatosis in humans is a clinical syndrome that is characterized by iron overloading as a consequence of deregulated absorption of iron at the level of the proximal small intestine (Mclaren et al. 1984). Iron is usually first deposited in the periportal region of the liver and as the disease progresses, it spreads centripetally toward the central vein in the same pattern as that noted in our experiments with increasing exposure to AZD7969. This inborn error of metabolism usually results from an inherited defect in a small number of genes that regulate the master iron regulating hormone of the body—hepcidin (Nemeth and Ganz 2009). Although the long-term toxic consequences of iron overloading characteristic of genetic hemochromatosis in man were not investigated in this study, treatment with AZD7969 did result in a potent and dose-responsive iron loading disorder (Hall et al. 2013) that might prove to be a useful tool compound to study these types of diseases in the laboratory. However, one limitation to this model might be a lack of understanding around the mechanism through which this effect is mediated. We therefore offer 2 possible explanations. Since GSK3β has not been reported to regulate iron metabolism, we suggest that this effect is due to off-target pharmacology. Other internal studies using GSK3β inhibitors from different structural series support this notion (personal observation). AZD7969 structural activity relationships suggest that hepcidin may be an off-target candidate, but we have not pursued this possibility. In addition, HIPK2 (homeodomain-interacting protein kinase 2) may represent an off-target candidate, since it is known that under conditions of genotoxic stress, activation of HIPK2 results in decreased ferritin H transcription (the major intracellular iron storage gene) via activation of ATF1 (activating transcription factor 1; Hailemariam et al. 2010). Inhibition of HIPK2 might therefore lead to increased ferritin H transcription and increased storage of iron bound to ferritin. However, AZD7969 and its metabolic products are not directly genotoxic (data not shown) and therefore it is uncertain whether under these conditions HIPK2 repression would markedly affect iron storage.

Summary

Taken together, we have presented data that show good concordance between the Han Wistar rat and the Beagle dog in both 3-week and 1-month toxicity studies. Our data are also consistent with the known effects of GSK3β inhibition within the adult stem cell compartment resulting in hyperplasia that was either organized and nontissue effacing, or disorganized and tissue effacing. We also show that treatment with AZD7969 increases hepatocyte glycogen accumulation consistent with the expected effects of inhibition of GSK3β in the liver. Finally, we demonstrate that treatment with AZD7969 results in iron overloading in both rats and dogs that might prove to be a useful tool compound for studying other types of iron loading disorders.

Our data also support the mechanistic rationale for use of GSK3β inhibitors to treat bone disorders, such as pathological bone loss secondary to myeloma, and disorders of glucose/glycogen metabolism (type 2 diabetes). GSK3β inhibitors have also been proposed as potential new drugs to treat a number of human cancers. However, our data suggest that inhibition of GSK3β induces proliferative effects in a wide range of normal organs and may therefore promote tumor growth, or the proliferation of precancerous lesions, in a subset of human cancers.

Footnotes

Acknowledgments

The authors would like to acknowledge the original study pathologists Jayne Harris, Jen Barnes, and Russell Westwood. We would also like to acknowledge John Foster for helpful comments in proofreading the manuscript.

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Abbreviations

References

Cadigan

K. M.

Liu

Y. I.

(2006). Wnt signaling: Complexity at the surface. J Cell Sci 119, 395–402.

Chen

R. H.

Ding

W. V.

McCormick

(2000). Wnt signaling to beta-catenin involves two interactive components. Glycogen synthase kinase-3beta inhibition and activation of protein kinase C. J Biol Chem 275, 17894–99.

Clement-Lacroix

Morvan

Roman-Roman

Vayssiere

Belleville

Estrera

Warman

M. L.

Baron

Rawadi

(2005). Lrp5-independent activation of Wnt signaling by lithium chloride increases bone formation and bone mass in mice. Proc Natl Acad Sci USA 102, 17406–11.

Clevers

(2006). Wnt/beta-catenin signaling in development and disease. Cell 127, 469–80.

Cline

G. W.

Johnson

Regittnig

Perret

Tozzo

Xiao

Damico

Shulman

G. I.

(2002). Effects of a novel glycogen synthase kinase-3 inhibitor on insulin-stimulated glucose metabolism in Zucker diabetic fatty (fa/fa) rats. Diabetes 51, 2903–10.

