SIRT3 activation promotes enteric neurons survival and differentiation

Enteric neuron degeneration has been observed during aging, and in individuals with metabolic dysfunction including obesity and diabetes. Honokiol, a naturally occurring compound, is an activator of Sirtuin-3 (SIRT3) that has antioxidant activity. Its role in modulating enteric neuron-specific neurodegeneration is unknown. We studied the effects of honokiol and its fluorinated analog, hexafluoro-honokiol, on enteric neuronal differentiation and survival. We used a previously established model of mouse primary enteric neuronal cells and an enteric neuronal cell line treated with palmitate (PA) and lipopolysaccharide (LPS) to induce mitochondrial dysfunction and enteric neuronal cell death. The effect of honokiol and hexafluoro-honokiol was assessed on neuronal phenotype, fiber density, differentiation, and pyroptosis. Honokiol and hexafluoro-honokiol significantly increased neuronal networks and fiber density in enteric neurons and increased levels of neuronal nitric oxide synthase and Choline acetyltransferase mRNA. Hexafluoro-honokiol and honokiol also significantly increased SIRT3 mRNA levels and suppressed palmitate and LPS-induced neuronal pyroptosis. SIRT3 knock-down prevented the hexafluoro-honokiol mediated suppression of mitochondrial superoxide release. Our data supports a neuroprotective effect of honokiol and its derivative and these could be used as prophylactic or therapeutic agents for treating enteric neurodegeneration and associated motility disorders.

www.nature.com/scientificreports/ that honokiol, an antitumor and antiangiogenic agent 15 , and hexafluoro-honokiol, a novel fluorinated synthetic honokiol analog, exert a protective effect against melanoma in vivo 16 and enhance SIRT3 expression 17 . Honokiol also demonstrated an anti-inflammatory effect in mice with diarrhea induced by enterotoxin by promoting the intestinal barrier and regulating apoptosis of the intestinal epithelium 18 . An advantage of honokiol is that it exerts a direct effect on neuronal cells in the CNS by crossing the blood-brain barrier easily 19 . A combinatorial administration of honokiol alone or with other compounds showed a neuromodulating and neuroprotective effect on the rodent model for diseases related to the CNS and peripheral nervous system 20 . The pleiotropic effects of honokiol in the CNS 14,16,19,21 and gastric mucosa 18,20 suggest a potential role for honokiol in the ENS. The mechanisms for the therapeutic effects of honokiol continue to be defined. Previous models showed that honokiol activates key proteins like SIRT3 and its downstream targets such as peroxisome proliferatoractivated receptor γ coactivator 1α (PGC-1α) to counter the production of ROS that are induced under stress/ injury 22 . The reduced levels of SIRT3 owing to aging, diseases, or poor lifestyle habits 23 as well as knock out of SIRT3 in experimental models of various diseases 23,24 show increased oxidative damage and severe alterations in the acetylation status of mitochondrial proteins, emphasizing the role of SIRT3 in maintaining mitochondrial homeostasis under stress 25 . A key target of SIRT3 is superoxide dismutase-2 (SOD2). In response to oxidative stress, SIRT3-dependent deacetylation of SOD2 is a novel post-translational regulation at lysine 68 (K68) and 122 (K122) which enhances enzymatic activity [26][27][28][29] . Oxidative stress-induced SOD2 expression is believed to be a chief cellular defense mechanism 30 . As SIRT3 activity decreases, SOD2 gets inactivated by acetylation leading to enhanced oxidative stress. In the case of the ENS, the role of SIRT3 in protecting neurons against oxidative stress has not been investigated. In this study, we investigated if honokiol can enhance the survival and differentiation of primary myenteric neurons and in vitro cultured neuronal cells and if the neuroprotective activity of honokiol is mediated by SIRT3.

