Shutaro Katsurabayashi a, *, Kohei Oyabu a, Kaori Kubota a, b, Takuya Watanabe a, b, Tomohisa Nagamatsu c, d, Norio Akaike d, e, f, Katsunori Iwasaki a, b
ABSTRACT
Adenosine triphosphate (ATP) is the most vital energy source produced mainly in the mitochondria. Age- related mitochondrial dysfunction is associated with brain diseases. Nicotinamide adenine dinucleotide (NAD+) is an essential cofactor for energy production in mitochondria. Here, we examined how the novel NAD+-assisting substance, 10-ethyl-3-methylpyrimido[4,5-b]quinoline-2,4(3H,10H)-dione (TND1128), modulates the morphological growth of cultured mouse hippocampal neurons. The morphological growth effect of TND1128 was also compared with that of β-nicotinamide mononucleotide (β-NMN). TND1128 induced the branching of axons and dendrites, and increased the number of excitatory syn- apses. This study provides new insight into TND1128 as a mitochondria-stimulating drug for improving brain function.
Keywords:Mitochondria;Dendrite;Axon;Synapse;Development
1.Introduction
The prevalence of neurodegenerative diseases such as Alz- heimer’s disease,Parkinson’s disease,and stroke is rapidly increasing with the aging population worldwide. One of the factors causing these severe neurodegenerative diseases is mitochondrial dysfunction [1]. ATP production occurs mainly in the mitochondria under aerobic conditions. Mitochondria are present in all animal cells and produce the energy required for cell activity via the electron transport system.Mitochondria and mitochondria- associated genes are vulnerable to oxidative stress caused by abnormal electron transport, resulting in reduced energy meta- bolism and cellular degeneration. Mitochondria are particularly susceptible to impairments in the brain, which requires high en- ergy levels. Recently, there has been concern about “brain mito- chondrial dysfunction” due to age-related abnormalities in mitochondrial electron transport. Namely, preventing mitochon- drial dysfunction in the brain may have anti-aging effects, leading to reduced susceptibility to disease.Flavin mononucleotides(FMNs)and flavin adenine di- nucleotides (FADs) are cofactors of flavoenzymes that function as redox catalysts in vivo. Structural changes in flavins have been re- ported in microorganisms [2]. 5-Deazaflavin, in which the N-5 of flavin is replaced with CH, has also been reported to function as a cofactor in several flavin-catalyzed reactions [3]. Furthermore, Nagamatsuetal. [4] have reported that several 5-deazaflavin ana- logues exhibited a strong redox function as a model of NAD(P)+ and antitumor activity by molecular docking to PTK (PDB code: 1t46) [5], and are therefore expected to Hepatitis E effectively ameliorate mito- chondrial dysfunction [6]. In the present study, we investigated the effects of the nicotinamide adenine dinucleotide (NADþ )-like redox function of 10-ethyl-3-methylpyrimido[4,5-b]quinoline- 2,4(3H,10H)-dione (10-ethyl-3-methyl-5-deazaflavin, TND1128) on the arborization of neuronal axons and dendrites, and on the number of excitatory synapses. We also assessed the effect of nicotinamide mononucleotide (β-NMN), which has been widely investigated for its anti-aging effects [7], as a positive control.
2.Materials and methods
2.1. Animals
All animal protocols were approved by the Ethics Committee of Fukuoka University (permit number: 1712128). All experimental protocols were performed according to the relevant guidelines and regulations of Fukuoka University and were performed in accor- dance with the National Institutes of Health guidelines for the Care and Use of Laboratory Animals. Timed-pregnant Jcl:ICR mice (Catalogue ID: Jcl:ICR, CLEA Japan, Inc., Tokyo, Japan) were pur- chased at gestational day 15 from the Kyudo Company (Tosu, Japan). A pregnant mouse was housed individually in a plastic mouse-cage in temperature-controlled rooms (23 ± 2 o C) at our animal facility with a 12-h light-dark cycle. Food (CLEA Rodent Diet, CE-2, CLEA Japan, Inc., Tokyo, Japan) and water were provided ad libitum. All in vivo work was carried out in compliance with the ARRIVE guidelines.
2.2.Autaptic neuron culture
Hippocampal neurons were obtained from ICR mouse brains at postnatal day 0. The dissociated neurons were platedat a density of 1500 cells/cm2 per well onto the micro island astrocytes. The cells were cultured in a humidiied incubator at 37 o C with 5% CO2. The dot-patterned astrocytes were prepared from the cerebral cortex derived from the mass culture in advance. Details of the prepara- tion of culture specimens have been described previously [8,9].
