Diterpenoids and Triterpenoids From Frankincense Are Excellent Anti-psoriatic Agents: An in silico Approach

Psoriasis is a chronic autoimmune disease that affects 2–3% of the global population and requires an effective treatment. Frankincense has been long known for its potent anti-inflammatory activities. In this study, a structural bioinformatics approach was used to evaluate the efficacy of individual active components of frankincense, macrocyclic diterpenoid derivatives (1-27), and boswellic acids (28-46) in the treatment of psoriasis. Initially, major druggable targets of psoriasis were identified. Subsequently, structure-based screening was employed by using three different docking algorithms and scoring functions (MOE, AutoDock Vina, and MVD) for the target fishing of compounds against 18 possible targets of psoriasis. Janus Kinase 1, 2, 3 (JAK 1/2/3), eNOS, iNOS, interleukin-17 (IL-17), and Tumor necrosis factor-α (TNF-α) were identified as the preferred molecular targets for these compounds. This computational analysis reflects that frankincense diterpenoids and triterpenoids can serve as excellent anti-psoriatic agents by targeting major cytokines (TNF-α, IL-17, IL-13, IL-23, and IL-36γ,) exacerbated in psoriasis, and inflammatory pathways particularly JAK1/2/3, eNOS, iNOS, MAPK2, and IFNγ. The results were compared with the reported experimental findings which correlates well with our in-silico verdicts.


