Introduction

In tardive dyskinesia, involuntary movements occur1,2. This condition often develops after prolonged use of certain drugs, including antipsychotics1,2. Symptoms frequently appear around the mouth, such as repeated lip pursing, side-to-side tongue movements, mouth protrusion, and teeth clenching1,2,3,4. Involuntary movements may also affect the hands and feet, making walking difficult. Other symptoms include stiffness or difficulty moving the hands and feet, along with poor posture due to muscle stiffness1,2,3,4.

Long-term administration of antipsychotics, including certain neurotransmitter receptor antagonists, is believed to increase the number of neurotransmitter receptors at postsynaptic regions in an effort to receive neurotransmitters like dopamine. This can result in excessive neurotransmitter stimulation, leading to tardive dyskinesia3,4. Therefore, potential solutions to relieve or treat symptoms of tardive dyskinesia include reducing the dosage of medication, switching medications, or using drugs that suppress or mitigate the symptoms3,4. In addition to these options, a key drug target is the ATP-dependent vesicular monoamine transporter-2 (VMAT-2, also known as solute carrier family 18 member A2 [SLC18A2]). VMAT-2 is localized in small vesicles in presynaptic regions5,6,7,8. By inhibiting VMAT-2 activity with chemical agents such as tetrabenazine, neurotransmitter uptake into presynaptic vesicles is reduced, helping to normalize or mitigate the neuronal functions responsible for generating involuntary movements5,6,7,8. VMAT-2 inhibitors are also used to treat various neurological disorders characterized by involuntary movements7,8,9,10,11,12.

While VMAT-2 inhibitors are highly effective in treating neurological diseases after the developmental stage has ended13,14,15,16,17,18, their applicability in treating diseases during the developmental stage remains unclear. First, we described that treatment with the VMAT-2 inhibitors reserpine or tetrabenazine reduces neuronal process elongation in both primary cortical neurons19,20 and the N1E-115 cell line, a model for process elongation21,22. Then, we investigated whether the effects of reserpine or tetrabenazine could be reproduced by knockdown of VMAT-2 with CRISPR/Cas13-compatible guide RNA (gRNA)23,24 specific to VMAT-2. Furthermore, we explored whether hesperetin (also known as vitamin P aglycon), a flavonoid with multiple neuroprotective effects25,26,27,28,29,30,31,32, could restore both the decreased process elongation and lower phosphorylation levels of neuronal differentiation-related Akt kinase29,30,31 induced by targeting VMAT-2.

Materials and methods

Key materials

The key materials, including antibodies, chemicals, and plasmids, are listed in Table 1.

Table 1 Key materials used in this study.
Full size table

Cell line culture and its induction of differentiation

The mouse N1E-115 cell line was purchased from the Japan Health Science Foundation (Tokyo, Japan). Cells were cultured in cell and tissue culture dishes (Nunc brand of Thermo Fisher Scientific, Waltham, MA, USA) in high-glucose Dulbecco’s modified Eagle medium (DMEM; Nacalai Tesque, Kyoto, Japan; Fujifilm, Tokyo, Japan), supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gibco brand of Thermo Fisher Scientific), and 1% Pen-Strep mixture (Nacalai Tesque; Fujifilm) in accordance with the manufacturer’s instructions. To induce neurite-like process elongation, cells were cultured in DMEM with 1% FBS and the penicillin-streptomycin mixture in 5% carbon dioxide at 37 °C for several days, with or without reserpine (15 micromolar concentration), tetrabenazine (60 micromolar concentration), and/or hesperetin (10 micromolar concentration), unless otherwise indicated. Cell morphologies were visualized using a microscope system equipped with i-NTER LENS (Micronet, Saitama, Japan) and i-NTER SHOT 2 software (ver. 12, Micronet). In each image field (each i-NTER field) captured with the microscope, cells with processes longer than three cell body lengths, as measured using Image J software (ver. Java 8; downloaded from the NIH image website, https://imagej.nih.gov/), were considered process-bearing cells21,22. The percentages for each field were statistically represented in graphs. Under these conditions, the percentage of attached cells incorporating trypan blue was estimated to be less than 5% in each experiment.

