Synthesis and Biological Evaluation of Diversified Hamigeran B Analogs as Neuroinflammatory Inhibitors and Neurite Outgrowth Stimulators

We describe the efficient synthesis of a series of new simplified hamigeran B and 1-hydroxy-9-epi-hamigeran B norditerpenoid analogs (23 new members in all), structurally related to cyathane diterpenoid scaffold, and their anti-neuroinflammatory and neurite outgrowth-stimulating (neurotrophic) activity. Compounds 9a, 9h, 9o, and 9q exhibited moderate nerve growth factor (NGF)-mediated neurite-outgrowth promoting effects in PC-12 cells at the concentration of 20 μm. Compounds 9b, 9c, 9o, 9q, and 9t showed significant nitric oxide (NO) production inhibition in lipopolysaccharide (LPS)-activated BV-2 microglial cells, of which 9c and 9q were the most potent inhibitors, with IC50 values of 5.85 and 6.31 μm, respectively. Two derivatives 9q and 9o as bifunctional agents displayed good activities as NO production inhibitors and neurite outgrowth-inducers. Cytotoxicity experiments, H2O2-induced oxidative injury assay, and ELISA reaction speculated that compounds may inhibit the TNF-α pathway to achieve anti-inflammatory effects on nerve cells. Moreover, molecular docking studies provided a better understanding of the key structural features affecting the anti-neuroinflammatory activity and displayed significant binding interactions of some derivatives (like 9c, 9q) with the active site of iNOS protein. The structure-activity relationships (SARs) were also discussed. These results demonstrated that this structural class compounds offered an opportunity for the development of a new class of NO inhibitors and NGF-like promotors.