Cobas

Wilson

Ernst

Mancini

S. J.

MacDonald

H. R.

Kemler

Radtke

(2004). Beta-catenin is dispensable for hematopoiesis and lymphopoiesis. J Exp Med 199, 221–9.

Eldar-Finkelman

(2002). Glycogen synthase kinase 3: An emerging therapeutic target. Trends Mol Med 8, 126–32.

Embi

Rylatt

D. B.

Cohen

(1980). Glycogen synthase kinase-3 from rabbit skeletal muscle. Separation from cyclic-AMP-dependent protein kinase and phosphorylase kinase. Eur J Biochem 107, 519–27.

Ernest Dodd

Ristich

V. L.

Ray

Lober

R. M.

Bollag

W. B.

(2005). Regulation of protein kinase D during differentiation and proliferation of primary mouse keratinocytes. J Invest Dermatol 125, 294–306.

10.

Esfandiari

Fathi

Gorabi

Kiani

Nemati

Baharvand

(2012). Glycogen synthase kinase-3 inhibition promotes proliferation and neuronal differentiation of human-induced pluripotent stem cell-derived neural progenitors. Stem Cells Dev 21, 3233–43.

11.

Fletcher

A. M.

Bregman

C. L.

Woicke

Salcedo

T. W.

Zidell

R. H.

Janke

H. E.

Fang

Janusz

W. J.

Schulze

G. E.

Mense

M. G.

(2010). Incisor degeneration in rats induced by vascular endothelial growth factor/fibroblast growth factor receptor tyrosine kinase inhibition. Toxicol Pathol 38, 267–79.

12.

Frame

Cohen

Biondi

R. M.

(2001). A common phosphate binding site explains the unique substrate specificity of GSK3 and its inactivation by phosphorylation. Mol Cell 7, 1321–7.

13.

Galli

Piemontese

Lumetti

Manfredi

Macaluso

G. M.

Passeri

(2012). GSK3b-inhibitor lithium chloride enhances activation of Wnt canonical signaling and osteoblast differentiation on hydrophilic titanium surfaces. Clin Oral Implants Res 24, 921–27.

14.

Gambardella

Nagaraju

C. K.

O'shea

P. J.

Mohanty

S. T.

Kottam

Pilling

Sullivan

Djerbi

Koopmann

Croucher

P. I.

Bellantuono

(2011). Glycogen synthase kinase-3α/β inhibition promotes in vivo amplification of endogenous mesenchymal progenitors with osteogenic and adipogenic potential and their differentiation to the osteogenic lineage. J Bone Miner Res 26, 811–21.

15.

Gilmour

P. S.

O’Shea

P. J.

Fagura

Pilling

J. E.

Sanganee

Wada

Courtney

P. F.

Kavanagh

Hall

P. A.

Escott

K. J.

(2013). Human stem cell osteoblastogenesis mediated by novel glycogen synthase kinase 3 inhibitors induces bone formation and a unique bone turnover biomarker profile in rats. Toxicol Appl Pharmacol 272, 399–407.

16.

Gong

Slee

R. B.

Fukai

Rawadi

Roman-Roman

Reginato

A. M.

Wang

Cundy

Glorieux

F. H.

Lev

Zacharin

Oexle

Marcelino

Suwairi

Heeger

Sabatakos

Apte

Adkins

W. N.

Allgrove

Arslan-Kirchner

Batch

J. A.

Beighton

Black

G. C.

Boles

R. G.

Boon

L. M.

Borrone

Brunner

H. G.

Carle

G. F.

Dallapiccola

De Paepe

Floege

Halfhide

M. L.

Hall

Hennekam

R. C.

Hirose

Jans

Juppner

Kim

C. A.

Keppler-Noreuil

Kohlschuetter

LaCombe

Lambert

Lemyre

Letteboer

Peltonen

Ramesar

R. S.

Romanengo

Somer

Steichen-Gersdorf

Steinmann

Sullivan

Superti-Furga

Swoboda

van den Boogaard

M. J.

Van Hul

Vikkula

Votruba

Zabel

Garcia

Baron

Olsen

B. R.

Warman

M. L.

Osteoporosis-Pseudoglioma Syndrome Collaborative Group. (2001). LDL receptor-related protein 5 (LRP5) affects bone accrual and eye development. Cell 107, 513–23.

