3.1. RA reduces polyQ and a-syn mediated OS damage and improves antioxidant defense in nematode models
As a deeper insight into the pathogenesis of ND, OS damage is considered a major contributor to disease pathogenesis, which is defined as an imbalance between the production of ROS and the body’s antioxidant defense capacity [21]. Therefore, natural antioxidants have been proposed as potential treatments for neurological disorders [22]. Our previous studies have revealed that RA effectively counteracted OS damage in nematodes, suggesting that RA may also be a potential prevention for ND [16]. To deeply explore the neuroprotective effect of RA in delaying NDs, it is necessary to further examine the effects of RA on OS damage and antioxidant defense capacity in NDs models. Firstly, we determined the ROS content, and the results showed that both HD and PD strains treated with RA had lower ROS content than that incubated with E. coli OP50 (Fig. 1a and 1b). Notably, excess ROS can damage cells by oxidizing the proteins, DNA, and lipids, leading to neurodegeneration and mitochondrial dysfunction [23]. Among these, MDA represents a common peroxidation product in patients with NDs, and some studies have even suggested that MDA testing could be used as a biomarker to determine the progression of neurodegenerative diseases [24]. Therefore, the present study further detected the MDA levels in nematodes. Interestingly, we observed significantly lower MDA levels in the RA group compared to the control, suggesting that RA has the potential to reduce the neuroprotective effects of polyQ and a-syn mediated OS damage (Fig. 1c and 1d). When the organism is attacked by OS, the antioxidant defense system serves as an important weapon to perform detoxification functions [25]. In particular, SOD plays an important role in resisting OS damage [25]. Unexpectedly, RA treatment significantly enhanced SOD enzyme activity in HD and PD models, indicating that RA can boost the antioxidant defense capacity of NDs (Fig. 1e and 1f), and the specific mechanism may involve defense against polyQ and a-syn induced OS damage by increasing SOD expression. Taken together, the above results adequately demonstrate that RA can ameliorate the oxidative damage of NDs and promote the antioxidant defense capacity in C. elegans, and it is worthwhile to further explore the RA-mediated neuroprotective mechanism.
3.2. RA improved polyQ and ɑ-syn mediated neuronal loss by activating Wnt signaling and Notch signaling
In NDs patients, OS and proteotoxicity are largely responsible for neurodegeneration. [26]. Given that RA exhibited strong antioxidant activity in C. elegans, we curiously wondered whether RA influenced the neuronal functional homeostasis in NDs models. The HA759 and BZ555 strains, which respectively labeled ASH neurons and dopaminergic neurons with GFP, were employed to detect the survival of neurons in the NDs model [27, 28]. As the associated pathogenic protein accumulates, its mediated proteotoxicity results in symptoms of loss of relevant neuronal fluorescent markers in both HA759 and BZ555 nematodes [27, 28]. Firstly, in the HD model HA759 strains, representative fluorescence images of ASH neurons were presented in Fig. 2a, with white arrows pointing to normally surviving neurons and red arrows pointing to dead neurons. The statistical results revealed that the survival rate of ASH neurons in the E. coli OP50 treatment group was only 37.78%, while the RA treatment group had a higher survival rate of 62.22%, indicating the potential of RA to improve ASH neuronal lesions (Fig. 2b). Further, since paraquat was one of the well-recognized drugs to induce PD syndrome in laboratory animals [5], we treated BZ555 strain with 0.4 mM paraquat to establish PD neuron injury model. And it had been reported that paraquat-treated animals exhibited major features of PD, including behavioral deficits and disruption or loss of DA neuronal morphology [5, 20]. In our study, massive DA neuron death was observed in paraquat-treated nematodes. The quantitative results showed that the number of normal DA neurons in the paraquat-treated group was significantly decreased compared with the normal-treated group, indicating that the PD nematode model was successfully established (Fig. 2d). Interestingly, once treated with RA, the DA neuronal damage was alleviated and the number of normal DA neurons increased significantly (Fig. 2d), which was consistent with Wang et al.’s research that RA could directly reduce 6-OHDA-mediated death of the nigra DA neurons [19]. Overall, the above experiments confirmed that RA could ameliorate neurological homeostatic imbalances and protect neurons from OS damage and proteotoxicity in NDs.
