Oral treatment with iododiflunisal delays beta amyloid plaque formation in a transgenic mouse model of Alzheimer Disease: a Longitudinal in vivo molecular imaging study


 Background: Transthyretin (TTR) is a tetrameric, Amyloid-beta (Aβ)-binding protein, which has been shown to reduce Aβ toxicity both in vitro and in vivo. The ability of TTR to interact with Aβ can be enhanced by a series of small molecules that stabilize its tetrameric form. Because of this, TTR stabilizers might act as disease modifying drugs in Alzheimer Disease (AD). In this work, we monitored the therapeutic efficacy of two TTR stabilizers, iododiflunisal (IDIF) and the repurposed drug tolcapone, by longitudinal assessment of Aβ deposition in an animal model of AD using positron emission tomography (PET) with [18F]florbetaben. Methods: Mice (AβPPswe/PS1A246E/TTR+/-; n=21) were divided into 3 groups (n=7 per group): iododiflunisal (IDIF)-treated, tolcapone-treated and non-treated. The treatment, administered in the drinking water at a dose of 100 mg/Kg/day, was started at 5 months of age. The level of Aβ deposition was assessed at ages=5, 9, 11, and 14 months by PET imaging using [18F]florbetaben. Treatment efficacy was determined based on radiotracer uptake in the hippocampus (HIP) and the cortex (CTX) with respect to the cerebellum (CB) and presented as standardized uptake value ratios (SUVr). Immunohistochemical (IHC) analysis was performed at 14 months of age to further support in vivo results. Results: SUVr of [18F]florbetaben in CTX and HIP of non-treated animals progressively increased from age=5 to 11 months and stabilized afterwards. In contrast, [18F]florbetaben uptake in HIP of IDIF-treated animals remained constant between ages=5 and 11 months and significantly increased at 14 months. At age=11 months, IDIF-treated group showed significantly lower SUVr values than those obtained for non-treated animals. In tolcapone-treated group, SUVr progressively increased with time, but at lower rate than in non-treated group. Moderate treatment effect of tolcapone suggests different mechanism of action than IDIF. No significant treatment effect was observed in CTX of IDIF- or tolcapone-treated animals. Results from IHC matched the in vivo data at age=14 months. Conclusions: The TTR stabilizer IDIF shows good Aβ-protective effect. Nevertheless, Aβ levels in treated and control animals reached similar values at the end of the study. Furthermore, differences in efficacy between IDIF and tolcapone, suggest different mechanisms of action.

cause of death among those over the age of 70. Alarmingly, these numbers are increasing and are estimated to reach 50 million dementia patients by 2050, worldwide.
Knowledge gained on AD enabled the development of a variety of mechanism-based therapeutic approaches, which aim at slowing down or stopping the disease progression. Most investigated treatment strategies include: (i) minimizing the amount of Aβ in the brain by reducing its generation or accelerating its clearance from the brain; (ii) minimizing aggregation or post-translational modifications of tau protein; and (iii) inhibiting apolipoprotein E (ApoE) (6). Other neuroprotective strategies involve the use of neurotrophins and target neuroinflammation or oxidative stress. Nevertheless, despite decades of efforts, there is still no cure for AD. The outcome is especially worrying because over the last decade more than 50 drug candidates successfully passed phase II clinical trials but all failed in more advanced phases (7)(8)(9)(10)(11). Currently, there are only 132 agents in clinical trials for the treatment of AD (8) compared to more than 3558 drugs employed in cancer trials (12). The absence of approved diseasemodifying therapy calls for an immediate intervention, by feeding new drug candidates into the currently exhausted AD drug development pipeline of stage I clinical trials.
Low success rate in finding appropriate treatment possibilities for AD could be overcome by the development of disease-modifying therapies (DMTs). One possibility of alleviating pathophysiological stress suggests reducing levels of Aβ and its toxic species by enabling their transport out of the brain through the help of intrinsic proteins.
Research in the past decade revealed that interactions of molecular chaperone proteins with toxic Aβ species minimize their harmful effects on the central nervous system (CNS) (13)(14)(15)(16). Several intrinsic proteins were shown to be capable of modifying the stability/aggregation, circulation and clearance characteristics of Aβ peptides (17).
Another protein that helps transport Aβ peptides across the blood brain barrier (BBB) is Transthyretin (TTR) (25)(26)(27)(28). TTR is a 55 kDa homotetramer (29) present in the serum and cerebrospinal fluid (CSF) and is the main Aβ binding protein in human CSF. It was demonstrated that the stability of the tetrameric form of TTR plays a pivotal role in amyloidogenic properties of the protein (30) and that unstable TTR complexes bind poorly to Aβ peptide (31). Studies on AD patients have shown that TTR has reduced ability to carry it's natural stabilizer, thyroxine (T4) in blood plasma (32) and that the ratio folded/monomeric TTR is reduced in AD patients (33). This indicates that TTR stability affect neuroprotective ability of the tetramer. Indeed, the presence of resveratrol led to deceleration of TTR clearance and restoration of normal concentration levels of TTR in the brain (34), possibly due to favored dimer-dimer interaction, which supports tetrameric form of TTR. The seemingly positive effect of increased stability of TTR tetrameric form on TTR-Aβ interaction opens a new avenue for targeting disease pathogenesis.
With the aim of providing a prioritized list of compounds that help enhance the TTR/Aβ interaction, we started a drug discovery program using a combination of computational modeling, in vitro affinity/selectivity/stability assays and structural studies (35). One of these compounds, iododiflunisal (IDIF; Figure 1) has proven efficient in promoting Aβ clearance from the brain and improving animal's cognitive functions when orally administered to AD transgenic mice (AβPPswe/PS1A246E/TTR+/-) daily for 2 months, starting just before the onset of the disease (36). Another study showed that the formation of TTR-IDIF complex enhances brain penetration of both TTR and IDIF (37).
These results suggest that IDIF stabilizes TTR in vivo and prevents Aβ deposition in the brain. On the other hand, tolcapone ( Figure 1), a selective, potent and reversible nitrocatechol-type inhibitor of the enzyme catechol-O-methyltransferase, originally used for the treatment of Parkinson disease, was shown to stabilize TTR in vitro (38) and is currently being repurposed for the treatment of TTR-related amyloidosis (39). previously used for imaging Aβ plaques in AD patients (40), different transgenic AD mouse models (41)(42)(43)(44)(45)(46) and also in longitudinal therapeutic efficacy studies of AD transgenic mouse models (47), was used for assessing protein aggregates distribution. repurposed drug tolcapone.