Introduction
In recent years, the prevalence of neurodegenerative disorders is increasing [1][2][3][4]. Dementias are responsible for the greatest burden of neurodegenerative diseases. For example, it is estimated that there are currently almost 47 million people worldwide living with dementia. Alzheimer's disease (AD) is the most common type of dementia and it causes aberrant synapse pruning in neurologic disorders. Individuals from the aging population are suffering from AD [5][6][7][8], without any available disease-modifying treatments to prevent or treat cognitive deficits associated [9][10][11]. There has been an explosion of new findings in the nervous system, from contributions to migration of cells and synapse elimination during development to detrimental damage of nerve cells in autoimmunity and aberrant synapse pruning in neurologic disorders [12][13][14]. Meanwhile, studies suggest that the promotion of an anti-inflammatory response may slow or prevent diseases [15]. Therefore, Mar. Drugs 2020, 18 there is a continuous need to search for high potent compounds as primary sources of medicines to optimize drug discovery and to develop more effective therapies. On the other hand, nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophins like NT-3, NT-4, NT-5, have attracted attention as potential therapeutics for severe neurodegenerative diseases such as AD or as regeneration-promoting compounds [16]. However, their protein properties such as complex structures and high molecular weight eliminated them for using as clinical medication. Hence, the search for small molecules with the NGF biological functions or the NGF-induced neurotrophic activity arouses high interest [17]. The hamigeran diterpenoids, possessing a unique [5,6,6] or [5,6,7] fused tricyclic framework (Figure 1), also known as A-B-C rings, are structurally intriguing, marine-derived natural products. Among them, hamigerans A and B (1 and 2), discovered in 2000 from the poecilosclerid sponge, Hamigera tarangaensis [18], share the first naturally occurring [5,6,6] skeleton. Interestingly, the cyathanes comprise a structurally diverse class of diterpenoids with over 100 members reported from higher fungi thus far [19], featuring a unique [5,6,7] tricarbocyclic scaffold (Figure 1). Except that the seven-membered ring in cyathane scaffold is replaced by a benzene ring in hamigerans, they all have tricarbocyclic ring system, multiple stereogenic centers, and are punctuated by carbons at a variety of oxidation states. Very recently, our group discovered a series of novel natural cyathane diterpenoids with neurotrophic and anti-neuroinflammatory effects from higher basidiomycetes such as Cyathus africanus, Sarcodon scabrosus [20,21], such as sarcodonin G (3), striatoid B (4), cyafricanin C (5), and allocyathin B 2 (6). The two small families of natural products have aroused considerable interest from the research communities of natural products, pharmacology, and synthetic chemistry because of their unique structures with intriguing biological potential [22][23][24][25]. Many synthetic endeavors have been devoted to the synthesis of hamigeran B (2), and our group achieved the total synthesis of hamigeran B in 2018 [23][24][25][26][27]. However, the novel pharmacological properties of hamigeran B analogs have not been assessed in anti-neuroinflammatory and neurotrophic activity so far. Because of the similarity of the ABC ring scaffold in their structures, we speculated the hamigeran B derivatives may also possess analogous neurological activity to the cyathanes. As a result, the substantial interest in the synthesis has resulted in the biological investigation on neurite-outgrowth stimulation and anti-neuroinflammatory activity. Herein, we present a concise synthesis of simplified hamigeran B and 1-hydroxy-9-epi-hamigeran B analogs and biological evaluation thereof.
Mar. Drugs 2020, 18, x FOR PEER REVIEW 2 of 23 continuous need to search for high potent compounds as primary sources of medicines to optimize drug discovery and to develop more effective therapies. On the other hand, nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophins like NT-3, NT-4, NT-5, have attracted attention as potential therapeutics for severe neurodegenerative diseases such as AD or as regeneration-promoting compounds [16]. However, their protein properties such as complex structures and high molecular weight eliminated them for using as clinical medication. Hence, the search for small molecules with the NGF biological functions or the NGF-induced neurotrophic activity arouses high interest [17]. The hamigeran diterpenoids, possessing a unique [5,6,6] or [5,6,7] fused tricyclic framework (Figure 1), also known as A-B-C rings, are structurally intriguing, marine-derived natural products. Among them, hamigerans A and B (1 and 2), discovered in 2000 from the poecilosclerid sponge, Hamigera tarangaensis [18], share the first naturally occurring [5,6,6] skeleton. Interestingly, the cyathanes comprise a structurally diverse class of diterpenoids with over 100 members reported from higher fungi thus far [19], featuring a unique [5,6,7] tricarbocyclic scaffold (Figure 1). Except that the seven-membered ring in cyathane scaffold is replaced by a benzene ring in hamigerans, they all have tricarbocyclic ring system, multiple stereogenic centers, and are punctuated by carbons at a variety of oxidation states. Very recently, our group discovered a series of novel natural cyathane diterpenoids with neurotrophic and anti-neuroinflammatory effects from higher basidiomycetes such as Cyathus africanus, Sarcodon scabrosus [20,21], such as sarcodonin G (3), striatoid B (4), cyafricanin C (5), and allocyathin B2 (6). The two small families of natural products have aroused considerable interest from the research communities of natural products, pharmacology, and synthetic chemistry because of their unique structures with intriguing biological potential [22][23][24][25]. Many synthetic endeavors have been devoted to the synthesis of hamigeran B (2), and our group achieved the total synthesis of hamigeran B in 2018 [23][24][25][26][27]. However, the novel pharmacological properties of hamigeran B analogs have not been assessed in anti-neuroinflammatory and neurotrophic activity so far. Because of the similarity of the ABC ring scaffold in their structures, we speculated the hamigeran B derivatives may also possess analogous neurological activity to the cyathanes. As a result, the substantial interest in the synthesis has resulted in the biological investigation on neurite-outgrowth stimulation and anti-neuroinflammatory activity. Herein, we present a concise synthesis of simplified hamigeran B and 1-hydroxy-9-epi-hamigeran B analogs and biological evaluation thereof.

Chemistry
The synthetic route for the synthesis of the hamigeran B skeleton is illustrated in Scheme 1. The synthesis of diol 9a commenced with preparation of chiral phenolic compound 7 starting from precursor S1 which was prepared from a readily affordable 2-methyl-1,3-cyclopentanedione over six steps with 13% total yield, according to the previously reported protocol [26]. The compound S1 was treated by OsO 4 and NaIO 4 to give the aldehyde 7 a 70% yield, which was unstable in air [26]. The tricyclic compound 8 is envisioned as being constructed through a key intramolecular Friedel-Crafts cyclization of aldehyde 7 with 70% yield. The exclusively high regioselectivity may be attributed to the formation of a coordination intermediate, which fixes the nucleophilic site at the ortho position to the phenolic OH. The compound 8 was treated with PCC to give the oxidative product 9b in 70% yield. After obtaining the dehydrated intermediate 10, the regioselective dihydroxylation of the double bond in the central B ring of 10 with OsO 4 and 4-methylmorpholine N-oxide (NMO) proceeded anti to the angular methyl group to give the diol 9a in 60% yield.