To delve into the RA-mediated underlying molecular mechanisms of regulation of neuron homeostasis and synaptic function, we analyzed signaling pathways closely related to neuronal disorders in NDs, such as the Notch pathway and the Wnt pathway. Firstly, the Wnt signaling pathway performed a critical role in regulating the structure and function of the adult nervous system [29, 30]. Increasing evidence suggested that Wnt signaling could regulate the formation and function of neuronal circuits via modulating dendrite development, synaptic function, and neuronal plasticity [29, 30]. In NDs, dysfunction of the Wnt pathway causes pathogenic protein-induced synaptic dissociation and may lead to memory impairment and cognitive decline [31, 32]. Given the crucial role of the Wnt signaling pathway in the synaptic homeostasis and pathogenesis of NDs, several molecular components of this pathway, such as Wnt ligands, Frizzled receptors, Dishevelled and β-catenin, have been proposed as targets for the treatment of NDs [31]. Thus, we examined the RNA level of key genes of the Wnt signaling pathway in C. elegans, including lin-44 (encoding Wnt ligands/LIN − 44, lin-17 gene (encoding Wnt Frizzled/LIN-17 receptor) and bar-1 gene (encoding β-catenin). The result presented that RA remarkably down-regulated the RNA expression of lin-44 in the HD model, and up-regulated the lin-17 gene’s expression in the PD model (Fig. 2e, 2f). Moreover, it had been reported that overexpression of the Wnt/Frizzled receptor could combine with regulated Wnt against the toxicity of Aβ oligomers and increase cell survival [33]. Hence, RA might experimentally protect neurons from polyQ and ɑ-syn mediated neurotoxicity by downregulating lin-44 expression in HD strains and upregulation of lin-17 expression in PD. Next, another important pathway, the Notch pathway also acted a critical role in neuronal development and regulating processes such as neurogenesis, synaptic plasticity, and long-term memory, which were associated with cognitive function [34]. What's more, dysregulation of the Notch signaling pathway was associated with the development of NDs including AD, while the altered Notch signaling pathway might induce Aβ accumulation [34, 35]. In this case, two genes lin-12 and glp-1, responsible for encoding the Notch related receptor proteins were also examined. Experimental results showed that RA treatment didn’t have a significant effect on lin-12 and glp-1 in the PD model, but increased glp-1 expression and decreased lin-12 expression in the HD model. This result also suggested that LIN-12/GLP-1 receptors were only involved in the RA-mediated improved survival of ASH neurons in HD, but not in the PD model. Moreover, chronic dysregulation of Notch signaling might lead to neuronal degeneration and ultimately result in specific learning and memory deficits [34, 35]. Besides, the Notch pathway was known to interact with the Wnt pathway to jointly participate in regulating neurogenesis, survival, and synaptic function [34]. Hence, the data suggested that RA-mediated improvement in neuron homeostasis involved the following two possibilities, including a reduction in polyQ-mediated ASH neuronal death via regulating the Notch/LIN-12/GLP-1 signaling combined with Wnt/LIN-44 signaling, and inhibition in ɑ-syn mediated DA neuronal death via upregulating Wnt/LIN-17 signaling.
3.3. RA significantly alleviated polyQ and ɑ-syn induced dyskinesia
In NDs, neurological disturbances or death in the brain could lead to a range of movement disorders such as resting tremors, bradykinesia, and uncoordinated movements [7]. Given that we have already demonstrated that RA could rescue ASH and DA neurons from pathogenic protein-induced neurotoxicity, it was necessary to further explore the regulatory role of RA on dyskinesia in the NDs’ nematode model. Next, several transgenic strains, including AM140 strains (expressing polyQ in body wall muscle cells), HA759 strains (expressing polyQ in ASH neurons), NL5901 (expressing ɑ-syn in body wall muscle cells), and BZ555 (paraquat inducing the death of DA neurons) were used to evaluate dyskinesia in NDs. As these proteins are expressed and aggregated in muscle cells and neurons, their mediated proteotoxicity can potentially result in muscle fiber damage and neuronal loss, ultimately leading to motor impairment and cognitive deficits [2]. Thus, we could qualitatively determine the expression aggregation of the pathogenic proteins through determining the paralysis rate of the relevant mutants. Firstly, as expected, once treated with the RA sample, the HA759’s paralysis curve was significantly right-shifted compared to the control (Fig. 3a). Additionally, there was a progressive increase in mean lifespan, median longevity, and maximum longevity of HA759 (Table 1). It was important to note that myoideum injury and neuronal damage induced motor abnormalities might affect nematode swallowing, head pendulum, and mobility, etc [2]. To deeply explore the effect of RA on HD nematodes’ mobility, the number of head swings, swallowing frequency, and mean movement velocity were deeply detected. The results were presented in Fig. 3b and 3c, which showed that head swings and swallowing frequency showed significant enhancement in both AM140 and HA759 strains on day 5. Furthermore, to test the effect of RA on nematodes’ body bending, wormlab software was employed to analyze the average movement speed, which scientifically reflected the movement of the nematodes. Equally, a significant increase in the average speed was observed in HD worm models in the RA treatment group compared to the control group (Fig. 3d). However, in our previous report, no changes were observed in the AM140 strains’ paralysis curve and paralysis lifespan, which behaved very differently from HA759 strains [16]. Since the motor impairment of AM140 resulted from polyQ aggregation, RA enhanced the integrated viability of AM140. So the possibility of this observed difference might be that RA-mediated reduction in polyQ aggregation in AM140 was sufficient to improve overall motor vigor, but was not adequate to slow down the overall paralysis process. Overall, RA had performed extremely well in improving polyQ-mediated neuronal damage and behavioral deficits in HD nematode models.