Compounds 2.1
IDIF meglumine salt was prepared as previously described (36). In brief, to a solution of Committees.
The mouse model AβPPswe/PS1A246E/TTR+/-(carrying only one copy of the TTR gene), was generated as previously described (36)  PET images were co-registered with a magnetic resonance imaging (MRI) template (M. Mirrione-T2, available in the π-MOD image processing tool) and different brain regions (cortex, hippocampus, cerebellum, whole brain) were automatically delineated. The Statistical significance of differences in between time points (for each treatment) or treatment (at a single time point) was calculated using t-student test analyses. P < 0.05 was considered significant.

PET-CT imaging 3.1
Longitudinal PET-CT imaging using [ 18 F]florbetaben was carried out to determine the Aβ plaque burden at the whole brain level and in selected brain subregions in vivo.
Animals submitted to different treatments were scanned at 5, 9, 11 and 14 months of age.
A steady decrease of [ 18 F]florbetaben uptake over time was observed in the whole brain in all animal groups (Figures 2a and 2d-2g). SUV values decreased from ca. 0.1 at age=5 months to ca. 0.05 at age=14 months. No significant differences among groups were observed at any age. Similar trends were observed for SUV values obtained for CTX ( Figure 2b) and HIP (Figure 2c).  (Figure 3a). Differences between groups at a given age or between ages within each group were not significant. SUVr values in CTX were always below 1, indicating that the radiotracer uptake in CTX is actually lower than the uptake in CB. In contrast, SUVr values determined for HIP showed significant differences between groups (Figure 3b) The effect of the treatment on Aβ deposition was studied by assessing Aβ burden in all animals after the last imaging session (age=14 months) by immunohistochemical (IHC) analyses followed by quantification. IHC did not show any significant differences between treated and non-treated animals at this time point, neither in CTX (Figure 4a) nor in HIP (Figure 4b). In all cases, high plaque density was observed, with significant variability among individuals, as observed in the photomicrographs (Figure 4c). photomicrographs illustrate immunohistochemical analysis of brain Aβ plaques using the 6E10 antibody. Brain slices (left) and selected areas for quantification (right) are shown. Slides within each group correspond to two representative animals.