Isolation and culture of primary cortical neurons

Primary cortical neuronal cells were isolated from the cerebrum of C57BL/6JJcl mice (Clea Japan, Inc., Tokyo, Japan) at embryonic days 16 to 17 and cultured as previously described19,20. Following incubation with 100 units/ml papain (Worthington Biochemical, Lakewood, NJ, USA) in Leibovitz’s L-15 medium (Fujifilm) at 37 °C for 15 min, cells were gently dissociated by pipetting the medium up and down. The dissociated cells were plated at 3 to 5 × 105/cm2 on polylysine (Fujifilm)-coated cell and tissue culture dishes. The culture medium consisted of Neurobasal medium supplemented with 2% B27 (Thermo Fisher Scientific), 1% GlutaMAX (Thermo Fisher Scientific), and 0.1 mg/ml gentamicin solution (Thermo Fisher Scientific). The cells were maintained with 5% carbon dioxide at 37 °C. The medium was replaced with fresh medium one day after isolation, and subsequently, half of the medium was replaced every 2 to 3 days. After maintaining neurons for 7 to 14 days, cultured cortical neuronal cells were detached using a 0.05% trypsin and 0.53 mM EDTA mixed solution (Thermo Fisher Scientific) once. The cells were stored in liquid nitrogen until the experiment. To initiate experiments for observing process elongation, neuronal cells were reattached to cell and tissue culture dishes and allowed to elongate their processes for several days. At day 1 after cell seeding, cells with processes longer than three cell body lengths were considered process-bearing cells. Cells with a single axon longer than 100 micrometer and dendrites shorter than 25 micrometer were counted as cells with an axon and cells with dendrites, respectively. Under these conditions, the percentage of attached cells incorporating trypan blue was estimated to be less than 5% in each experiment.

Plasmid transfection

Cells were transfected with plasmids using the HilyMax transfection kit (Kumamoto, Tokyo, Japan) in accordance with the manufacturer’s instructions. The medium was replaced 4 h after transfection and cells were generally used for 48 h or more post-transfection for biological and biochemical experiments, unless otherwise indicated. In some experiments, multiple sets of transfected cells were observed using fluorescent microscopy. Under these conditions, the percentage of attached cells incorporating trypan blue was estimated to be less than 5% in each experiment.

Cell lysis, polyacrylamide electrophoresis, and Immunoblotting

Cells were lysed in lysis buffer (50 mM HEPES-NaOH, pH 7.5, 150 mM NaCl, 5 mM MgCl2, 1 mM dithiothreitol, 1 mM phenylmethane sulfonylfluoride, 1 µg/ml leupeptin, 1 mM EDTA, 1 mM Na3VO4, 10 mM NaF) containing 0.5% NP-4021,22 and centrifuged for collecting supernatants. These supernatants were used for immunoprecipitation with antibody-conjugated protein G resin (Thermo Fisher Scientific). The samples (20 µg per sample) or the immunoprecipitates were separated on sodium dodecylsulfate polyacrylamide gels (Nacalai Tesque). The electrophoretically separated proteins were transferred to polyvinylidene fluoride membranes (Fujifilm), blocked with Blocking One (Nacalai Tesque), and immunoblotted using primary antibodies, followed by peroxidase enzyme-conjugated secondary antibodies. Peroxidase-reactive bands were detected using X-ray film (Fujifilm) or by adding TMB solution for Western blotting (Nacalai Tesque) on each membrane and captured with an image scanner, CanoScan LiDE400 (Canon, Tokyo, Japan), using CanoScan software (ver. 2024, Canon). The blots shown in the figures are representative of three independent experiments. We performed multiple sets of experiments in immunoblotting studies, quantified the immunoreactive bands, and normalized them to control immunoreactive bands using Image J software. 5. The control group data of immunoblotting was calculated as 1.0.

Statistical analyses

Values are presented as means ± standard deviation from separate experiments. Intergroup comparisons were conducted using either a Student’s or Welch’s t-test in Excel (ver. 2024, Microsoft, Redmond, WA, USA). For multiple comparisons, a one-way analysis of variance (ANOVA) was performed, followed by Tukey’s honestly significant difference (HSD) test, using Excel with StatPlus add-in software (ver. 2021, AnalystSoft, Alexandria, VA, USA). Differences were considered statistically significant at p < 0.05.