Table 1
Statistical analysis of the paralysis of C. elegans
Strains
|
Treatment
|
Mean
paralysis
|
Medium
paralysis
|
Maximum
paralysis
|
Mean fold
increase%
|
Uncensored/n
|
HA759
|
OP50
|
8.66 ± 0.47 a
|
10.07 ± 0.31a
|
19.66 ± 0.47a
|
/
|
173/204
|
RA
|
12.00 ± 0.00b
|
11.94 ± 0.43b
|
24.00 ± 0.82b
|
20.44
|
154/201
|
NL5901
|
OP50
|
13.48 ± 0.57 a
|
13.67 ± 0.33 a
|
21.67 ± 0.57 a
|
/
|
155/180
|
|
RA
|
17.07 ± 0.34
|
16.96 ± 0.49
|
24.33 ± 0.35
|
21.03
|
148/185
|
BZ555
|
OP50
|
10.58 ± 0.45 b
|
10.67 ± 0.33 b
|
19.00 ± 0.00 b
|
/
|
168/183
|
|
OP50 + PQ
|
8.75 ± 0.42 a
|
8.33 ± 0.35 a
|
16.67 ± 0.57 a
|
/
|
145/189
|
|
RA + PQ
|
10.67 ± 00.49 b
|
10.06 ± 0.39 b
|
19.33 ± 0.36 b
|
17.99
|
158/190
|
The data were analyzed by one-way ANOVA (SPSS 23), and different letters in a column denote values that are significantly different (p < 0.05). The mean fold increase was calculated by (T-C)/C*100, where T is the mean paralysis time of C. elegans treated with RA and C is the mean survival time of Escherichia coli OP50 (E. coli OP50) treatment. In BZ555 strains, T is the mean paralysis time of C. elegans treated with (RA + PQ) RA and PQ (paraquat), and C is the mean survival time of (OP50 + PQ) E. coli OP50 and PQ treatment. |
Secondly, to evaluate the effect of RA on PD-like behavioral deficits induced by ɑ-syn, we first detected the NL5901 and BZ555 strains’ paralysis behaviour. And RA treatment remarkedly ameliorated both NL5901 and BZ555 strains’ dyskinesia, with remarkably right shifting the paralysis curves of the two strains compared with the control group (Fig. 3e and 3f). Similarly, we examined head swings, swallowing frequency and motility speed of the PD worms to further investigate the role of RA in regulating the motility of various parts of the nematode body. As revealed in Fig. 3g, compared to the controls, significant increases in head swings were observed in conditions with the RA in both NL591 and BZ555. In addition, RA enhanced BZ555 strains’ swallowing frequency, while little difference was observed in NL5901 (Fig. 3h). Finally, RA improved the NL5901 and BZ555 strains’ overall motility speed, which could presumably reflect the movement of the whole body of the nematode(Fig. 3i). Thus, the present study elucidated the improvement effect of RA on the behavioral disorders triggered by the deleterious pathogenic protein aggregation in NDs from different aspects, including head pendulum, swallowing, paralysis behavior, and global movement speed. And the data supported that RA had the characteristics of rescuing dyskinesia and improving sports performance in HD and PD strains, respectively.