INTRODUCTION
Psoriasis is a chronic and most inexplicable autoimmune skin disease that affect 2-3% of the population worldwide (Baliwag et al., 2015). It is characterized by increased propagation of the epidermis that can multiply up to 10 times faster than normal in psoriasis with dilation of dermal capillaries. As underlying cells reach the skin's surface and die, their sheer volume creates red plaques covered with white scales that cause itchy and scaly skin, swelling, pain, and disfiguring skin lesions (MacDonald and Burden, 2007). Psoriasis can occur at any age and equally in men and women. The disease is more common in adults than children. The estimates in children vary between 0.7% (Augustin et al., 2010) in Europe to almost none in Asia (Bø et al., 2008;Chen et al., 2008). The variation in the prevalence of psoriasis has also been linked to geographical locations, as it is less common in countries closer to the Equator (Egypt, Sri-Lanka, Taiwan) as opposed to countries that are further away (Europe, Australia, and North America) (Parisi et al., 2013).
Currently, there are three different types of treatment used to reduce the inflammation and skin irritation/itching, including topical treatments, light therapy, and systemic medications (Winterfield et al., 2005;Gisondi et al., 2017;Golbari et al., 2018). Topical treatments serve as first-line therapies that include use of topical corticosteroids, vitamin D analogs, anthralin, retinoids, and calcineurin inhibitors, and they are frequently prescribed to treat mild to moderate psoriasis. The overuse of corticosteroids causes thinning of the skin. Vitamin D analogs (Calcipotriene, Calcitriol) and anthralin reduce skin cell growth, remove scales, and make skin smoother. These analogs treat mild to moderate psoriasis along with other treatments; however, they promote skin irritation. Similarly, topical retinoids may decrease inflammation, but cause skin irritation and increase sensitivity to sunlight. Moreover, oral retinoids cause risk of birth defects and are not recommended for pregnant and breast-feeding women. Calcineurin inhibitors, particularly tacrolimus and pimecrolimus, also reduce inflammation and plaque accumulation; however, an increased risk of skin cancer and lymphoma is associated with these inhibitors, and they are therefore not recommended for long-term or continuous use (Choi et al., 2017).
Second-or third-line therapies, including phototherapy and systemic therapies, are given to patients with severe psoriasis or treatment-resistant disease. In phototherapy, natural or artificial ultraviolet (UV) light is used to treat mild psoriasis by exposing skin to controlled amounts of natural sunlight, artificial ultraviolet A (UVA), or UV B (UVB) light, either alone or in combination with medications. Exposure to UV in sunlight or artificial light slows skin cell turnover and reduces scaling and inflammation. Daily exposure to small amounts of sunlight may improve psoriasis, but intense sun exposure can worsen symptoms and cause skin damage. Controlled doses of UVB light improves the symptoms of mild to moderate psoriasis; however, it may cause short-term side effects like redness, itching, and dry skin. UVA light penetrates deeper into the skin than UVB, improves skin, and is often used to treat severe psoriasis. Despite this, it can cause short-term side effects, such as nausea, headaches, burning, and itching, or long-term side effects, such as dry/wrinkled skin, freckles, increased sun sensitivity, and increased risk of skin cancer and melanoma (Pardasani et al., 2000).
Patients with severe psoriasis are treated with systemic treatment (including retinoids, methotrexate, and cyclosporine), which are associated with severe side effects. Retinoids may cause lip inflammation and hair loss. Methotrexate helps psoriasis by decreasing the production of skin cells and suppressing inflammation; however, it may also cause stomach upset, loss of appetite, and fatigue. With long-term use, methotrexate can cause liver damage and decreased production of red and white blood cells and platelets. Cyclosporine suppresses the immune system and is similar to methotrexate in effectiveness but can only be taken short term because it may increase risk of infection, cancer, kidney problems, and high blood pressure at high doses or long-term therapy (Hoffman et al., 2016).
The presence of cytokines, dendritic cells, and T lymphocytes in psoriatic lesions has encouraged the development of biologic therapies for psoriasis (Schadler et al., 2019). These therapies include monoclonal antibodies (mAB) against tumor necrosis factor-α (TNF-α) (infliximab, adalimumab, golimumab), interleukin (IL)-12 and IL-23 (ustekinumab), IL-17A (secukinumab and ixekizumab), and inhibitors of TNF-α (etanercept) and phosphodiesterase 4 (apremilast). These drugs are usually used to treat psoriatic patients who have failed to respond to traditional therapy or are associated with psoriatic arthritis. However, these drugs have strong effects on the immune system and may permit life-threatening infections. In particular, people taking these treatments must be screened for tuberculosis. All these treatments are only used to manage the disease at each time it surfaces (Tollefson et al., 2018). Therefore, new and safer chemical agents are urgently required for the effective treatment of psoriasis.
Frankincense is known for its superior anti-inflammatory potential Al-Harrasi et al., 2018a). The active constituents of frankincense, including incensole and several boswellic acid derivatives, suppress the expression of tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and nuclear factor-κβ (NF-κβ) (Moussaieff et al., 2007). Due to our deep interest in exploring medicinal properties of frankincense, a computational pipeline was created to investigate the anti-psoriatic potential of the active components of frankincense. In this study, computational target fishing was applied for the identification of potential druggable molecular targets of psoriasis and the binding potential of cembrenoid diterpenoids and triterpenoids found in several species of Boswellia was scrutinized by in silico reverse molecular docking. The computational analyses reveal the promising binding potential of these compounds with the proteins associated with psoriatic pathways.

MATERIALS AND METHODS
The computational experiments were performed on a Windows 10 workstation with Intel R Core TM i7-7700HQ CPU@2.80GHz processor and 12 GB RAM. For docking, MOE (Molecular Operating Environment), MVD (Molegro Virtual Docker) and ADT Vina (AutoDock Tools Vina) were used. Protein-ligand interactions were visualized on Chimera software (Pettersen et al., 2004).

Selection and Preparation of Ligands for Docking
A set of 46 macrocyclic diterpenoids and triterpenes from different species of Boswellia was chosen from our in-house compound's library for reverse docking analysis. The structures of the selected compounds are shown in Table 2. The 2Dstructure of each ligand was drawn on ChemDraw (https://www. perkinelmer.com/category/chemdraw) and converted into a 3Dformat by Molecular Operating Environment (MOE 2014.09) (Florence et al., 2014) (https://www.chemcomp.com/). Hydrogen atoms were added, partial charges were applied (based on MMFF94x force field) on each structure, and the structures were minimized using the MMFF94x force field (eps = r, Cutoff until the RMS gradient of 0.1 kcal/mol/Å was achieved). The protonation state of each compound was assigned based on the neutral pH. All the compounds were imported into MOE database to be used in docking.