Ethics statement

Techniques using genetically modified techniques were performed in accordance with a protocol approved by the Tokyo University of Pharmacy and Life Sciences Gene and Animal Care Committee (Approval Nos. LS28-20 and LSR3-011).

Results

A VMAT-2 inhibitor does not significantly affect morphologies in long term culture following the induction of differentiation

To confirm that treatment with reserpine or tetrabenazine, both VMAT-2 inhibitors, does not affect length of mature neurite-like processes in N1E-115 cells13,14, we treated the cells with each inhibitor after long-term culture following the induction of differentiation. As a result, treatment with reserpine did not significantly affect their process length (Fig. 1A and B). Similar results were obtained with tetrabenazine (Fig. 2A and B), which is consistent with the fact that few side effects are reported when the drug is taken by adults13,14.

Fig. 1
figure 1

Reserpine does not affect process length after long-term culture following the induction of differentiation. (A, B) At 7 days after the induction of differentiation, N1E-115 cells were treated with reserpine or its vehicle and cultured for an additional 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented.

Full size image
Fig. 2
figure 2

Tetrabenazine does not affect process length after long-term culture following the induction of differentiation. (A, B) At 7 days following the induction of differentiation, N1E-115 cells were treated with tetrabenazine or its vehicle and cultured for an additional 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented.

Full size image

Treatment with a VMAT-2 inhibitor or knockdown of VMAT-2 exhibits decreased process elongation

We examined whether inhibition of VMAT-2 affects neurite process elongation in primary cortical neurons. Treating neurons with reserpine inhibited single axon elongation by day 3 following cell seeding on culture dishes. Reserpine treatment also affected the number of dendrites, specifically increasing the number of cells with three or more dendrites and decreasing the number of cells with a single dendrite (Fig. S1A and B). Similar effects were observed with tetrabenazine treatment (Fig. S2A and B). These results suggest that both reserpine and tetrabenazine significantly affect axon and dendrite morphogenesis.

Therefore, we conducted further analyses using the N1E-115 cell line. As expected, treatment with reserpine led to a reduction in process elongation following the induction of differentiation (Fig. 3A and B). These results were supported by decreased expression levels of neuronal differentiation markers growth-associated protein 43 (GAP43) and Tau. In contrast, the expression levels of the control actin protein were comparable in the presence or absence of reserpine (Fig. 3C and D). Similar results were obtained with tetrabenazine treatment (Fig. 4A–D), suggesting that inhibition of VMAT-2 reduces the process outgrowth possibly seen during development.

Fig. 3
figure 3

Treatment with reserpine leads to decreased process elongation. (A, B) Following the induction of differentiation, N1E-115 cells were treated with reserpine or its vehicle and cultured for 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented (** p < 0.01; n = 10 fields). (C, D) The lysates of cells treated with reserpine or its vehicle were used for immunoblotting with antibodies against GAP43, Tau, and control actin. The expression levels of GAP43 and Tau were normalized to those of actin and graphically represented (** p < 0.01; n = 3 blots).

Full size image
Fig. 4
figure 4

Treatment with tetrabenazine leads to decreased process elongation. (A, B) Following the induction of differentiation, N1E-115 cells were treated with tetrabenazine or its vehicle and cultured for 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented (** p < 0.01; n = 10 fields). (C, D) The lysates of cells treated with tetrabenazine or its vehicle were used for immunoblotting with antibodies against GAP43, Tau, and control actin. The expression levels of GAP43 and Tau were normalized to those of actin and graphically represented (** p < 0.01; n = 3 blots).