3.4. The IIS pathway, MAPK signaling, and HSF-1signaling involved in the RA-mediated mechanism of polyQ and ɑ-syn reduction
In HD, an abnormal expansion of polyQ within the first exon of Htt leads to an adverse event that the mutant Htt clusters into multiple toxic oligomers and fibrils in a polyQ length and concentration-dependent pattern, ultimately leading to behavioral and cognitive impairment [3, 36]. Similarly, the aggregation of harmful ɑ-syn proteins is a key pathological feature of PD patients [37]. Therefore, it was necessary to further explore the impact of RA on the deposition of these pathogenic proteins. However, it was worth noting that the paralysis experiments above could only indirectly reflect polyQ and ɑ-syn aggregations, but could not directly and quantitatively analysis the aggregations of pathogenic protein. Thus, to delve into the effect of RA on the aggregation of polyQ and ɑ-syn in the NDs model, the AM140 and NL5901 strains, respectively co-expressing polyQ and ɑ-syn with a fluorescent protein in body wall muscle cells, were employed to visualize and quantity the aggregation of pathogenic proteins. As displayed in Fig. 4a, the overall fluorescence of strain AM140 was observably darker in the RA-treated group compared to the control group, and quantitative results also revealed that the RA group markedly reduced the fluorescence intensity of the strain (Fig. 4b). This result indicated that RA reduced the misfolding aggregation of the pathogenic protein polyQ in HD, which was consistent with the previous results of reduction of ASH neuron damage and improvement of motor capacity in HD strains. Similarly, in the PD model NL5901 strain, the red arrow pointed to the protein fibrils formed by ɑ-syn aggregates, which were highly toxic to neurons (Fig. 4c). Interestingly, after the RA intervention, the number of ɑ-syn fibrils was significantly reduced, suggesting that RA successfully diminished the abnormal aggregation of ɑ-syn (Fig. 4d). Overall, RA has exhibited outstanding performances in improving polyQ and ɑ-syn-mediated neuronal damage, behavioral deficits and overall motility in nematodes. And these excellent performances could mainly result from the RA-mediated reduction of abnormal aggregation of pathogenic proteins polyQ and ɑ-syn. In all, these above results highlighted that RA displayed neuroprotective activity in NDs models, so it was worthwhile to further explore the mechanisms of action of RA-mediated retardation of NDs.
Next, to further elucidate the details of RA's retardation of HD and PD pathogenic protein aggregation, we performed molecular docking to predict the binding mechanisms and key interactions between RA and Htt as well as ɑ-syn. Firstly, the Htt tetramer in the research of Kotler et. al. [38] consisting of an amphiphilic domain, a polyQ chain, and a proline-rich sequence, was regarded as a minimalistic construct involved in oligomerization of HD patients and remains monomeric states (PDB code 6N8C), which is of great benefit to analyze the detailed interactions of RA and Htt. Therefore, we first employed the AutoDock software to perform molecular docking analysis of RA and Htt tetramer. As summarized in Fig. 4e, RA (pink) could directly bind to the Htt tetramer (red, PDB code 6N8C) through hydrogen bonds and an attractive charge between the ligand and amino acid residues with a total binding energy of -5.5 kcal/mol. Specifically, RA bound Htt tetramer via the formation of one conventional hydrogen bond at LysA:5, one carbon-hydrogen bond at ThrA:2 and an attractive charge at GluA:4. Further, we detected the interaction between RA (pink) and α-syn aggregates (blue, PDB code 5F1T) which closely associated with increased loss of DA neurons in PD patients [39]. Theoretically, RA could directly bind to the α-syn through a hydrogen bond at DthA:15 with a binding energy of -7.57 kcal/mol. Overall, our molecular docking results demonstrated that RA and the pathogenic proteins Htt and ɑ-syn might be linked together to form stable complexes. In NDs, pathogenic proteins are transformed from native conformation to β-sheet structure [40]. The Thr, Lys, and Glu residues in Htt and Dth residue in α-syn are important components of the β-chain [41]. Further, with the increase in the number of polyQ residues and α-syn, these extended sequences have a stronger affinity for each other and interact to form β-sheet chains [42]. Once the β-sheet-rich protein is formed and the protein concentration exceeds a critical value, pathogenic protein aggregates with fibrillar morphology, and neuronal inclusions are formed, which are protease-resistant and toxic to neuronal cells [42]. In our study, RA could interact with Thr, Lys, and Glu of Htt protein and Dth of α-syn protein to form stable complexes. The formation of these complexes can weaken the stability of the β-sheet structure and prevent the expansion of Htt long chains and α-syn aggregates via affecting chemical bonds. In addition, it had been reported that drugs with the capacity to bind to the polyQ sequence in Htt might stabilize the natural conformation of the protein, thereby preventing or delaying the production of β-folded structures in the protein [40]. Therefore, our study suggested that RA might act as a key driver that interferes with the conformation transformation of natural proteins into β-sheet structures, ultimately reducing the abnormal aggregation of pathogenic proteins. Overall, the computer simulation data were consistent with the above results that RA diminished polyQ and α-syn aggregations in NDs nematodes, further elucidating the mechanism of RA-mediated reduction of pathogenic proteins specifically.