2.3.Immunostaining
Autaptic neurons were immunostained, as described previously [9]. Primary antibodies were used at the following dilutions: anti- microtubule-associated protein 2(MAP2),1:1000(guinea-pig polyclonal, antiserum, Synaptic Systems, Go(€)ttingen, Germany),anti-tau, 1:5000 (mouse monoclonal, antiserum, Cell Signaling Technology, Danvers, MA, USA), anti-vesicular glutamate trans- porter 1 (VGLUT1), 1:2000 (rabbit polyclonal,afinity-puriied,Synaptic Systems, Go(€)ttingen, Germany). Appropriate secondary antibodies conjugated to Alexa Fluor 488 or 594 (Thermo Fisher Scientiic, Waltham, MA, USA) this website were usedat a dilution of 1:400. Cell nuclei were visualized by counterstaining with 40 ,6-diamidino-2- phenylindole (DAPI) contained in the mounting medium (Pro- LongH Gold antifade mounting reagent, Thermo Fisher Scientiic, Waltham, MA, USA).
2.4.Morphological analysis
Autaptic neurons were observed using a confocal microscope (LSM710, Carl Zeiss, Oberkochen, Germany) with a 40 x objective lens (C-Apochromat,numerical aperture1.2). We analyzed dendrite and axon branching using Sholl’s method [10], which is widely used in neuroscience research to quantify the complexity of dendrite and axon branching. This method is a simple method for quantifying neuronal morphology. For quantiication, we used Image J software and the Sholl analysis plug-into draw a minimum circle with a radius of 20 μm centered at the cell body, and then increased the radius in concentric circles by 10 μm. The number of intersections of the dendrites with each circle was determined (Fig. 1A). The number of intersections with the smallest circle in- dicates the number of dendrites growing out of the cell body. An increase in the number of intersections with each concentric circle means more branches of protrusions. VGLUT1 puncta were detec- ted with a size threshold >5 pixels using ImageJ software. To identify the synaptic puncta, a Gaussian blur ilter was applied to remove background noise [11].
2.5.Chemicals
TND1128 was prepared as previously described [12]. All chem- icals except for TND1128 used in this study were purchased from Sigma-Aldrich, USA, dissolved in DMSO, and stored at —20 o C before use.
2.6.Statistical analysis
Data collected in the study were analyzed and plotted with Kaleida Graph 3.6 (Synergy Software, Reading, PA, USA). Values were expressed as the mean ± standard error mean (SEM). For the determination of the statistical signiicance between the two groups, Student’s t-tests were performed. For multiple comparisons between the treated conditions to the control, one-way ANOVA post hoc Tukey’s multiple comparisons tests were performed. For line graphs, two-way ANOVA post hoc Tukey’s multiple compari- sons tests were performed using Prism 6.04 software (GraphPad, La Jolla, CA, USA). The criterion for statistical signiicance was p < 0.05.
Fig. 1. Effects of β-NMN on dendrite branching. A, Scheme of Sholl analysis of a single hippocampal neuron. Concentric circles were drawn at 10 μm intervals around the cell body. B, Effect of β-NMN (1.0 μM) on dendrite crossings at 4 days in vitro. The x-axis indicates the distance from the center of cell body up to 150 μm (black line: Control, n ¼ 20; red line: β-NMN, n ¼ 19). Data were obtained from four independent cultures. C, The total number of crossings after β-NMN application (black bar: Control, n ¼ 20; red bar: β-NMN, n ¼ 19). Data were obtained from the neurons in B. (For interpretation of the references to colour in this igure legend, the reader is referred to the Web version of this article.)
3. Results
Control, p < 0.0001 for 0.3 μM vs. Control, p < 0.0001 for 1.0 μM vs. Control; Fig. 2B). The number of crossings over the smallest circle,
We irst examined the effect of β-NMN on dendrites of hippo- campal neurons. β-NMN (1.0 μM) was added 1 day after plating single neurons. Dendrites were immunostained with MAP2 anti- bodies 3 days after the addition of β-NMN (i.e., on day 4 in vitro). The results showed that β-NMN slightly increased dendrite branching in some regions, but the difference was not signiicant compared with the control (Fig. 1B). Additionally, β-NMN applica- tion did not change the total number of crossings(Control, 62.1 ± 7.3; β-NMN, 72.8 ± 6.8; Fig. 1C). Next, to investigate the pharmacological effect of TND1128 on dendrites, TND1128 (0.1 μM, 0.3 μM, and 1.0 μM) was added for 3 days, 1 day after plating the hippocampal neurons (Fig. 2A). The dendrites were immunostained with MAP2 antibodies, and the number of branches was quantiied. TND1128 signiicantly increased the number of branching dendrites in a concentration-dependent manner (p = 0.0625 for 0.1 μM vs.