Discussion
TTR is an important transporter protein for the brain defense against pathophysiological stress caused by Aβ deposition. Previous studies on AD animal models have shown that oral administration of IDIF, a TTR tetramer-stabilizing drug, results in decreased Aβ plaque deposition and ameliorates cognitive status at early stages of the disease (36). To fully evaluate the potential of IDIF to act as a diseasemodifying drug, longitudinal therapeutic efficacy and long-term treatment effect on L166P driven by the neuron-specific Thy-1 promoter) from 8 to 13 months of age. The reasons behind the discrepancy between the two studies are unknown, but the most plausible cause for this phenomenon may stem from the inherent differences between the two animal models. Our hypothesis is that there are physiological factors that contribute to decreased tracer uptake in parallel to disease progression. Such physiological changes have indeed been reported in AD mice models (54).
In evaluation of treatment efficacy, the effects of intrinsic genetic differences between animal models were avoided by presenting the data as SUV of [ 18 F]florbetaben in CTX and HIP with respect to CB (SUVr) (41)(42)(43)45). SUVr values in CTX and HIP of nontreated animals progressively increased with animal age (Figures 3a-3b). Similar to previously reported studies in a different animal model (42), significant differences were only observed in HIP but not in CTX. PET images obtained at 9, 11 and 14 months (Figures 3c-3f) clearly showed an increase in SUVr values in HIP of non-treated animals from 9 to 11 months, while the increase in plaque load from 11 to 14 months was not apparent, according to quantification data.
As for the treatment groups, IDIF-treatment delayed Aβ plaque build-up in HIP until the age=11 months, but could not fight severe plaque accumulation at later stages. IHC analysis at the end point confirmed that there were no significant differences between IDIF-treated and non-treated animal groups ( Figure 4). The absence of differences between non-treated and IDIF-treated animals at the end point is most likely not related to inability of stabilizers to complete their task. There are many possible reasons behind these results, including the occurrence of other disease-related processes that possibly take over the main role at more advanced stages of the disease. Although not proven, factors such as a decrease in fluid intake as a consequence of the phenotype (which has been reported for other AD models (55)) or hindered mobility, may lead to lower drug intake and hence limited therapeutic efficacy.
Importantly, no clear evidence of Aβ protein clearance was observed in tolcaponetreated group. Compared to non-treated animals, tolcapone helped slow down the rate of plaque deposition in HIP, suggesting some protective effects of the treatment. Even though previous report showed that tolcapone and entacapone inhibit Aβ fibrilization in a specific and concentration-dependent manner (56), our results suggest that direct effect of tolcapone was not sufficient to produce significant differences between treated and non-treated animals. Difference in effectiveness, when compared to IDIF, suggests a different mechanism of action. Although in vitro studies have shown that tolcapone stabilizes TTR tetramer, limited chaperoning ability in TTR/Aβ interaction could be the cause of the inability of tolcapone-TTR complex to promote Aβ clearance from the brain.
This hypothesis is supported by our recent work describing the lack of chaperoning capabilities of TTR/Aβ interaction of the orphan drugs Tafamidis and diflunisal in vitro, although they are both TTR stabilizers (57).
The positive IDIF treatment effect observed in HIP was not matched in CTX. It seems that CTX was not affected by either of the treatments. This is not surprising, since some studies show differences in the amount of TTR in different brain regions (58).
Furthermore, our recent PET-imaging study shows that entrance of TTR into the brain after intravenous administration starts at the third ventricles, which suggests that TTR traffic occurs partially via the cerebrospinal fluid-brain barrier (CSFBB) and not only through the blood brain barrier (BBB) (37). Ultimately, this led to lower and delayed TTR presence in peripheral areas of the brain (37), such as CTX, where a significant concentration of TTR could be observed only at 6 hours after administration. Assuming the same transport mechanism of IDIF-and tolcapone-stabilized TTR this delay could be one of the reasons for insufficient influx of the complex into CTX. In turn, this would lead to ineffective Aβ clearance, resulting in the absence of differences between treated and non-treated mice.
Of note, in vivo results show lower increase in SUVr in CTX compared to HIP over time.
Ex vivo IHC staining at 14 months of age did not corroborate these findings. Contrarily, no significant difference in abundance of Aβ was found between CTX and HIP. The differences between in vivo and ex vivo results could be due to different tissue permeability in different brain regions. This would impede uniform radiotracer distribution throughout the brain in vivo and could cause differences in [ 18 F]florbetaben wash-out rates. Investigation of possible reasons for the observed findings is out of the scope of this work and will be addressed in future studies.

Conclusions
In conclusion, this is the first large-scale longitudinal A-PET study of cerebral amyloidosis in a transgenic AD mouse model, treated with small molecules that enhance TTR/A interaction. Our work confirms positive effects of TTR stabilizers IDIF and (to a minor extent) tolcapone on delay and/or slowing down Aβ deposition. Furthermore, this study offers the first evidence of how the ability to stabilize TTR complexes affects the degree of amyloidosis in the brain longitudinally. The results suggest that IDIF behaves as chaperone of the TTR-Aβ interaction and could be used to ameliorate Aβ aggregate-related pathological stress by promoting amyloid plaque clearance from CNS.
Furthermore, the present study provides with the basis for the design of a doseresponse study and translation of this new disease-modifying approach to clinical trials for AD therapy. It also shows a great significance of development of small, T4-like structures for new therapeutic strategies against AD. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request