17.

Hailemariam

Iwasaki

Huang

B. W.

Sakamoto

Tsuji

(2010). Transcriptional regulation of ferritin and antioxidant genes by HIPK2 under genotoxic stress. J Cell Sci 123, 3863–71.

18.

Hall

A. P.

Davies

Stamp

Clamp

Bigley

(2013). Comparison of computerized image analysis with traditional semiquantitative scoring of Perls' Prussian Blue stained hepatic iron deposition. Toxicol Pathol 41, 992–1000.

19.

Hall

A. P.

Westwood

F. R.

Wadsworth

P. F.

(2006). Review of the effects of anti-angiogenic compounds on the epiphyseal growth plate. Toxicol Pathol 34, 131–47.

20.

T. C.

Chan

T. A.

Vogelstein

Kinzler

K. W.

(1999). PPARdelta is an APC-regulated target of nonsteroidal anti-inflammatory drugs. Cell 99, 335–45.

21.

T. C.

Sparks

A. B.

Rago

Hermeking

Zawel

da Costa

L. T.

Morin

P. J.

Vogelstein

Kinzler

K. W.

(1998). Identification of c-MYC as a target of the APC pathway. Science 281, 1509–12.

22.

Hoeflich

K. P.

Luo

Rubie

E. A.

Tsao

M. S.

Jin

Woodgett

J. R.

(2000). Requirement for glycogen synthase kinase-3beta in cell survival and NF-kappaB activation. Nature 406, 86–90.

23.

Hofbauer

L. C.

Maisch

Schaefer

J. R.

(2002). High bone density due to a mutation in LDL-receptor-related protein 5. N Engl J Med 347, 943,4; author reply 943–4.

24.

Huelsken

Vogel

Erdmann

Cotsarelis

Birchmeier

(2001). beta-Catenin controls hair follicle morphogenesis and stem cell differentiation in the skin. Cell 105, 533–45.

25.

Jope

R. S.

Johnson

G. V. W.

(2004). The glamour and gloom of glycogen synthase kinase-3. Trends Biochem Sci 29, 95–102.

26.

Kang

Wei

Honaker

Yamaguchi

Appella

Hung

M. C.

Piwnica-Worms

(2008). GSK-3 beta targets Cdc25A for ubiquitin-mediated proteolysis, and GSK-3 beta inactivation correlates with Cdc25A overproduction in human cancers. Cancer Cell 13, 36–47.

27.

Kannoji

Phukan

Sudher Babu

Balaji

V. N.

(2008). GSK3beta: A master switch and a promising target. Expert Opin Ther Targets 12, 1443–55.

28.

Kato

Patel

M. S.

Levasseur

Lobov

Chang

B. H.

Glass

D. A.

2nd Hartmann

Hwang

T. H.

Brayton

C. F.

Lang

R. A.

Karsenty

Chan

(2002). Cbfa1-independent decrease in osteoblast proliferation, osteopenia, and persistent embryonic eye vascularization in mice deficient in Lrp5, a Wnt coreceptor. J Cell Biol 157, 303–14.

29.

Kavukcu

Soylu

Turkmen

(2002). The clinical value of urinary N-acetyl-beta-D-glucosaminidase levels in childhood age group. Acta Med Okayama 56, 7–11.

30.

Kim

W. Y.

Snider

W. D.

(2011). Functions of GSK-3 Signaling in development of the nervous system. Front Mol Neurosci 17, 4:44.

31.

Kinzler

K. W.

Vogelstein

(1996). Lessons from hereditary colorectal cancer. Cell 87, 159–70.

32.

Korinek

Barker

Moerer

van Donselaar

Huls

Peters

P. J.

Clevers

(1998). Depletion of epithelial stem-cell compartments in the small intestine of mice lacking Tcf-4. Nat Genet 19, 379–83.

33.