Reverse Molecular Docking by MOE
A reverse docking approach was applied to search the appropriate druggable binders of macrocyclic diterpenoids and triterpenes by using MOE docking suit (Lee et al., 2016;Xu et al., 2018). After the preparation of protein and ligand files, docking was performed by using the Triangle Matcher docking algorithm and London dG scoring function. Induced Fit protocol (implemented in MOE) was applied during docking. The ligand binding residues were selected to define the active/ligand binding site ( Table 1). By default, 30 docked conformations of each ligand were retained after docking. At the end of docking, the interactions of each ligand with the binding residues were visualized by Protein-ligand Interaction Fingerprints (PLIF) setup of MOE and Chimera. Later, conformational sampling was performed, and the best docked pose of each ligand was selected based on the docking score, rank, and binding interactions with the target. PLIF calculates hydrogen bonding, protein-ligand solvent bridging, ionic attraction, surface contact, metal ligation, arene attraction, and protein-ligand solvent interactions between protein and ligand. The 2D interactions of ligands with their targets were visualized by the MOE protein-ligand interaction tool. Subsequently, Chimera was used to depict the 3D view of protein-ligand interactions.

Molegro Virtual Docker
For each docking, protein and ligand structures were imported into the MVD (version 2019.7.0.0) workspace in "pdb" and "mol2" format, respectively (Thomsen and Christensen, 2006). The ligands were prepared by MOE by applying charges and explicit hydrogen, and the valences, bond orders, and protons were thus correct in the ligands. The ligand binding site in the protein was defined by cavity detection algorithm of MVD. Thirty docking runs were performed for each ligand with a maximum iteration of 1,500 and maximum population size of 50, which resulted in 30 docked poses of each ligand after docking.     Later, the compounds were ranked on the basis of MolDock and re-rank scores.

Analysis Measures and Conformational Sampling
To select the most appropriate inhibitor against selected target, a consensus approach was used. The docked library was sorted on the basis of docking scores, and the suitable inhibitor for a particular target was selected when it is ranked among the top 10 ligands of the screened dataset by two out of three programs. The most optimal binding mode of each compound was chosen by conformational sampling. All the docked poses of each compound (generated by each program) were compared, and the binding mode that was analogous in all the docking methods was declared as the probable binding mode.

Re-docking
The robustness of the docking methods was scrutinized by redocking experiments. For this purpose, 10 (IL-17, IL-36γ, PPARγ, MAPK2, JAK1/2/3, TNF-α, iNOS, and eNOS) out of 18 selected targets were chosen due to the presence of co-crystallized ligands (inhibitors) in their PDB structures. All the programs successfully re-produced the X-ray confirmed orientation of each ligand with root mean square deviation (RMSD) of ≤2.5 Å. Additionally, the screening reliability was justified by the ranking position of known inhibitors of all the target, and it was observed that all the positive controls were ranked at the top of the screened library. The results indicate that the used docking methods are reliable in predicting the binding orientation and in the selection of known inhibitors. The re-docking results are summarized in supporting information Table S1.