In NDs, protein conformational disorders triggered by disorders in the protein quality control system are one of the causes of the abnormal accumulation of pathogenic proteins. Here, in an attempt to further unravel the mechanisms behind RA-mediated amelioration in protein homeostasis in nematodes, the effects of RA on several key genes related to protein aggregations and quality control were investigated. It had been reported that reducing the insulin/insulin-like growth factor 1 signaling (IIS) could delay aging and protect model animals from pathogenic protein induced toxicity and OS [43–45]. Firstly, we determined the RNA expression level of the key genes in IIS, including daf-2, daf-16, and hsf-1. Our results showed that RA evidently down-modulated the RNA level of daf-2 in PD strains, but had no influence on the daf-2 gene in HD mutants, indicating that RA acted through DAF-2/IIS pathway to regulate ɑ-syn aggregation in PD (Fig. 4g and 4h). It was reported that daf-16 and hsf-1, as the downstream genes of daf-2, also acted important roles in suppressing misfolding proteins aggregation and their induced neurotoxicity [43, 44]. Additionally, RA significantly reduced daf-16 RNA level and enhanced hsf-1 expressions in both HD and PD nematodes, suggesting that daf-16 and hsf-1 might be key regulators of protein toxicity reduction induced by RA (Fig. 4g and 4h). These results also were in accord with Cohen and Dillin’s research that a fork head box O transcription factor (DAF-16) and heat-shock factor 1 (HSF-1) regulated the counter-toxic effect in IIS [43]. However, our previous study found that RA significantly upregulated the expression of the daf-16 gene in strain Bristol N2, in contrast to the NDs model [16]. The main possibilities for this difference are as follows. Firstly, DAF-16, as a key transcription factor in the IIS pathway, could also receive signals from multiple pathways, which occurred the complexed relationships [43, 46]. Further, the regulation of daf-16 by RA in different models could be influenced by many factors, thus leading to differences in daf-16 gene expression in N2 and NDs models. To get insight into the regulatory role of RA on downstream genes of the IIS and HSF-1 pathway, changes in RNA levels in downstream genes of daf-16 and hsf-1 were further analyzed, including OS genes (sod-1, sod-3) and small heat shock protein (sHSPs) genes (hsp-16.1 and hsp-16.2). Surprisingly, the genes hsp-16.1 and hsp-16.2 were up-regulated in both HD and PD worms, while the sod-1 and sod-3 were not influenced by RA (Fig. 4g and 4h). The HSF-1 activated the expression of the sHSPs genes to act as a molecular chaperone and controlled the correct folding of the protein and prevented the formation of toxic oligomers [47]. And the molecular chaperone networks usually serve as a monitoring mechanism in protein quality control systems, aiming to ensure that proteins achieve stable folded conformational states and remove abnormally folded proteins [47, 48]. As a major activator of chaperone expression, hsf-1 not only regulated the expression of genes involved in protein degradation, but also promoted the expression of genes encoding chaperones in response to cellular stress [47–49]. Moreover, overexpression of the HSP-16 family of small chaperone proteins decreases the number of toxic Aβ aggregates [50]. Hence, the data revealed that RA activated the molecular chaperone monitoring system via upregulating hsf-1 expression to activate hsp-16.1 and hsp-16.2 genes, ultimately reducing the accumulation of polyQ and ɑ-syn associated with OS. In all, the IIS and HSF-1 pathways were involved in the RA-mediated reduction of polyQ and ɑ-syn aggregations and neurotoxicity.