Kugimiya

Kawaguchi

Ohba

Kawamura

Hirata

Chikuda

Azuma

Woodgett

J. R.

Nakamura

Chung

U. I.

(2007). GSK-3beta controls osteogenesis through regulating Runx2 activity. PLoS One 2, e837.

34.

Kuhnert

Davis

C. R.

Wang

H. T.

Chu

Lee

Yuan

Nusse

Kuo

C. J.

(2004). Essential requirement for Wnt signaling in proliferation of adult small intestine and colon revealed by adenoviral expression of Dickkopf-1. Proc Natl Acad Sci U S A 101, 266–71.

35.

Kulkarni

N. H.

Onyia

J. E.

Zeng

Tian

Liu

Halladay

D. L.

Frolik

C. A.

Engler

Wei

Kriauciunas

Martin

T. J.

Sato

Bryant

H. U.

Y. L.

(2006). Orally bioavailable GSK-3alpha/beta dual inhibitor increases markers of cellular differentiation in vitro and bone mass in vivo. J Bone Miner Res 21, 910–20.

36.

Lee

S. H.

Yoon

Shin

S. H.

Zahoor

Kim

H. J.

Park

P. J.

Park

W. S.

Min do

Kim

H. Y.

Choi

K. Y.

(2012). Valproic acid induces hair regeneration in murine model and activates alkaline phosphatase activity in human dermal papilla cells. PLoS One 7, e34152.

37.

Little

R. D.

Carulli

J. P.

Del Mastro

R. G.

Dupuis

Osborne

Folz

Manning

S. P.

Swain

P. M.

Zhao

S. C.

Eustace

Lappe

M. M.

Spitzer

Zweier

Braunschweiger

Benchekroun

Adair

Chee

FitzGerald

M. G.

Tulig

Caruso

Tzellas

Bawa

Franklin

McGuire

Nogues

Gong

Allen

K. M.

Anisowicz

Morales

A. J.

Lomedico

P. T.

Recker

S. M.

Van Eerdewegh

Recker

R. R.

Johnson

M. L.

(2002). A mutation in the LDL receptor-related protein 5 gene results in the autosomal dominant high-bone-mass trait. Am J Hum Genet 70, 11–9.

38.

Lochhead

P. A.

Coghlan

Rice

S. Q. J.

Sutherland

(2001). Inhibition of GSK-3 selectively reduces glucose-6-phosphatase and phosphoenolpyruvate carboxykinase gene expression. Diabetes 50, 937–46.

39.

Marsell

Sisask

Nilsson

Sundgren-Andersson

A. K.

Andersson

Larsson

Nilsson

Ljunggren

Jonsson

K. B.

(2012). GSK-3 inhibition by an orally active small molecule increases bone mass in rats. Bone 50, 619–27.

40.

McLaren

G. D.

Muir

W. A.

Kellermeyer

R. W.

(1983). Iron overload disorders: natural history, pathogenesis, diagnosis, and therapy. Crit Rev Clin Lab Sci. 19, 205–66.

41.

Mishra

(2010). Glycogen synthase kinase 3 beta: Can it be a target for oral cancer. Molecular Cancer 9 JUN 11 2010. Article No.:144.

42.

Nemeth

Ganz

(2009). The role of hepcidin in iron metabolism. Acta Haematol 122, 78–86.

43.

Nikoulina

S. E.

Ciaraldi

T. P.

Mudaliar

Carter

Johnson

Henry

R. R.

(2002). Inhibition of glycogen synthase kinase 3 improves insulin action and glucose metabolism in human skeletal muscle. Diabetes 51, 2190–98.

44.

Patel

Doble

Woodgett

J. R.

(2004). Glycogen synthase kinase-3 in insulin and Wnt signalling: A double-edged sword? Biochem Soc Trans 32, 803–8.

45.

Patyna

Arrigoni

Terron

Kim

T. W.

Heward

J. K.

Vonderfecht

S. L.

Denlinger

Turnquist

S. E.

Evering

(2008). Nonclinical safety evaluation of sunitinib: A potent inhibitor of VEGF, PDGF, KIT, FLT3, and RET receptors. Toxicol Pathol 36, 905–16.

46.