Molecular Targeting of Cembrane Diterpenes and Boswellia Triterpenes
In the current study, a reverse docking approach was employed to predict the binding potential of selected cembrenoid diterpenoids and Boswellia triterpenoids with the 18 drug targets. For this purpose, three different docking algorithms and scoring functions were used. Those proteins, which were collectively ranked with higher score by all the docking methods, were considered as "best" targets. Based on the docking score, JAK1/2/3, eNOS, and iNOS were proposed as the best targets for all the compound by MOE and ADT vina, while MVD only picked 17/46 and 12/46 compounds as excellent inhibitors for JAK1 and eNOS, respectively. Similarly, ADT Vina ranked JAK1 and eNOS among the top ranked targets for all the compounds, while JAK2, JAK3, and iNOS were retrieved as good binders for 42/46, 40/46, and 39/46 compounds, respectively. MVD demonstrated that 40/46, and 37/46 compounds can efficiently target JAK2 and 3, respectively. While >39, 26, and 50% of the screened library may target JAK1, eNOS, and iNOS, respectively. Thus, JAK1/2/3, eNOS and iNOS were considered as the most appropriate druggable candidates for the selected diand tri-terpenoids. Among the selected cytokines, IL-17 was retrieved as an interesting target since MVD showed that all the compounds may neatly fit at the binding pocket of IL-17, while MOE and ADT Vina suggested that 42/46 and 40/46 compounds can effectively target IL-17, respectively. According to MOE and ADT Vina, 45/46 compounds possess high to moderate binding affinities for TNF-α, while MVD showed that >63% of the screened database holds good inhibitory potential against TNF-α. Therefore, IL-17 and TNF-α were also considered as good targets for these compounds.
Based on the in-silico outcomes, JAK1/2/3, eNOS, iNOS, IL-17, and TNF-α are the most appropriate targets for the purified chemical constituents of frankincense. Moreover, the consensus Frontiers in Chemistry | www.frontiersin.org Frontiers in Chemistry | www.frontiersin.org Frontiers in Chemistry | www.frontiersin.org    results of all the docking methods depicted that MAPK2, PPARγ, IL-13, IL-23, IL-36γ, and IFNγ can serve as possible binding proteins for several di-and tri-terpenes. Meanwhile, IL-1α, IL-1β, NF-κB, IL-22, and STAT3 were identified as the least potential targets for these cembrenoid derivatives. The best inhibitors for each target are given in supporting information Table S2 and the docking scores are tabulated in Tables S3-S5. The predicted binding potential of each compound with the selected target is presented as a heatmap in Figure 1.

JAK1/2/3
Various cytokines are regulated by the JAK-STAT pathway, particularly TH17 signaling, which is important in the pathogenesis of psoriasis. JAK inhibitors are efficacious against psoriasis, alopecia areata, and atopic dermatitis (Ciechanowicz et al., 2019). When docked at the active site of JAK1, all the diand tri-terpenoids were predicted as the potent inhibitors by MOE and ADT vina; however, MVD ranked 17/46 compounds (12, 28-34, 36-38, 40-41, and 43-46) as the best binders of JAK1. Among them, four molecules (6, 30-32) were mutually ranked by all the docking methods among the top 10 ligands; however, interaction analysis suggest that only 30-32 can serve as potent inhibitor of JAK1. The binding modes of 30-32 suggests that these compounds neatly fit at the active site of JAK1 where Lys908, Arg1007, Asn1008, Phe886, His885, and Gly887 particularly stabilizes these compounds through H-bonding. The carboxylic group of 3-α-acetyl-β-Boswellic acid (30) mediates bidentate interactions with the side chains of Asn1008 and Arg1007, while the acetyl moiety interacts with the side chain of Lys908. Similarly, carboxylic group of 3-α-acetyl-11-keto-β-Boswellic acid (31) interacts with the side chains of Asn1008 and Asp1021, whereas the acetyl group mediates H-bonding with the amino group of Gly887. However, the carboxylic group of α-Boswellic acid (32) (3) showed that the acetate moiety of the compound mediates H-bonding with the side chain of Lys882, which is homologous to Lys908 and Lys855 in JAK1 and JAK3, respectively. Lys882 also mediate H-bonding with the acetate groups of 14 and16. The docked view of 20 reflects that this compound has a chance to bind with several residues (Leu932, Tyr931, and Arg980) due to the presence of several polar -OH groups in its structure. The -OH and the carbonyl groups of 23 interact with the side chains of Lys882 and Asp994, respectively. This indicates that Lys882 is a crucial residues in the stabilization of diterpenoids at the active site of JAK2. The carboxylic groups of triterpenoid 28 and 43 were found to be H-bonded with the side chain of Arg980, while the carboxylic moiety of 37 interacts with the side chain of Asn981 via H-bonding. Arg980 and Asn981 of JAK2 are homologous to Arg1007 and Asn1008 of JAK1, respectively. This suggest that these Arg and Asn residues play important roles in the binding of boswellic acids in the active site of JAK1/2.
For JAK3, only 9/46 and 6/46 compounds showed the least inhibitory potential in MVD and ADT vina docking, respectively, while rest of the compounds exhibited higher binding potential. Meanwhile, compounds 16, 20, 24, and 36 were communally ranked at the top of the screened dataset by all three scoring functions. It was observed that compounds 16 and 20 can inhibit both the enzymes. 3-α-acetyl-11-keto-β-boswellic acid (31) is  known to inhibit the activation of STAT3 by inhibiting the phosphorylation of JAK2 and Src. For the dephosphorylation of STAT3, compound 31 induces Src homology region 2 domaincontaining phosphatase 1 (SHP-1) (Kunnumakkara et al., 2009). Our docking results are consistent with these experimental findings. Compound 31 was identified as a good inhibitor of JAK1/2/3 by all the scoring functions. The binding modes of selected hits (16, 20, 24, and 36) showed that the methyl acetate of 16 binds with Arg911 and Cys909, while the -OH group of 20 and the carbonyl group of 24 interacts with the amino group  of Leu905. The binding mode of 11-α-methoxy-β-boswellic acid (31) depicted that this compound did not produce significant binding interactions within the active site of JAK3. The binding modes of best predicted inhibitors in the active site of JAK2/3 are shown in Figure 3.