Materials and methods
Animals. This study followed ARRIVE guidelines on the use of experimental animals. All animal experiments were approved by the Atlanta Veteran Affairs Medical Center Institutional Animal Care and Use Committee and conducted according to recommended guidelines.
Immunofluorescence staining and imaging of myenteric neurons. Neuronal cells were fixed for 20 min at room temperature in 4% paraformaldehyde in phosphate-buffered saline (PBS) and permeabilized at 4 °C for 15 min with 0.3% Triton-X 100. The cells were then blocked with 5% Bovine serum albumin (BSA) in PBS for 1 h and incubated at 4 °C with gentle shaking with a 1:1000 dilution of mouse anti-β-III-tubulin (Tuj-1) antibody (#ab78078, Abcam, Waltham, MA, USA) or 1:400 dilution of rabbit anti-nNOS antibody (#ab76067, Abcam) in blocking buffer. After overnight incubation, the cells were incubated for 1 h at room temperature with a 1:200 dilution of Alexa Fluor 488 conjugated donkey anti-rabbit IgG secondary antibody (#A-21206, Molecular Probes, Eugene, OR, USA) or 1:400 dilution of Alexa Fluor 594 conjugated anti-rabbit IgG secondary antibody (#A-21207, Molecular Probes) before the nuclei were labeled with 4,6-diamidino-2-phenylindole (DAPI, #D3571, Molecular Probes). They were then mounted in Prolong Gold antifade mounting medium (#P36930, Invitrogen, Eugene, OR, USA) and visualized with the aid of an Olympus IX51 microscope (Olympus, Tokyo, Japan) equipped with the cellSens Standard 1.12 imaging software (Olympus) for fluorescence imaging or Nikon A1R (Nikon Instruments, Melville NY, USA) issued with NIS Elements software for confocal imaging. Neuronal cells were scored for fiber length, the number of nNOS-positive cells, and number of DAPI cells using ImageJ software 34 . At least 8-10 fields were examined per group and all experiments have been replicated a minimum of three independent times. Quantitative real-time polymerase chain reaction (qRT-PCR). Total RNA was isolated using the RNeasy Mini kit (Qiagen, Hilden, Germany) and first-strand cDNA was synthesized using the SuperScript VILO siRNA transfection. Neuronal cells grown for at least 24 h in a complete cell culture medium without antibiotics were seeded at recommended densities in tissue culture plates and reverse transfected with 50 nM ON-TARGETplus SMARTpool Mouse SIRT3 siRNA (Horizon Discovery, Cambridge, United Kingdom) or ON-TARGETplus Non-targeting Control siRNA (Horizon Discovery) using Lipofectamine RNAiMAX Reagent (Invitrogen, Carlsbad, CA, USA). The transfection media were replaced after 24 h with a complete medium supplemented with antibiotics and with or without honokiol, hexafluoro-honokiol, BSA-conjugated palmitate, and LPS. The transfected cells for total RNA were isolated for gene expression analyses and MitoSOX assay for superoxide measurement was used after 24 h while proteins were harvested for Western blotting after 48 h.
Assay for mitochondrial ROS. Neuronal cells were cultured in 96 well plates at 80% confluency with and without SIRT3-siRNA transfection, then incubated for 10 min in the dark in a 37 °C incubator loaded with 5 µM MitoSOX (#M36008, Invitrogen) reagent working solution. The cells were washed and counterstained with Hoechst 33,342 (#62,249, Thermo Scientific, Rockfold, IL, USA) in fixation solution 15 min at room temperature. After two washes with PBS the cells imaged on a Olympus IX51 microscope (Olympus).
Statistical analysis. GraphPad Prism 9.0 (https:// www. graph pad. com/; GraphPad Software, La Jolla, CA, USA) software was employed for data analysis by unpaired t-test and 2-way ANOVA as appropriate. The significant difference was considered by P-values observation as follows: P-values of < 0.05 (*), < 0.01 (**), and < 0.001 (***). The mean ± standard error of the mean (SEM) was obtained from at least three separate experiments.

Hexafluoro-honokiol enhances nNOS neurons and protein expression under stress.
We have previously demonstrated that PA and LPS can potentially damage enteric neurons in vitro 41 . We investigated whether hexafluoro-honokiol, an equipotent and easier to extract compound than honokiol and a key enhancer of SIRT3 17 could prevent the neuronal damage induced by PA and along with lipopolysaccharide (LPS) on the primary myenteric neuronal and IM-FEN cells. Primary myenteric neuronal cells were cultured for 5 days and treated with vehicle, hexafluoro-honokiol (10 μM), PA (0.5 mM) with LPS (0.5 ng/ml), and Hexafluoro-honokiol with PA and LPS treated for 24 h, and probed for nNOS, Tuj-1, and DAPI markers. nNOS neurons significantly increased with the treatment of hexafluoro-honokiol. PA and LPS reduced nNOS neurons, although this was not statistically significant with the vehicle, this effect was reversed by hexafluoro-honokiol treatment (hexafluoro-honokiol, 2.34 ± 0.370 fold, P < 0.05; PA + LPS, 0.78 ± 0.21 fold, ns; Hexa + PA + LPS, 2.98 ± 0.67 fold, P < 0.001) compared with vehicle ( Fig. 2A). We then used Western blotting to examine nNOS protein expression in IM-FEN cells. Furthermore, when compared to vehicle or hexafluoro-honokiol, PA and LPS treatment alone increased nNOS protein expression considerably. While nNOS protein expression peaked significantly when hexafluoro-honokiol was combined with PA and LPS treatment compared to other groups (hexafluorohonokiol, 0.880 ± 0.008 fold, P < 0.001; PA + LPS, 1.518 ± 0.02 fold, P < 0.001; Hexa + PA + LPS, 3.272 ± 0.008 fold, P < 0.001) (Fig. 2B). These findings suggest that hexafluoro-honokiol has a positive impact on nNOS neurons and protein expression in the complicity of stress caused by PA with LPS.