Chemistry
The synthetic route for the synthesis of the hamigeran B skeleton is illustrated in Scheme 1. The synthesis of diol 9a commenced with preparation of chiral phenolic compound 7 starting from precursor S1 which was prepared from a readily affordable 2-methyl-1,3-cyclopentanedione over six steps with 13% total yield, according to the previously reported protocol [26]. The compound S1 was treated by OsO4 and NaIO4 to give the aldehyde 7 a 70% yield, which was unstable in air [26]. The tricyclic compound 8 is envisioned as being constructed through a key intramolecular Friedel-Crafts cyclization of aldehyde 7 with 70% yield. The exclusively high regioselectivity may be attributed to the formation of a coordination intermediate, which fixes the nucleophilic site at the ortho position to the phenolic OH. The compound 8 was treated with PCC to give the oxidative product 9b in 70% yield. After obtaining the dehydrated intermediate 10, the regioselective dihydroxylation of the double bond in the central B ring of 10 with OsO4 and 4-methylmorpholine N-oxide (NMO) proceeded anti to the angular methyl group to give the diol 9a in 60% yield. The synthetic route for the synthesis of 1-hydroxyl-9-epi-hamigeran B intermediate is illustrated in Scheme 2. The synthesis of intermediate diol 9p, began with the known chiral cyclopentaneone 11 as the previous study [28]. Using imidazole and TBSCl to treat compound 11 afforded the α-hydroxy protected compound 12. Isopropylation of the compound 12 with i-PrI followed by a base enolization with LiHMDS and Comins' reagent (PhNTf2 is comins' reagent, named after a person, Daniel Comins) provided enol triflate 13 in two steps with 70% yield. For the coupled product 14, we preferred to examine the Suzuki cross-coupling of 13 with pinacol boronate instead of the corresponding boronic acid that was used previously in Trost's synthesis [29]. Oxidation of the olefinic 14 was conducted by using a combination of OsO4 and NaIO4 to give 90% yield of the aldehyde 15, which was unstable in the air [26]. The compound 16 was envisioned as being constructed via an intramolecular Friedel-Crafts cyclization as the key reaction of 15 followed by protection of the phenol hydroxyl with MeI. We are failed to obtain the pure 16 as the by-product has the similar polarity. To our delight, the by-product can be removed next step. The compound 16 was affected by PCC to give the oxidative product 9o. A couple of enantiomers with ipsilateral dihydroxyl obtained by regioselective dihydroxylation of TBS protected 17 are difficult to purify. Therefore, the compound 17 were constructed through removing the protective group followed by dehydration. The regioselective dihydroxylation of the double bond in the B ring of 17 with OsO4 and NMO anti to the angular methyl group afforded the intermediate diol 9p. We have established a concise route for the synthesis of 1-hydroxyl-9-epi-hamigeran B cores 9p and 9o that proceeds via an eight-step sequence from the known ketone 11. To the best of our knowledge, this synthetic route represents The synthetic route for the synthesis of 1-hydroxyl-9-epi-hamigeran B intermediate is illustrated in Scheme 2. The synthesis of intermediate diol 9p, began with the known chiral cyclopentaneone 11 as the previous study [28]. Using imidazole and TBSCl to treat compound 11 afforded the α-hydroxy protected compound 12. Isopropylation of the compound 12 with i-PrI followed by a base enolization with LiHMDS and Comins' reagent (PhNTf 2 is comins' reagent, named after a person, Daniel Comins) provided enol triflate 13 in two steps with 70% yield. For the coupled product 14, we preferred to examine the Suzuki cross-coupling of 13 with pinacol boronate instead of the corresponding boronic acid that was used previously in Trost's synthesis [29]. Oxidation of the olefinic 14 was conducted by using a combination of OsO 4 and NaIO 4 