eNOS and iNOS
eNOS and iNOS are involved in immune response, and they are important enzymes in the pathogenesis of psoriasis (Mittal et al., 2009;Rácz and Prens, 2009). The post docking analysis showed that all the compounds bind at the entrance of the active site of eNOS and iNOS. All the derivatives were ranked as good inhibitors of eNOS and iNOS by MOE. Similarly, ADT vina also showed that all the cembrenoids possess greater inhibitory affinities for eNOS; however, except for 1, 2, 5, 7, 10, 11, and 15, the rest of the compounds exhibits excellent binding tendency for iNOS. Despite this, the MolDock score filtered only few compounds as good binders of eNOS ( Table S2) and 50% of the docked library (2, 3, 8, 12, 13, 18, 21, 26, 29, 31-34, 36-38, and  40-46) for iNOS. Among all, compounds 15-16, 28-29, and 37-38 as well as compounds 8, 21, 36-37, and 40 were of utmost importance for eNOS and iNOS, respectively. It is shown that compound 37 targets both eNOS and iNOS effectively, suggesting it could be the most important inhibitor of eNOS and iNOS. The binding modes of the most active compound 28-29 and 36-37 and 40 in the active site of eNOS and iNOS, respectively, are shown in Figure 4.

IL-17
IL-17 and TNF-α emerged as good targets for most of the compounds by all the docking methods. An IL-17 dimer binds with IL-17RA and forms a protein-protein interface, which is also blocked by recently identified cyclo-/linear-peptide inhibitors (Blauvelt and Chiricozzi, 2018;Brembilla et al., 2018). Based on the CA criterion, compounds 16, 18, 28, 29, and 32 were retrieved as the most promising inhibitors for IL-17. The mode of interactions reveals that 18, 28, 29, and 32 are well accommodated at the antagonist binding site of IL-2 where Gln94, Glu95, and Leu97 play important roles in the stabilization of these compounds (supporting information Figure S1).

TNF-α
TNF-α plays a crucial role in the exacerbation of inflammation in psoriasis. The management of psoriasis is done by inhibiting the production of TNF-α by several inhibitors like adalimumab, certolizumab pegol, etanercept, golimumab, and infliximab. MOE and ADT vina ranked TNF-α as a moderate binder for all the compounds except compound 26 and 30, respectively. While MVD predicted that 63% of compounds can exhibit goodto-moderate inhibition of TNF-α. According to CA, compounds 20, 36, 38, 40, 43, and 45 were regarded as the most effective binders of TNF-α. mediated favorable interactions with the Ser60, Leu120, Gly121, and Tyr151 and displayed excellent to moderate inhibition of TNF-α in silico. The interactions of ligands with TNF-α are shown in Figure S2. Diterpenes and Boswellia Triterpenes   MAPK2, PPAR-γ, IL-13, IL-23, IL-36γ, and IFNγ Several di-and tri terpene derivatives exhibited excellent binding affinities for MAPK2, PPAR-γ, IL-13, IL-23, IL-36γ, and IFNγ in our molecular docking experiments (Table S2).