Full size image

We knocked down VMAT-2 using clustered regularly interspaced short palindromic repeat (CRISPR)/Cas1323,24-fitted guide RNA (Fig. S3) and explored whether VMAT-2 itself mediates process elongation. The knockdown of VMAT-2 reduced process elongation, even when differentiation was induced (Fig. 5A and B). Concomitantly, this knockdown decreased the expression levels of GAP43 and Tau, while those of actin remained comparable between the control gRNA and VMAT-2 groups (Fig. 5C and D). These results illustrate that VMAT-2 itself is involved in the regulation of process elongation following the induction of differentiation. Compared to inhibitor treatment or control gRNA transfection, cells transfected with VMAT-2 gRNA exhibited cell morphologies with a larger cell body diameter. While chemical inhibitors are known to inhibit VMAT-2 activity, they also inhibit VMAT-1 and/or SLC18A3 (VMAT-2 homologues), although the strength of inhibition may vary. Since knockdown of VMAT-2 using gRNA is specific to VMAT-2, the decreased expression of VMAT-2 could be responsible for the larger cell body diameter. Alternatively, inhibition of all VMAT-2 homologues may lead to cell morphologies with a smaller diameter, although the diameter of cells transfected with the control gRNA is approximately equivalent to that of cells treated with each inhibitor.

Fig. 5
figure 5

Knockdown of VMAT-2 results in decreased process elongation. (A, B) N1E-115 cells were transfected with plasmids encoding Cas13 with control gRNA (gControl) or VMAT-2 one (gVMAT-2). Following the induction of differentiation, cells were cultured for 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented (** p < 0.01; n = 10 fields). (C, D) The lysates of the respective cells were used for immunoblotting with antibodies against GAP43, Tau, and control actin. The expression levels of GAP43 and Tau were normalized to those of actin and graphically represented (** p < 0.01; n = 3 blots).

Full size image

Hesperetin recovers decreased process elongation induced by VMAT-2 Inhibition or knockdown

Since hesperetin is known to be a flavonoid with multiple neuroprotective effects25,26,27,28,29,30,31,32, we tested whether it could recover phenotypes induced by VMAT-2 inhibition. Treatment with hesperetin resulted in the recovery of decreased process elongation induced by reserpine or tetrabenazine (Fig. 6A–D; Fig. S4A and B). In addition, treatment with hesperetin recovered the phenotypes following the knockdown of VAMT-2 (Fig. 6E and F; Fig. S4C), suggesting that hesperetin may also be useful for cellular inhibitory phenotypes induced by inhibition of VMAT-2. In contrast, hesperetin itself did not significantly affect process elongation.

Fig. 6
figure 6

Hesperetin recovers decreased process elongation under various experimental conditions. (A, B) Following the induction of differentiation, N1E-115 cells were treated with reserpine and hesperetin (Hes) or its vehicle and cultured for 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented (** p < 0.01; n = 10 fields). (C, D) Following the induction of differentiation, cells were treated with tetrabenazine and hesperetin or its vehicle and cultured for 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented (** p < 0.01; n = 10 fields). (E, F) Following the induction of differentiation, cells transfected with Cas13 and VMAT-2 gRNA (gVMAT-2) were treated with hesperetin or its vehicle and cultured for 3 days. The morphology of the cells at day 0 is also depicted in the diagram. Cells with processes were counted and graphically represented (** p < 0.01; n = 10 fields).

Full size image

We asked whether hesperetin can upregulate phosphorylated Akt, whose levels correlate the activities required for process elongation29,30,31. Hesperetin indeed led to both upregulation of phosphorylated Akt levels after the treatment with reserpine or tetrabenazine (Fig. 7A–D) and upregulation of phosphorylated Akt following the knockdown of VAMT-2 (Fig. 7E and F). Hesperetin itself had no apparent effect on Akt phosphorylation. These results suggest that hesperetin has the ability to recover decreased process elongation with the increased Akt phosphorylation.