Meanwhile, activation of mitogen activated protein kinases (MAPK) pathway had been observed to perform ameliorative effects on various inflammatory cell stresses, such as proinflammatory cytokines released to activated microglia due to the accumulation of misfolded proteins [51]. Three key transcription factors in MAPK signaling, such as UNC-43 (a type II Ca2+/calmodulin-dependent kinase (CaMKII)), SEK-1 (a MAPK kinase), and SKN-1(the homologs of the mammalian NF-E2-associated factor 2 (Nrf2)), acted together to against OS and regulate misfolding protein-mediated neuronal cell survival [52, 53]. Therefore, we further examined the effect of RA on the expression of those three key transcription factors of the MAPK pathway. In the HD model, RA down-modulated sek-1 gene and up-regulated skn-1 gene (Fig. 4g). However, the increasing expressions of both skn-1 and unc-43 genes were observed in PD strains (Fig. 4h). Therefore, the regulatory effects of RA treatment on the key transcription factors of the MAPK signaling pathway were different in both the HD and PD models. The network of RA-mediated modulation of MAPK signaling was complex, and the interactions between MAP, CaMKII, and Nrf2 were also different between HD and PD models, so further detailed investigations were required.
The main pathogenesis of NDs is a neurological disorder triggered by protein misfolding aggregation. The above results confirmed that RA could significantly reduce polyQ and ɑ-syn aggregation and their mediated neurotoxicity. Moreover, it was demonstrated through computer technology that RA could specifically bind to Htt and ɑ-syn to prevent the protein from developing larger aggregates and inhibited further protein denaturation. Further, the possible mechanism of action was that RA activated key genes of IIS (daf-16 and daf-2), MAPK signaling (skn-1 and sek-1), and HSF-1 signaling (hsp-16.1, hsp-16.2, and hsf-1) to reduce polyQ accumulation in HD, and activated key genes of IIS (daf-16 and daf-2), MAPK signaling (skn-1 and unc-43), HSF-1 signaling (hsp-16.1 and hsf-1) to reduce ɑ-syn accumulation in PD.
3.5. RA ameliorated polyQ disease and PDs’ common mitochondrial dysfunction
It was evidenced that mitochondrial dysfunction mediated by protein toxicity was one of the major symptoms of NDs [9]. In HD, mutant Htt induced by polyQ would disrupt mitochondrial dynamics and biogenesis, and also interact directly with mitochondrial proteins, ultimately disrupting the mitochondrial structural integrity to reduce ATP synthesis and mitochondrial bioenergetic activity [54]. Alternatively, the ɑ-syn of PD would interact with the mitochondrial proteins, which hindered the ATP/ADP exchange between the mitochondria and the cytoplasm and impaired the mitochondrial protein input, finally destroying the mitochondrial function [37]. Mitochondria in neurons, acting as highly dynamic organelles in structure and function, played a crucial role in various neuronal functions [1, 54, 37]. For example, the transport of neurotransmitters, release of cargo in the synaptic cleft, and maintenance of neuronal ATP levels all depended on mitochondrial activity and integrity [10, 1]. Thus, mitochondrial disorders in NDs could trigger a range of functional abnormalities, including ATP descending, MMP decrease, ROS increase, and abnormal mitochondrial structure. And all of that would be very dangerous for the death of neurons in NDs [9, 1]. In particular, the mitochondria were susceptible to OS and contained multiple electronic vectors capable of producing ROS, and an abnormal increase in ROS levels was inevitable when its function was compromised [1]. Apropos, this research had confirmed RA evidently reduced ROS levels in worms, suggesting RA could attenuate oxidative damage induced by mitochondrial dysfunction in NDs and display a strong anti-oxidative activity relating to neuroprotection. Further, we analyzed the changes in mitochondrial ATP and MMP levels. In present case, for HD, significant recoveries of ATP and MMP were observed in both AM140 and HA759 (Fig. 5a and 5c). Besides, in PD, the ATP and MMP content of BZ555 treated with paraquat were significantly lower than those treated with E. coli OP50, while both ATP and MMP were evidently raised in BZ555 and NL5901 as treated with RA (Fig. 5b and 5d). The above results implied that RA had the properties of improving the mitochondrial dysfunction in HD and PD models, including an increase in ROS, and a decrease in ATP and MMP.