Hexafluoro-honokiol prevents cell pyroptosis from palmitate and LPS-induced stress.
To determine if hexafluoro-honokiol mediated reduction in mitochondrial stress leads to suppression of enteric neuronal cell death we assessed neuronal pyroptosis. Pyroptosis is defined by the caspase 1-dependent generation of plasma-membrane pores, which results in pathogenic ion fluxes, cellular lysis, and the release of proinflammatory intracellular substances 46 . We have previously demonstrated PA and LPS induced neuronal pyroptosis 47 . To assess the role of hexafluoro-honokiol in the prevention of cell pyroptosis, we measured the levels of cleaved caspase-1 in IM-FEN cells that are exposed to PA with LPS and treated with hexafluoro-honokiol or vehicle for 24 h. Hexafluoro-honokiol decreased the expression of cleaved caspase-1 significantly and suppressed pyroptosis in palmitate plus LPS induced cell stress (hexafluoro-honokiol, 0.46 ± 0.03-fold, ns; PA + LPS, 0.97 ± 0.04 fold, P < 0.001; Hexa + PA + LPS, 0.51 ± 0.09 fold, P < 0.001) (Fig. 5) representing a protective role of hexafluoro-honokiol in enteric neuronal cells. www.nature.com/scientificreports/

Discussion
In this study, we have shown for the first time that honokiol and its analog, hexafluoro-honokiol, exert beneficial effects on the ENS. Previous studies have reported that honokiol protected against oxidative stress by restoring mitochondrial functions and Na + , K + -ATPase levels 48 , Honokiol also suppressed inflammation in the brains of mice with cerebral ischemia and post injury 20,49 and it activated SIRT3 in mice with cardiac hypertrophy 16,32 . These reports demonstrate that honokiol activates SIRT3 to deacetylate downstream mitochondrial targets and preserve mitochondrial health. Reduced levels of SIRT3 have been associated with aging, disease, or poor lifestyle habits 23 . The knockdown of SIRT3 in experimental models of various diseases 23,24 increased oxidative damage, and caused severe alterations in the acetylation status of mitochondrial proteins, further emphasizing the role of SIRT3 in maintaining mitochondrial homeostasis under stress. In our study, honokiol increased the survival and differentiation of primary myenteric neuronal cells isolated from mice. Also, honokiol notably enhanced the expression of the genes that code for the nNOS (inhibitory neurotransmitter 50 ) and ChAT (excitatory neurotransmitter 50 ). The ENS can be damaged by high-fat diets rich in saturated fatty acids such as PA and LPS endotoxins which are an integral part of gram-negative bacteria and produce pro-inflammatory cytokines 41 . We have previously shown that PA and LPS induce the loss of nitrergic www.nature.com/scientificreports/ neurons in vitro to alter the gut microbiome and affect motility in vivo 41 . Nitrergic neurons produce nitric oxide (NO), an inhibitory neurotransmitter that relaxes gastrointestinal smooth muscle 40 . PA and LPS induced oxidative stress can cause a modest upregulation of nNOS expression. Hexafluoro-honokiol significantly enhances the expression of nNOS neurons and proteins. Our data indicate that treatment with honokiol and hexafluorohonokiol increases SIRT3 in mRNA and protein levels in neuronal cells (IM-FEN). Furthermore, we have also shown that honokiol and hexafluoro-honokiol are dependent on SIRT3 to exert an anti-pyroptotic effect on enteric neurons in the presence of PA and LPS. Further, blocking SIRT3 prevented hexafluoro-honokiol mediated amelioration of mitochondrial superoxide release (Fig. 6). This agrees with previous findings that showed that honokiol activates SIRT3 and suppressed ROS production reducing cellular stress 16 . Thus, SIRT3 is essential for honokiol's activity in the ENS and could be a potential therapeutic target for neuropathies in the ENS. We have previously demonstrated that a high-fat diet causes the loss of nitrergic neurons in the colon and ileum by activating apoptosis via the caspase-3 signaling pathway, leading to ENS neurodegeneration 41 . In this study, we assessed if enteric neurodegeneration caused by PA and LPS could be suppressed by honokiol or hexafluorohonokiol treatment. We observe that honokiol and its analog hexafluoro-honokiol suppressed