Fig. 7
figure 7

Hesperetin recovers decreased Akt phosphorylation under various experimental conditions. (A, B) Following the induction of differentiation, N1E-115 cells were treated with reserpine and hesperetin (Hes) or its vehicle and cultured for 3 days. The immuneprecipitates using antibodies against phosphorylated Akt (pAkt) and the lysates of cells were used for immunoblotting with antibodies against total Akt (Akt). The levels of Akt immunoprecipitated with an anti-phosphorylated Akt antibody were normalized to those of total Akt or actin and graphically represented (** p < 0.01; n = 3 blots). The immunoreactive levels of Akt were also normalized to those of actin. (C, D) Following the induction of differentiation, cells were treated with tetrabenazine and hesperetin (Hes) or its vehicle. The levels of phosphorylated Akt immunoprecipitates were normalized to those of total Akt or actin and graphically represented (** p < 0.01; n = 3 blots). The immunoreactive levels of Akt were also normalized to those of actin. (E, F) Cells were transfected with plasmids encoding Cas13 with and VMAT-2 one (gVMAT-2). Following the induction of differentiation, cells were treated with or without hesperetin (Hes). The levels of phosphorylated Akt immunoprecipitates were normalized to those of total Akt or actin and graphically represented (** p < 0.01; n = 3 blots). The immunoreactive levels of Akt were also normalized to those of actin.

Full size image

Discussion

Tardive dyskinesia involves the involuntary movement of body parts and is often observed in individuals taking antipsychotics for an extended period1,2,3,4. Treatment options include reducing the dosage of medications, changing drugs, or using drugs that can help alleviate symptoms3,4. As a candidate for the third treatment option, inhibiting VMAT-2 activities reduces neurotransmitter uptake into presynaptic vesicles and attenuates overall neurotransmitter receptor activities in highly sensitized postsynaptic membranes, which are believed to be critically responsible for involuntary movements1,2,3,4. VMAT-2 is part of the VMAT molecular family, which includes VMAT-1, encoded in a different genomic region than the vmat-2 gene in humans and rodents. VMAT-1 is primarily expressed on the membrane of dense nuclear vesicles in neuroendocrine cells, such as chromaffin cells in the adrenal medulla, as well as in the nervous system. In contrast, VMAT-2 is primarily found on the membranes of synaptic vesicles in nerve terminals in the central nervous system and sympathetic neurons9,10,11,12. Thus, specific inhibition of VMAT-2 can be linked to a more targeted effect of reducing receptor activity in postsynaptic membranes. In our in vitro experiments, reserpine and tetrabenazine exhibited similar effects on process elongation. Both reserpine and tetrabenazine are well-known inhibitors of VMAT-25,6,7,8, but reserpine shows strong binding activities to the monoamine recognition site of both types of VMAT proteins, blocking the transport of monoamines into intracellular vesicles33. On the other hand, tetrabenazine is thought to exert its inhibitory effect by binding specifically to a VMAT-2 surface site that differs from that of reserpine33,34, which is why tetrabenazine continues to be used clinically to treat tardive dyskinesia5,6. Despite the clear differences in the clinical uses of these two drugs13,14, our studies provide possible direct evidence that either reserpine or tetrabenazine should be used with caution, as they may inhibit or delay the elongation of processes during developmental stages. Further studies in mice can help translate the in vitro phenomena to in vivo situations and provide mechanistic evidence regarding their effects during developmental stages.

Hesperetin is known to exhibit multiple neuroprotective effects in models of Alzheimer’s disease and Parkinson’s disease25,26,27,28,29,30,31,32. In this study, we found that hesperetin has a protective effect against decreased morphological differentiation caused by inhibition of VMAT-2 in cells. However, it remains unclear how hesperetin can recover these phenotypes. While hesperetin is known to act as an antioxidant, similar to other flavonoids, as an inhibitor of some phosphatases, and/or as an activator of protein kinase29,30,31,32, the protective effect may stem from one or more of these mechanisms. This situation is reminiscent of existing examples where a therapeutic drug is administered alongside a protective or supplementary compound35,36. The simultaneous administration of a tardive dyskinesia drug and hesperetin could pave the way for such applied research.