to give 90% yield of the aldehyde 15, which was unstable in the air [26]. The compound 16 was envisioned as being constructed via an intramolecular Friedel-Crafts cyclization as the key reaction of 15 followed by protection of the phenol hydroxyl with MeI. We are failed to obtain the pure 16 as the by-product has the similar polarity. To our delight, the by-product can be removed next step. The compound 16 was affected by PCC to give the oxidative product 9o. A couple of enantiomers with ipsilateral dihydroxyl obtained by regioselective dihydroxylation of TBS protected 17 are difficult to purify. Therefore, the compound 17 were constructed through removing the protective group followed by dehydration. The regioselective dihydroxylation of the double bond in the B ring of 17 with OsO 4 and NMO anti to the angular methyl group afforded the intermediate diol 9p. We have established a concise route for the synthesis of 1-hydroxyl-9-epi-hamigeran B cores 9p and 9o that proceeds via an eight-step sequence from the known ketone 11. To the best of our knowledge, this synthetic route represents the first time that can accomplish the synthesis of 1-hydroxyl-9-epi-hamigeran B intermediates within 10 steps from a known compound. Moreover, most of the transformations employed herein could be performed on multigram scale. This makes our synthetic route highly useful for mass production. The hamigeran B skeleton-based derivatives 9b-n and 1-hydroxyl-9-epi-hamigeran B-based 9ow obtained ( Figure 2) were prepared according to the methods in the section of Materials and Methods. Protecting the hydroxyl of 9a afforded the corresponding compound 9h-n. Hydrogenation of 9h and 9g proceeded very efficiently to provide the hydrogenated product 9c and 9e. The compounds 9h, 9m, and 9n were constructed via esterification from 9a. Hydrolysis and concomitant oxidation proceeded 9c to produce 9d, followed by the condensation which afforded 9g. Using trifluoroacetic acid to treat compound 9e afforded the product 9f. Compounds 9q and 9s were constructed via removing methyl on the phenolic group of 9o and 9r. The compound 9t was established through simply eliminating 1-Hydroxy of 9s. Protecting the hydroxyl of 9p afforded the corresponding compound 9u-w. There are also six annellation compounds (9e, 9g, 9i-j, 9l, 9w) and two trisubstituted ester molecules (9m bearing Boc-alaester and 9n bearing nitrobenzoate). After chromatographic purification, all compounds including 9a-w were subjected to spectroscopic analysis, and their structures and purity (≥95%) for biological analysis were confirmed by means of HRESIMS and 1 H and 13 C NMR analysis. The hamigeran B skeleton-based derivatives 9b-n and 1-hydroxyl-9-epi-hamigeran B-based 9o-w obtained ( Figure 2) were prepared according to the methods in the section of Materials and Methods. Protecting the hydroxyl of 9a afforded the corresponding compound 9h-n. Hydrogenation of 9h and 9g proceeded very efficiently to provide the hydrogenated product 9c and 9e. The compounds 9h, 9m, and 9n were constructed via esterification from 9a. Hydrolysis and concomitant oxidation proceeded 9c to produce 9d, followed by the condensation which afforded 9g. Using trifluoroacetic acid to treat compound 9e afforded the product 9f. Compounds 9q and 9s were constructed via removing methyl on the phenolic group of 9o and 9r. The compound 9t was established through simply eliminating 1-Hydroxy of 9s. Protecting the hydroxyl of 9p afforded the corresponding compound 9u-w. There are also six annellation compounds (9e, 9g, 9i-j, 9l, 9w) and two trisubstituted ester molecules (9m bearing Boc-alaester and 9n bearing nitrobenzoate). After chromatographic purification, all compounds including 9a-w were subjected to spectroscopic analysis, and their structures and purity (≥95%) for biological analysis were confirmed by means of HRESIMS and 1 H and 13 C NMR analysis.