However, possessing normal mitochondria structure was one of the important prerequisites for maintaining normal mitochondrial function and enabling ATP synthesis, normal ROS production and keeping MMP at normal levels [55]. Therefore, we hypothesized that RA might improve mitochondrial function via maintaining the normal structure of the mitochondria in NDs. In addition, to delve into the effects of RA on mitochondrial structure, the nematode strain PD4251 with GFP-labelled mitochondria in myoblasts was hired to detect changes in mitochondrial structure. Mitochondria are highly dynamic organelles that control mitochondrial morphology, length, size and number through constant division and fusion with each other [56, 10]. Normal mitochondria in nematode muscle cells have been reported to be tubular in shape and when mitochondrial dynamics are abnormal, the number of fragmented punctate mitochondrial structures will arise, affecting normal mitochondrial function [56, 10]. As presented in Fig. 6a, we observed that the nematodes in the experiment mainly had three morphological mitochondria, including normal tubular mitochondria, abnormally inflated and punctiform ones, and the other ones in an intermediate state, which was consistent with literature reports [56]. As expected, statistical results showed that RA treatment could not only greatly increase the proportion of normal mitochondria, but also reduce the proportion of abnormal mitochondria compared to paraquat model group, which confirmed our conjecture that RA rescued mitochondrial dysfunction in NDs models by reducing abnormal mitochondrial dilatation and deformation (Fig. 6b). The translocase of the outer mitochondrial membrane (TOM complex) mediates the translocation of proproteins through the mitochondrial outer membrane, which plays an important role in mitochondrial biogenesis and maintenance [57]. TOM-7 is an important component of the TOM complex and a core element of the protein conductive pore [57]. As shown in Fig. 6c, TOM-7 in the transgenic nematode DLM14 was co-expressed with GFP. A significant decrease in fluorescence intensity occurred in the paraquat model group compared to the E. coli OP50 group in Fig. 6d, while a significant recovery of fluorescence intensity was observed after RA treatment (Fig. 6d). The improvement of TOM-7’s expression mediated by RA was closely related to maintenance of mitochondrial structure and function, since TOM-7 played an important role in mitochondrial normal biogenesis and normal structure maintenance [57]. In addition, mitochondrial intimal complexes I, II, III, IV, and V play important roles in mitochondrial electron transport chains and redox reactions [58]. As shown in Fig. 6c in the transgenic strain CB7272, the green fluorescent protein expressed in the body wall and pharyngeal muscle along with complex I (ccIs4251), complex II (mIs12) and complex III (dpy-17), while the red fluorescent protein co-expressed in the epidermis and pharyngeal muscle with complex IV (frIs7) and complex V (uIs69). Quantification results showed that RA improved the expression of complexes I, II, and III, but had no effect on complexes IV or V. Since complex I and III were the main sites of ROS production in the mitochondrial electron transport chain (Lenaz et al., 2006), we speculated that RA mediated decrease of ROS showed a close connection with its ability to improve the expression of complex I, II and III. Our results highlighted that RA enhanced the expression of TOM-7 and complexs I, II, and III in NDs (Fig. 6d).
Overall, RA could indeed improve ATP, and MMP levels and reduced ROS content in HD and PD models. At the same time, RA raised the structural homeostasis of mitochondria typically and diminished the proportion of abnormal mitochondria in nematodes. RA had shown excellent performance in improving the expression of the outer mitochondrial membrane proteins TOM-7 and inner mitochondrial membrane complexes I, II and III. Therefore, this study comprehensively illustrated the excellent efficacy of RA in improving mitochondrial structure and dysfunction in NDs, respectively.