Rahmani

Lopes

Rachidi

Delabar

J. M.

(1998). Expression of the mnb (dyrk) protein in adult and embryonic mouse tissues. Biochem Biophys Res Commun 253, 514–18.

47.

Rubinfeld

Albert

Porfiri

Fiol

Munemitsu

Polakis

(1996). Binding of GSK3beta to the APC-beta-catenin complex and regulation of complex assembly. Science 272, 1023–26.

48.

Satoh

Uesugi

Kawabata

Mori

Fujii

Kashimoto

Kajimura

Furuhama

(2001). Morphological classification of dental lesions induced by various antitumor drugs in mice. Toxicol Pathol 29, 292–9.

49.

Kang

Luu

H. H.

Park

J. K.

Luo

Song

W. X.

Jiang

Luo

Yin

Montag

A. G.

Haydon

R. C.

T. C.

(2006). CCN1/Cyr61 is regulated by the canonical Wnt signal and plays an important role in Wnt3A-induced osteoblast differentiation of mesenchymal stem cells. Mol Cell Biol 26, 2955–64.

50.

Sreehari

Naik

D. R.

Eapen

(2011). Osteopetrosis: A rare cause of anemia. Hematol Rep 3, e1.

51.

Stambolic

Woodgett

J. R.

(1994). Mitogen inactivation of glycogen synthase kinase-3 beta in intact cells via serine 9 phosphorylation. Biochem J 303, 701–4.

52.

Tang

Q. L.

Xie

X. B.

Wang

Chen

Han

A. J.

Zou

C. Y.

Yin

J. Q.

Liu

D. W.

Liang

Zhao

Z. Q.

Yong

B. C.

Zhang

R. H.

Feng

Q. S.

Deng

W. G.

Zhu

X. F.

Zhou

B. P.

Zeng

Y. X.

Shen

J. N.

Kang

(2012). Glycogen synthase kinase-3beta, NF-kappaB signaling, and tumorigenesis of human osteosarcoma. J Natl Cancer Inst 104, 749–63.

53.

Tanwar

P. S.

Kaneko-Tarui

Zhang

Rani

Taketo

M. M.

Teixeira

(2010). Constitutive WNT/beta-catenin signaling in murine Sertoli cells disrupts their differentiation and ability to support spermatogenesis. Biol Reprod 82, 422–32.

54.

Tetsu

McCormick

(1999). Beta-catenin regulates expression of cyclin D1 in colon carcinoma cells. Nature 398, 422–6.

55.

Thakur

S. C.

Thakur

S. S.

Chaube

S. K.

Singh

S. P.

(2003). Subchronic supplementation of lithium carbonate induces reproductive system toxicity in male rat. Reprod Toxicol 17, 683–90.

56.

Van Den Berg

D. J.

Sharma

A. K.

Bruno

Hoffman

(1998). Role of members of the Wnt gene family in human hematopoiesis. Blood 92, 3189–202.

57.

Wang

H. L.

Hart

Fan

Mustafi

Bissonnette

(2011). Upregulation of glycogen synthase kinase 3beta in human colorectal adenocarcinomas correlates with accumulation of CTNNB1. Clin Colorectal Cancer 10, 30–6.

58.

Yoshida

(2008). Role for DYRK family kinases on regulation of apoptosis. Biochem Pharmacol 76, 1389–94.

Preclinical Toxicity of AZD7969

Abstract

Keywords

Introduction

Materials and Methods

Test Material

Animals and Husbandry

Study Design

Rat 3-week dose range finding (DRF) study

Rat 1-month study

Dog 3-week study

Dog 1-month study

Clinical Signs and Measurements

Results

Tolerability and Clinical Signs

Rat studies

Dog studies

Clinical Pathology

Rat studies

Dog studies

Necropsy Findings

Rat studies

Dog studies

Organ Weights

Rat studies

Dog studies

Major Histological Findings

Proliferative Changes

Femorotibial joint, sternum, and incisor teeth

Liver and Gastrointestinal System

Kidneys and Adrenals

Degenerative Changes

Miscellaneous Changes

Lymphohematopoietic System

Discussion

Summary

Footnotes

Acknowledgments

Abbreviations

References