Biological Evaluation
According to previously reported methods [30,31], we used rat pheochromocytoma PC12 cells as a model system of neuronal differentiation to measure the effects of the synthesized compounds hamigeran B (2), 9a-n and 9o-w on neurite outgrowth, with NGF (10 ng/mL) as positive control. Specifically, as shown in Figure 3, most of them in combination with NGF (10 ng/mL) showed NGF-mediated neurite-outgrowth promoting activities, and 9a, 9h, 9o, and 9q exhibited significant neuritogenic effects with the percentage of neurite-bearing PC12 cells of 15.94%, 14.99%, 24.23%, and 16.68%, respectively, compared to NGF control cells (9.35%) and the parent molecule 2 (11.34%). Among them, 9o, possessing an ether bond, and in particular large bulky TBS and OMe groups, was the most potent neurotrophic NGF-like inducer.

Biological Evaluation
According to previously reported methods [30,31], we used rat pheochromocytoma PC12 cells as a model system of neuronal differentiation to measure the effects of the synthesized compounds hamigeran B (2), 9a-n and 9o-w on neurite outgrowth, with NGF (10 ng/mL) as positive control. Specifically, as shown in Figure 3, most of them in combination with NGF (10 ng/mL) showed NGFmediated neurite-outgrowth promoting activities, and 9a, 9h, 9o, and 9q exhibited significant methods [32,33]. The results of inhibitory effects are depicted in Figure 4. As a result, most of the tested compounds showed inhibitory effects on NO production, and 9b-c, 9f, 9o, 9q, and 9t exerted significant effects, with IC50 values in the approximate range 5.8-24 μm. Among them, the most potent inhibitors, 9c (IC50 = 5.85 μm) and 9q (IC50 = 6.31 μm) showed a comparable inhibition potency to natural product quercetin (IC50 = 4.3 μm). The most important thing is that they can be cheaply stored and transported as the less deliquescent than quercetin. Cell with one or more neurites whose lengths were at least twice the diameter of the cell body were counted as neurite-bearing cells. The purple bars mean statistically significant compounds. In all panels, error bars indicate ± SD (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 compared with the nerve growth factor (NGF) group. Cell with one or more neurites whose lengths were at least twice the diameter of the cell body were counted as neurite-bearing cells. The purple bars mean statistically significant compounds. In all panels, error bars indicate ± SD (n = 3). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 compared with the nerve growth factor (NGF) group.
The inhibitory effects of the hamigeran B derivatives (9a-n) and 1-hydroxyl-9-epi-hamigeran B ones (9o-w) on LPS-stimulated NO production in BV2 cells were assessed according to our reported methods [32,33]. The results of inhibitory effects are depicted in Figure 4. As a result, most of the tested compounds showed inhibitory effects on NO production, and 9b-c, 9f, 9o, 9q, and 9t exerted significant effects, with IC50 values in the approximate range 5.8-24 µm. Among them, the most potent inhibitors, 9c (IC 50 = 5.85 µm) and 9q (IC 50 = 6.31 µm) showed a comparable inhibition potency to natural product quercetin (IC 50 = 4.3 µm). The most important thing is that they can be cheaply stored and transported as the less deliquescent than quercetin. The effect on NO production could technically be due to toxicity on the cells. To exclude the possibility that their inhibitory activity was simply due to the cytotoxicity of the tested compounds, a cytotoxicity assay was conducted in BV-2 cells. As shown in Figure 5, when compared to the vehicle control at 10 μm, all 16 compounds failed to affect cell viability significantly. While the compounds display mild toxicity, the toxicity cannot account for the other observed effects. It is further proved that 16 compounds have certain anti-neuritis activity, which has nothing to do with cytotoxic activity.
The accumulation of H2O2 can lead to amyloid production, dopamine oxidation, and cerebral ischemia. Therefore, the formed H2O2 can be easily converted into highly toxic hydroxyl radicals generated by Fenton's chemical method to damage lipids, proteins, and DNA. The oxidative damage may lead to mitochondrial dysfunction, calcium imbalance, inflammation, and apoptosis of nerve cells [34]. At a concentration of 10 μm, compound 9o has better protective activity in the oxidative stress response of PC12 cells caused by H2O2 and other compounds have weak activity in Figure 6.
TNF-α is a cytokine produced in response to infection and is closely related to the NF-κB pathway in the inflammatory response [35]. Based on the above experiments, in order to further prove the anti-neuritis activity of 9b, 9c, 9o, and 9q, we used enzyme-linked immune response (ELISA) to determine the secretion of TNF-α in BV-2 cells stimulated by LPS under the action of the above compounds. When BV-2 were treated with different concentrations of 9b, 9c, 9o, and 9q, TNF-α was reduced to varying degrees in a dose-dependent manner in Figure 7. The effect on NO production could technically be due to toxicity on the cells. To exclude the possibility that their inhibitory activity was simply due to the cytotoxicity of the tested compounds, a cytotoxicity assay was conducted in BV-2 cells. As shown in Figure 5, when compared to the vehicle control at 10 µm, all 16 compounds failed to affect cell viability significantly. While the compounds display mild toxicity, the toxicity cannot account for the other observed effects. It is further proved that 16 compounds have certain anti-neuritis activity, which has nothing to do with cytotoxic activity.
Mar. Drugs 2020, 18, x FOR PEER REVIEW 8 of 23 The accumulation of H 2 O 2 can lead to amyloid production, dopamine oxidation, and cerebral ischemia. Therefore, the formed H 2 O 2 can be easily converted into highly toxic hydroxyl radicals generated by Fenton's chemical method to damage lipids, proteins, and DNA. The oxidative damage may lead to mitochondrial dysfunction, calcium imbalance, inflammation, and apoptosis of nerve cells [34]. At a concentration of 10 µm, compound 9o has better protective activity in the oxidative stress response of PC12 cells caused by H 2 O 2 and other compounds have weak activity in Figure 6.  TNF-α is a cytokine produced in response to infection and is closely related to the NF-κB pathway in the inflammatory response [35]. Based on the above experiments, in order to further prove the anti-neuritis activity of 9b, 9c, 9o, and 9q, we used enzyme-linked immune response (ELISA) to determine the secretion of TNF-α in BV-2 cells stimulated by LPS under the action of the above compounds. When BV-2 were treated with different concentrations of 9b, 9c, 9o, and 9q, TNF-α was reduced to varying degrees in a dose-dependent manner in Figure 7.