It remains unclear how the inhibition or knockdown of VMAT-2 activity decreases process elongation in neuronal cells. Monoamine neurotransmitters such as dopamine, adrenaline, and noradrenaline, secreted from vesicles containing VMAT-2, could have positive autocrine effects on process elongation. Recent findings have shown that dopamine maintains axonal morphogenesis and function34. It is known that the D2-type dopamine receptor stimulates neurite outgrowth through cyclic AMP signaling37, contributing to morphological differentiation in neuronal cells38. Conversely, L-3,4-dihydroxyphenylalanine (L-DOPA) a dopamine precursor, binds to G protein–coupled receptor 143 (GPR143) and inhibits neuritogenesis39. It remains unknown whether GPR143 is involved in the downregulation of cyclic AMP signaling39; however, it is clear that dopamine and its metabolic intermediates play roles in promoting or inhibiting neuritogenesis. Adrenaline and its metabolic intermediates are also known to stimulate neuritogenesis through calcium signaling40,41,42. Similar to dopamine, it is likely that adrenaline and its intermediates contribute to neuronal cell morphological differentiation40,41,42. Given that monoamine neurotransmitters often have a positive effect on neuritogenesis during both development and adulthood, it is predicted that monoamines derived from VMAT-2 vesicles are responsible for neuritogenesis. Therefore, the effects of inhibition of VMAT-2 on neuronal cells are thought to impact not only developmental stages but also neuritogenesis and possibly neurogenesis in adult stages9,10,13,14. It may be important to determine the relationship between the levels of monoamine neurotransmitters, such as dopamine or adrenaline, in the brain or, if possible, in the spinal fluid or blood, and the dosage of VMAT-2 inhibitions administered. It may also be important to consider the relationship between disease severity and drug dosage. Although this study is at the cellular level, hesperetin may help broaden the therapeutic window by co-administering it with VMAT-2 inhibitors.

In addition to VMAT-2, another member of the SLC18A family member molecule SLC18A3 (also known as vesicular acetylcholine transporter [VAChT]) is expressed in the central nervous system (see the Human Protein Atlas website, https://www.proteinatlas.org/). This protein is widely expressed in neurons, including motor neurons, the facial nucleus, and hypoglossal nerves. Cells expressing SLC18A3 are cholinergic neurons, and this transporter functions to uptake acetylcholine, but not dopamine, into synaptic vesicles43,44. Since reserpine and tetrabenazine inhibit the monoamine neurotransmitter uptake activity of SLC18A343,44, future studies should carefully investigate whether the effects of these drugs are delayed in developing neurons by affecting both SLC18A3 and VMAT-2. Some mutations of SLC18A3 are known to be associated with less severe presynaptic congenital myasthenic syndrome type 21 (CMS21)45, but it is likely that the symptoms are specifically observed in peripheral tissues.

Congenital mutations of the vmat-2 gene are associated with infantile-onset parkinsonism-dystonia type 2 (PKDYS2) in humans46,47,48. Patients with PKDYS2 show a decrease in monoamine neurotransmitters. This condition is an autosomal recessive infantile-onset neuropathy characterized by a complex array of symptoms, including parkinsonism, dystonia, abnormal movements, and autonomic dysfunction, often accompanied by varying degrees of developmental delay. PKDYS2-associated mutations in VMAT-2 are likely linked to a loss of function in its transporting activities46,47,48. Also, heterozygous VMAT-2 knockout mice exhibit marked hypersensitivity to the locomotor effects of psychostimulants, while homozygous mice display abnormal brain structures and die within the first or second week of life49,50. Importantly, incomplete VMAT-2 activity affects not only abnormal movements but also multiple developmental defects, including those in the brain49,50. These symptoms are at least partly consistent with our results showing that inhibition of VMAT-2 leads to decreased morphological changes in cells. Here, we demonstrate for the first time that the treatment with the VMAT-2 inhibitors reserpine or tetrabenazine, or the knockdown of VMAT-2, decreases process elongation in neuronal cells. The phenotypes resembling the side effects targeting VMAT-2 can be recovered with concomitant treatment with hesperetin. Further studies can promote our understanding of the detailed mechanism by which VMAT-2 is responsible for morphological differentiation in the early developmental stage and investigate whether inhibition of VMAT-2 has little or no effect on morphogenesis in adulthood, using various types of isolated primary cells from brain tissues and genetically modified mice. Additional studies in this area may contribute to the development of drugs with fewer possible side effects for tardive dyskinesia and aid in identifying new drug target molecules.