**6. Potential Monitoring of AD Progression through NADH FLIM**

Cumulative evidence from patients, as well as cellular and animal models, have suggested that analyzing the content of NADH and NADPH may be useful to monitor AD progression and oxidative stress. Accordingly, mass spectrometry analysis of brains from triple-transgenic mice (3xTg-AD) showed that this AD model is associated with lower number of metabolites from NAD(P)+/NAD(P)H-dependent reactions [142]. In coherence with this result, it was reported that the brain cortex of 3xTgAD/Polβ+/− mice (in which DNA damage is further exacerbated) has reduced NAD+/NADH ratios [143]. The underlying cause of decreased NAD+/NADH ratio might be explained by an increase in oxidative stress due to PARP the activation. Accordingly, it is expected that the consumption of NAD+ by PARP rises under high oxidative stress and DNA damage [144]. The potential mechanistic relevance of PARP activation during AD pathogenesis has been partially supported by experiments in cultured hippocampal astrocytes treated with β-amyloid, which further activated PARP, while decreasing NAD(P)H autofluorescence as well as mitochondrial oxygen consumption [145]. Furthermore, exogenous treatment of AD patient-derived fibroblasts with NAD, which not only restores NAD+ levels but also inhibits PARP, decreased oxidative stress manifested as a rise in 8-Hydroxy-2 -deoxyguanosine (DNA oxidative damage) and mitochondrial ROS [143]. This is highly similar to what our group has previously reported by showing that the inhibition of PARP-1 reduces H2O2-induced cell death in MCI and AD lymphocytes [67]. Together, these results suggest that under high oxidative stress conditions manifested during AD, a PARP-mediated decrease in NAD+ content could be sensed by label-free microscopy as a drop in either free/protein-bound NADH or NADPH levels. In support of this possibility, it was determined by FLIM that cultured hippocampal neurons from both 3xTg-AD as well as aged mice have diminished cytoplasmic and mitochondrial concentrations of free NADH, which is the direct source of electrons for the mitochondrial complex I [146]

In a complementary approach that supports the potential relevance of a diagnostic tool based on FLIM, it has been shown that cultured neurons from 3xTg-AD mice manifest an early oxidized redox state and lower GSH defense before macromolecular ROS damage is evident [29]. Strikingly, this oxidative damage was reflected in lower resting levels of NAD(P)H/FAD fluorescence ratio and was fully reversible through treatment with NAM. Interestingly, it has been proposed that NAM, as well as other PARP-1 inhibitors, may be used as a treatment for AD at early stages [103]. In order to further test this therapeutic chance, it would be interesting to analyze the content of NADPH in AD patient-derived ONPs by FLIM during the treatment with PARP-1 inhibitors.

The content of NAD+ and NADH in the aging human brain have been non-invasively evaluated by means of magnetic resonance (MR)-based in vivo NAD assay [129]. In coherence with a progressive loss of mitochondrial activity and lower oxidative stress management during normal aging, an age-dependent decline in the content of NAD, NAD+, and NAD+/NADH ratio coupled to increased levels of NADH was revealed in healthy elderly subjects [129]. Interestingly, the decline in NAD+ levels during human aging has been linked to the development and progression of age-related diseases such as AD [147]. Thus, decreased NAD+ levels associated with aging and neurodegeneration are strikingly compatible with the results observed in AD transgenic mice (described above). The limited information from AD patients in this field, despite the promising results in animal models, stresses the need to improve our knowledge of the disease by using patient-derived cellular models.

We sustain that analyzing AD patient-derived ONPs through NADH FLIM is a valuable approach based on the following arguments. First, oxidative stress is an early feature of

AD which is manifested in the olfactory system as well as in cultured patient-derived ONPs. Accordingly, patient-derived ONPs are cells of neuronal lineage and can be easily cultured and non-invasively isolated, constituting a cost-effective way to obtain significant amounts of biological material. Second, the use of NADH and NADPH autofluorescence enables the non-invasive imaging of biological samples, minimizing the perturbation of normal physiological conditions and in a less time-consuming manner. With this approach, AD-related oxidative stress could be sensed as an increased FAD/NAD(P)H ratio or reduced levels of NADH or NADPH, which sustain the synthesis of cytosolic and mitochondrial antioxidant molecules. For all these measurements, FLIM not only provides the exclusive technology to discriminate between NADH and NADPH autofluorescence, but also enables obtaining a higher discrimination between the cytosolic and mitochondrial contribution [99]. Thus, we consider that analyzing AD patient-derived ONPs through NADH FLIM has a great translational potential.