3.6. RA improved mitochondrial dynamic homeostasis through the regulation of genes related to mitochondrial dynamics
In the nervous system, the high energy requirements of neurons were closely related to the function, maintenance, and dynamics of the mitochondria [54]. The key roles of mitochondria in neurons, including ATP generation, ROS generation, and antioxidant activity, were closely related to mitochondrial biogenesis, fission, and fusion [54, 1]. To gain further insight into the underlying mechanisms of the RA-improved mitochondrial dysfunction effect, we examined the genes associated with the mitochondrial dynamics. Firstly, phb-1 and phb-2 genes were respectively responsible for encoding prohibitin (PHB) 1 and PHB2, which together formed large circular complexes acting as membrane scaffolding in the inner membrane [59]. Besides, the PHB is essential for mitochondrial fusion, as well as neuronal ultrastructure and genomic stability [59, 60]. RA upregulated the expression of phb-2 genes in the PD model, but upregulated phb-1 in the HD model, which indicated that the phb-2 gene might involve in RA-mediated mitochondrial function and structural maintenance of the PD model, while phb-1 was involved in another HD model (Fig. 7a and 7b). Since PHB-overexpression could inhibit OS-induced damage, including inhibiting mitochondrial ROS production, promoting ATP production, and improving MMP [59, 61], RA might improve neuron homeostasis and mitochondrial function through enhancing PHB1 and PHB2 expression in HD and PD models, respectively. Additionally, it was reported that the PHB complex was required for the maintenance of optic atrophy 1 protein (OPA1) stability [59]. PHB could also modulate OPA1, a protein required for inner membrane fusion of mitochondrial membranes and cristae morphogenesis, to control cell proliferation and apoptosis [38, 62]. Moreover, in the nervous system, neuronal cells’ functions greatly depend on mitochondria fusion and fission due to the high energy requirements of these cells [63]. Thus, we further investigated the mRNA expression levels with genes encoding proteins associated with mitochondrial fusion and fission. And the two genes responsible for regulating fission are drp-1 and fis-1, encoding dynamin-related protein 1 (DRP-1) and fission 1, respectively. The two genes responsible for regulating fusion are mmf-1 and eat-3, encoding mitofusins 1 and the homolog of OPA1) [63, 64]. In HD, RA increased the expression of fis-1 and mmf-1, and downregulated the expression of the drp-1 and eat-3 genes (Fig. 7a). Additionally, in PD, RA also upregulated fis-1 and mmf-1 expression, but it increased drp-1 and eat-3 gene expression, which was contrary to the results of the HD model (Fig. 7b). We speculated that RA-mediated differences in the gene expressions related to mitochondrial dynamics in HD and PD models were related to these two following facts. First, mitochondria were highly dynamic organelles, while their fusion and fission processes were also complex and dynamically variable [10]. Second, despite the profiles that HD and PD had similar protein aggregation toxicity, the related pathogenic proteins and their mediated mitochondrial dysfunction differed [9, 8]. Thus, RA mediated distinct targets of intervention in two different NDs models during the complex mitochondrial dynamics. Another protein, ANC-1, which affected the mitochondrial structure, contained several helical regions, including a nuclear membrane localization domain, and an actin-binding domain [65]. ANC-1 controlled mitochondrial localization and connected the nucleus to the cytoskeleton by interacting with actin in the cytoplasm [65]. However, RA had no effect on anc-1 gene mRNA expression in either the HD or PD model, suggesting anc-1 didn’t participate in RA-induced amelioration of mitochondrial function (Fig. 7a and 7b). Another atp-2 gene was responsible for encoding the active site or β-subunit of complex V in C. elegans [66]. The complex V participated in forming mitochondrial respiratory chain which was capable of generating cellular energy in the form of ATP and regulating ROS production [58, 1]. The above reports highlighted the important role of the atp-2 gene in regulating mitochondrial function [66, 58]. Thus, we finally detected the expression of atp-2. As shown in Fig. 7a and 7b, RA upregulated the expression of atp-2 in the PD model, but had no change in HD. The results indicated atp-2 participated in the underlying mechanism of RA-induced reduction in mitochondrial ROS and increased in ATP in PD model.
In all, these experiments suggested that RA played an important role in improving the mitochondrial function and structural dysfunction mediated by polyQ and ɑ-syn, including improving the ATP and MMP levels and reducing ROS production in the HD and PD models, maintaining the normal mitochondrial structure, and maintaining the stable expression of TOM7 in mitochondrial outer membrane and mitochondrial inner membrane complexes. Further, RA enhanced multiple genes to improve mitochondrial dynamics network homeostasis and function in both the HD and PD models. Specifically, RA improved mitochondrial dysfunction through upregulating of phb-2, and combining with up-regulation of mmf-1 and fis-1 and down-regulation of drp-1 and eat-3 in HD nematode models. Second, in PD, RA regulated mitochondrial dynamics via modifying phb-2 gene and combining with the overall up-regulation of fis-1, mmf-1, drp-1 and eat-3, and the expression of atp-2 gene was up-regulated to adjust mitochondrial respiratory chain function.