Discussion
Based on the NO inhibition profile of these hamigeran B analogs (Figure 4), the structure-activity relationships (SARs) are summarized in Figure 8: (1) For whether hamigeran B or 9-epi-hamigeran B series, the [5,6,6]-tricarbocyclic system (ABC ring) is essential to activity, and introduction of an N-or O-heterocyclic ring (D ring) at C-7/C-8, or acylation or ether bond of three hydroxyl groups in 9a were detrimental to or led to a loss in activity (9e, 9g, 9k, 9w); (2) for hamigeran B series, hydrogenation of the double bond at C-3/C-4 is crucial to improved activity (9c IC 50 = 5.85 µm) relative to 9h, hydrogenation of the keto group at C-8 enhanced activity (9f, IC 50 = 24.91 µm vs. 9d, IC 50 > 50 µm); for 9-epi-hamigeran B series, the presence of the phenol hydroxyl at C14 was important for activity (9q, IC 50 = 6.31 µm vs. 9o, IC 50 = 11.9 µm, the IC 50 concentration curves in Supplementary Materials), and introduction of a large steric hindrance group (a sterically hindered group TBS), that is, TBS, at C-1 was able to highly increase activity (9o) compared to 9r; the large conjugated system with benzene-ring was important for enhancing the activity, for example, 9t (IC 50 = 11.98 µm) with a conjugated cyclpentan-1,3-diene unit was far more active in comparison with 9s.
Mar. Drugs 2020, 18, x FOR PEER REVIEW 10 of 23 and the carboxide of acetate group at C-7 position facilitated hydrogen bonding interaction with GLN-263 (binding-length: 1.948 Å). The carboxide at the C-7 position in 9o was favorably placed in the pocket facilitating hydrogen bonding interaction with TYR-373 (binding-length: 3.548 Å). Intriguingly, the protected α-hydroxy with TBS in 9q facilitated hydrogen bonding with GLN-263 (binding-length: 1.996 Å and 3.311 Å). The results showed that the four compounds are stabilized by different and significant binding interactions (i.e., hydrogen bonding). These interactions were determined between the synthesized compounds and the active site of the iNOS enzyme model; furthermore, the calculated binding-lengths were in good agreement with the experimental IC50 values. The two most active compounds, that is, 9c and 9q, in these series exhibited several strong interactions with the active site of the enzyme (Figure 9).   Cytotoxicity experiments showed that 16 compounds had no cytotoxic activity, proving that anti-neuritis activity was not caused by cytotoxicity. Therefore, compounds 9b, 9c, 9o, and 9q have no neuroprotective activity under the action of oxidative stress, further indicating that the 4 compounds have good anti-neuritis activity. Using ELISA reaction, it was found that 9b, 9c, 9o, and 9q and 4 compounds all inhibited the secretion of TNF-α in LPS activated BV-2 cells. Among them, the inhibitory effect of 9b was particularly obvious, and the tumor necrosis factor (TNF-α) was down-regulated in a dose-dependent manner. TNF-α is an important factor in the NF-κB inflammation pathway, and it is speculated that compounds may inhibit the pathway to achieve anti-inflammatory effects on nerve cells.
In order to investigate the putative binding mode of the potent inhibitors 9b, 9c, 9o, and 9q on the inducible nitric oxide synthase (iNOS), a molecular modeling study was carried out. As a result, the binding mode predicted for 9b, 9c, 9o, and 9q in the docking studies is shown in Figure 9, the phenolic hydroxyl group at the C-14 position in 9b facilitated hydrogen bonding interaction with TYR-373 (binding-length: 2.124 Å) and ASP-382 (binding-length: 2.079 Å and 2.325 Å). The carboxide of acetate group at the C-8 position in 9c was favorably placed in the pocket facilitating hydrogen bonding interaction with TRP-346 (binding-length: 3.304 Å) and TYR-373 (binding-length: 1.984 Å), and the carboxide of acetate group at C-7 position facilitated hydrogen bonding interaction with GLN-263 (binding-length: 1.948 Å). The carboxide at the C-7 position in 9o was favorably placed in the pocket facilitating hydrogen bonding interaction with TYR-373 (binding-length: 3.548 Å). Intriguingly, the protected α-hydroxy with TBS in 9q facilitated hydrogen bonding with GLN-263 (binding-length: 1.996 Å and 3.311 Å). The results showed that the four compounds are stabilized by different and significant binding interactions (i.e., hydrogen bonding). These interactions were determined between the synthesized compounds and the active site of the iNOS enzyme model; furthermore, the calculated binding-lengths were in good agreement with the experimental IC 50 values. The two most active compounds, that is, 9c and 9q, in these series exhibited several strong interactions with the active site of the enzyme (Figure 9).