i.e., the number of dendrites growing out of the cell body, did not change signiicantly. However, the total number of dendrite cross- ings was signiicantly increased after TND1128 application in a concentration-dependent manner(Control=35.4±3.8, 0.1 μM = 42.4 ± 3.6, 0.3 μM = 63.8 ± 8.8, 1.0 μM = 58.8 ± 5.9; Fig. 2C). We next investigated the effect of TND1128 on mature dendrites. In this experiment, TND1128 was added for 3 days from the 11th day of culture. The dendrites were immunostained with MAP2 antibodies after the addition of TND1128 (i.e., on day 14 in vitro). The results showed that TND1128 slightly increased dendrite branching near the soma in a concentration-dependent manner (p = 0.1981 for 0.1 μM vs. Control, p < 0.0001 for 0.3 μM vs. Control, p = 0.0106 for 1.0 μM vs. Control; Fig. 2D). However, the difference observed was not as large as that obtained when the neurons were immature (see Fig. 2B). Consequently, TND1128
Fig. 2.Effects ofTND1128 on Medicine quality dendrite branching. A, Representative images of dendrites at 4 days in vitro after 3 days ofTND1128 application (Control, 0.1, 0.3, and 1.0 μM). Scale bars, 100 μm. B, Effect of TND1128 on the number of dendrite crossings at 4 days in vitro (black line: Control, n = 24; blue line: 0.1 μM, n = 26; green line: 0.3 μM, n = 19; red line: 1.0 μM, n = 23). Data were obtained from three independent cultures. C, The total number of crossing dendrites after TND1128 application (black bar: Control, n = 24; blue bar: 0.1 μM, n = 26; green bar: 0.3 μM, n = 19; red bar: 1.0 μM, n = 23). Data were obtained from the neurons in B. **p < 0.01 vs. Control. D, Effect of TND1128 on the number of crossing dendrites at 14 days in vitro (black line: Control, n = 44; blue line: 0.1 μM, n = 46; green line: 0.3 μM, n = 46; red line: 1.0 μM, n = 41). Data were obtained from three independent cultures. E, The total number of crossing dendrites after TND1128 application (black bar: Control, n = 44; blue bar: 0.1 μM, n = 46; green bar: 0.3 μM, n = 46; red bar: 1.0 μM, n = 41). Data were obtained from the neurons in D. (For interpretation of the references to colour in this igure legend, the reader is referred to the Web version of this article.)application did not change the total number of crossings (Control = 100.0 ± 7.4, 0.1 mM = 108.3 ± 7.1, 0.3 mM = 124.0 ± 8.4, 1.0 mM = 113.3 ± 6.2; Fig. 2E). These results indicate that TND1128 markedly accelerates dendrite branching in the early develop- mental stage (immature stage). Additionally, 0.3 mM TND1128 was pharmacologically suficient for inducing morphological changes in neurons.
We next examined the effect of TND1128 on axons of hippo- campal neurons. To this end, TND1128 was added to hippocampal neurons 1 day after plating, neuronal axons were immunostained with tau antibody 3 days later, and the number of axon branches was quantiied (Fig. 3A). The axons were analyzed by subtracting MAP2-positive images from tau-positive images, because the tau antibody recognizes some dendrites. The results showed that TND1128 signiicantly increased the number of axon branches in a concentration-dependent manner (p = 0.0048 for 0.1 mM vs. Con- trol, p < 0.0001 for 0.3 mM vs. Control, p < 0.0001 for 1.0 mM vs. Control; Fig. 3B). The total number of crossings was signiicantly increased by TND1128 application in a concentration-dependent manner(Control=45.5±6.3,0.1 mM=61.4±9.1, 0.3 mM = 97.5 ± 13.0, 1.0 mM = 107.9 ± 13.4; Fig. 3C). Furthermore, TND1128 strongly promoted axon extension for 4 days in vitro, which was drastic than the effect on dendrites. This indicated that the axon elongation and crossing became more complex as the culture duration increased in the limited area like the dot- patterned astrocytes. Therefore, Sholl analysis of axons was not physically possible in longer culture periods, such as 14 days in vitro (data not shown). Excessive complexity of axons can lead to double counting of crossings, making it dificult to quantify correctly. Therefore, we did not further examine the TND1128 effect on the developmental changes of axons. Instead, we quantiied the num- ber of excitatory synapses formed on the axon terminals by immunostaining excitatory synapses with VGLUT1 antibodies at 14 days in vitro (Fig. 4A). TND1128 signiicantly increased the number of excitatory synapses in a concentration-dependent manner (Control = 169.1±22.0,0.1mM=223.1±27.3, 0.3 mM = 254.9 ± 27.7, 1.0 mM = 252.7 ± 29.0; Fig. 4B). Because the effect on axons was signiicantat 0.3 mM and 1.0 mM (Fig. 3B and C), it is likely that the number of synapses increased concomitantly with axon branching. These results suggest that TND1128 increases not only dendrites but also axon branching as well as the number of excitatory synapses, and thereby promotes the development of hippocampal neuronal morphology.