pyroptosis caused by the neurotoxic insults of PA and LPS and reduced the expression of the pyroptosis marker-cleaved caspase-1 www.nature.com/scientificreports/  www.nature.com/scientificreports/ protein. Honokiol and hexafluoro-honokiol increased the expression of SIRT3 and could potentially reduce mitochondrial oxidative stress caused by PA and LPS thereby preventing pyroptosis in neurons. SIRT3 modulates genotoxic, metabolic, and oxidative stress using different and potentially exclusive molecular mechanisms 51 . Contrary to our data, a study showed that a mice model of inflammation (DNBS-colitis) with SIRT3 deletion as well as SIRT3 inhibition did not increase oxidative damage in their myenteric neurons when compared to the wildtype mice 52 . The authors postulate that damage control in enteric neurons caused by ROS might utilize mechanisms that did not need SIRT3 as opposed to the necessity of SIRT3 in the CNS and other organs in the body to counteract ROS. In this mouse model of inflammation, the authors examined the role of SIRT3 in preventing inflammation-induced oxidative stress. The authors did not find any role for SIRT3 in preventing oxidative stress under these conditions. Our model is based on saturated fatty acids induced neuronal injury and the differences we observe are potentially related to the different stress inducers and the subsequent signaling pathways that are activated including the increase in mitochondrial reactive oxygen species. In our study, we found that SIRT3 expression was critical to protect against PA and LPS-induced injury of enteric neurons. Our results may differ from the previous study due to differences in the models and inducers of inflammation.
Potentially, a lack of SIRT3 is compensated by other sirtuins with similar transcriptional activators, substrates, and mechanistic functions such as SIRT1 53 . Like SIRT3, SIRT1 is an NAD + dependent deacetylase that regulates metabolic processes by acting as an energy sensor, and is widely expressed in the nucleus of enteric neuronal cells 54 , and has shown to prevent intestinal inflammation caused by acute colitis 55,56 . Further studies are needed to examine the role of other sirtuins in the ENS.
We propose that honokiol and hexafluoro-honokiol can increase the antioxidant response of SIRT3 in the ENS during metabolic stress by promoting the transcription of SIRT3 by binding to the antioxidant response element in SIRT3 which releases antioxidants and other neuroprotective chemicals and thus counteracts oxidative damage 57 . Alternatively, honokiol may increase SIRT3 transcription directly via PGC-1α (Peroxisome proliferator-activated receptor gamma coactivator 1-alpha), a master regulator of mitochondrial energy metabolism 32 . It has been shown that PGC-1α works with Nrf2 as a transcriptional coactivator to regulate SIRT3 expression 58 . More studies on how honokiol mediates neuroprotection via Keap1/Nrf2/SIRT3-ARE signaling cascade will be useful to position it as a pharmacological intervention to treat enteroneuropathies.
In summary, our findings demonstrate an important role for honokiol and hexafluoro-honokiol in promoting enteric neuronal survival and differentiation through increased SIRT3 expression. These studies show www.nature.com/scientificreports/ promising effects of using honokiol or its derivative as a potential therapeutic strategy in treating ENS-related neurodegenerative disorders.

Data availability
All data obtained or analyzed throughout this study are included in the manuscript and supplementary files.
Received: 27 April 2022; Accepted: 18 December 2022 Figure 6. Model illustrating a potential mechanism of Honokiol/Hexafluoro mediated activation of SIRT3 and preventing neurodegeneration in enteric neurons. High-fat diets rich in saturated fatty acids like palmitic acid (PA) and endotoxins like lipopolysaccharide (LPS) can cause metabolic stress in the enteric nervous system via induced ROS, which can lead to SIRT3 depletion mitochondrial sirtuin 3 (SIRT3). SIRT3 depletion causes the generation of pyroptotic markers like cleaved caspase-1. Honokiol and hexafluoro treatment may be increasing SIRT3 expression to reduced mitochondrial ROS and suppress pyroptosis.