Cell Viability Assay
BV-2 cells were seeded into 96-well plates at 2 × 10 4 cells/well. After 24 h, drugs were added and incubated for another 24 h. 10 µL CCK8 (APExBIO, Houston, TX, USA) was added to each well. After 4 h, the microplate reader measured absorbance at 450 nm.

Neuroprotective Effects against H 2 O 2 -Induced Oxidative Injury in PC12 Cells
PC12 cells were maintained in DMEM supplemented with 10% heat-inactivated horse serum and 5% fetal bovine serum in humidified 5% CO 2 /95% air at 37 • C. All cells were cultured in collagen coated culture dishes or flasks. The medium was changed every other day. Before treatment, cells were plated at an appropriate density on culture plates or dishes according to each experimental scale and cultured for 24 h. Cells were pretreated for 6 h with various concentrations of samples. Then, the medium was refreshed without adding sample, and the cells were exposed to 400 µm H 2 O 2 for another 16 h [34].

Cytokine Release Was Measured Using an ELISA Kit
BV-2 microglial cells were cultured in a 24-well plate at 2 × 10 5 cells/mL and 500 µL for 24 h. Then stimulated with the control group, model group, and administration group for 24 h. The supernatant was collected from the culture, and supernatant was used to measure the concentrations of TNF-α using an ELISA kit (Boster Biological Technology, Pleasanton, CA, USA) [35].

Molecular Docking Studies
Molecular docking simulations were performed using the software SYBYL-X 2.0 as the previous studies [45]. We performed a re-docking of the extracted ligand present in the iNOS complex. The similarity range were 0.3-0.6 (<0.8), which indicated the docking protocol was feasible. The three-dimensional (3D) crystal structure of iNOS (PDB code: 3E7G) was obtained from the RCSB Protein Data Bank.

Conclusions
In summary, we have enantioselectively synthesized for the first time hamigeran B norditerpenoid analogs with high yield. This methodology provides a route for the rapid assembly of the hamigeran pool in a practical fashion and sufficient quantities to facilitate drug discovery. In this work, the neurotrophic and anti-neuroinfammatory activities of hamigeran B analogs were first investigated. Cytotoxicity experiments, H 2 O 2 -induced oxidative injury assay, and ELISA reaction speculated that compounds may inhibit the TNF-α pathway to achieve anti-inflammatory effects on nerve cells. The SARs outlined provided an insight into the interactions between iNOS and a new class of the most potent NO inhibitors (9c, 9q) to facilitate the further structural modification of this compound class, and two derivatives 9q and 9o have potential as dual-role therapeutic agents for AD treatment. Our investigations would expand and enrich the chemical and biological diversity of hamigeran-like norditerpenoid analogs.

Conflicts of Interest:
The authors declare no conflict of interest.