4.Discussion
The Sirtuin gene, which has been shown to have anti-aging ef- fects and extend lifespan, is present in most eukaryotes, including humans. Therefore, activating the Sirtuin gene is expected to delay aging and extend the lifespan of organisms. So far, seven mammalian Sirtuin gene subtypes have been discovered. Sirt3, Sirt4, and Sirt5 are mainly expressed in mitochondria, whereas Sirt1 is mainly expressed in the nucleus, but has been reported to be involved in the repair and regeneration of damaged mitochondria [13]. These indings suggest that various Sirtuin genes may be involved in lifespan extension via activation of mitochondrial function [14]. It is believed that Sirtuin genes activate the mito- chondrial signals in cells. Increasing the mitochondria function can prevent dementia, atherosclerosis, and deafness, burn fat, promote cell repair, and remove reactive oxygen species. It has also been reported that Sirt2, a heterochromatin component that stops transcription at telomeres and ribosomal DNA, was identiied as an NAD+-dependent histone deacetylase [15]. Sirt2 protein can restore the binding of DNA to histones weakened by acetylation via
Fig. 3.Effects ofTND1128 on axon branching. A, Representative images of axons at 4 days in vitro after 3 days ofTND1128 application (Control, 0.1, 0.3, and 1.0 mM). Scale bars, 100 mm. B, The number of crossings over concentric circles was quantiied at 4 days in vitro. The analysis was the same as for dendrites (black line: Control, n = 20; blue line: 0.1 mM, n = 19; green line: 0.3 mM, n = 19; red line: 1.0 mM, n = 19). Data were obtained from three independent cultures. C, The total number of crossing axons after TND1128 application (black bar: Control, n = 20; blue bar: 0.1 mM, n = 19; green bar: 0.3 mM, n = 19; red bar: 1.0 mM, n = 19). Data were obtained from the neurons in B. ***p < 0.001 vs. Control. (For interpretation of the references to colour in this igure legend, the reader is referred to the Web version of this article.) Fig. 4. Effects ofTND1128 on the number of excitatory synapses. A, Representative images of MAP2 (green) and VGLUT1 (red) immunostaining at 14 days in vitro after 3 days of TND1128 application (Control, 0.1, 0.3, and 1.0 mM). Scale bars, 100 mm. B, Effect of TND1128 on the number of VGLUT1 puncta (glutamatergic synapses) at 14 days in vitro (Control, n = 45; 0.1 mM, n = 46; 0.3 mM, n = 46; 1.0 mM, n = 40). Data were obtained from three independent cultures. *p < 0.05 vs. Control. (For interpretation of the references to colour in this igure legend, the reader is referred to the Web version of this article.)deacetylation, increase DNA wrapping around histones, and repress transcription. Imai et al. have further proposed that enhanced biosynthesis of NAD+, an activator of NAD+-dependent histone deacetylases, may have anti-aging effects. They have also reported that NMN, an NAD+ synthesis intermediate, promotes NAD+ biosynthesis, which may be involved in the anti-aging effect [16]. Additionally, Sirt6 is thought to act as an important regulator of the extracellular signal-regulated kinases (ERK) 1/2 signaling pathway by mediating the expression of immediate early genes, leading to dendrite development [17]. Therefore, we speculated that the effect of TND1128 may be mediated by the underlying mechanisms of Sirt6. However, full validation of the molecular mechanism requires further pharmacological experiments. Furthermore, Mills et al. [18] have reported that feeding mice a diet supplemented with NMN for an extended period enhanced the production of NAD+in tissues and inhibited the decline in physi- ological functions of aging individuals. These indings are consis- tent with the fact that NAD+ is an essential substance for energy production in the irst step of oxidative phosphorylation in mito- chondria and oxidative reactions in the tricarboxylic acid cycle (TCA) cycle. Although in our study β-NMN application did not have a strong effect, the slight increase in dendritic branching (Fig. 1B) may have resulted from NAD+ production. However, the TND1128 effect was more signiicant than that of β-NMN, suggesting that TND1128 may be a more effective drug than β-NMN for enhancing vitality as well as preventing aging effects (including the preven- tion and improvement of various cerebral neurological disorders).Considering synapses form bridges between neurons, increasing the number of synapses is expected to improve brain function. Further experiments are necessary to elucidate whether the effect of TND1128 can also be obtained in in vitro models of cell damage. Finally,the eficacy of TND1128 as a treatment for neurodegenerative diseases in vivo needs to be conirmed.