PPAR-γ
PPAR-γ regulates lipid and glucose metabolism. Recently, the pathophysiological role of PPAR-γ was reported in the inflammatory immune response, and it was identified as a major drug target for the treatment of psoriasis, benign epidermal tumors, and atopic dermatitis (Sertznig and Reichrath, 2011). The docked compounds at the active site of PPAR-γ demonstrated that 9-11, 15-16, 18, 20, and 22 hold greater binding affinities with PPAR-γ. Among the selected hits, compounds 15 and 20 are excellent inhibitors of PPAR-γ. Liu et al. (2013) described that compound 31 down-regulates PPAR-γ2 expression and loss of phenotypic markers of mature human adipocytes and mobilizes lipolysis. MVD and ADT vina placed compound 31 in the list of good inhibitors (Liu et al., 2013).

Interleukin-36γ
IL-36γ binds with its receptor (IL-36RA) to produce immune response and found to be upregulated in psoriatic and dermatitis skin (D'erme et al., 2015;Braegelmann et al., 2016). The diterpenoids and boswellic acid derivatives were docked at the recently identified antagonist binding site of IL-36γ, which is also a binding site for the IL-36γ receptor. According to the docking scores, MOE, MVD, and ADT vina detected IL-36γ as a good target for 35, eleven, and eight compounds, respectively. The compounds 7, 13-16, 20, and 22-24 displayed highest potency against IL-36γ. Interestingly, the diterpenoids displayed more binding potential than triterpenoids for MAPK2, PPAR-γ and IL-36γ.

Interferon γ
IFNγ is involved in the innate and adaptive immunity against viral and some bacterial infections. Elevated level of IFNγ has been observed in the serum of psoriatic patients (Abdallah et al., 2009). MVD showed that all the compounds can target IFNγ-Rα binding site, while MOE and ADT vina retrieved six and 26 molecules as good inhibitors of IFNγ, respectively. The false positives were removed by consensus results, which showed that five compounds (17, 20, 22, 39, and 46) are the most prominent inhibitors of IFNγ.

DISCUSSION
The importance of the immune system in the pathogenesis of psoriasis is revealed by efficacy of several immunosuppressive agents, including cyclosporine, denileukin diftitox, and alefacept. TNF-α is usually elevated in psoriatic lesions; however, overexpression of several other cytokines, such as IL-13, IL-17, IL-22, IL-36γ , and IFN-γ , has also been observed in the skin cells of psoriatic patients. These cytokines mediate the potentiation of keratinocytes on psoriatic inflammation, and the use of antibodies against IL-23, TNF-α, and IL-17 in the treatment of psoriasis has shown a key role of these cytokines in the pathogenesis of psoriasis (Lowes et al., 2014;Harden et al., 2015). Moreover, studies have confirmed that the NF-κB pathway is activated in psoriatic lesions and is downregulated after successful treatment (Wang et al., 2017). Furthermore, several other proteins, including IL-1α/β, eNOS, iNOS, PPAR-γ, MAPK2, JAK1/2/3, and STAT3, were identified as possible novel therapeutic targets, which are overexpressed in psoriasis patients (Mittal et al., 2009;Rácz and Prens, 2009).
It is reported that serratol (1) and incensole acetate (3) inhibit the expression of TNF-α, NF-κB, and IL-1β (Moussaieff et al., 2008). Several derivatives of cembrenoid diterpenoids and boswellic acid are well known for their inhibitory activities against the production of TNF-α (Al-Harrasi et al., 2018a). Our docking results are consistent with those experimental findings. The transcription factor NF-κB orchestrates inflammation and regulates the activity of several proinflammatory genes. NF-κB serves as a crucial mediator in the pathogenesis of psoriasis (Goldminz et al., 2013;Moorchung et al., 2014). Several studies showed that incensole (2) and incensole acetate (3) suppress the activation of NF-κB by inhibiting TAK/TAB-mediated IκB kinase (IKK) activation loop phosphorylation (Berrrrpohl et al., 2007;Moussaieff et al., 2007). It is also reported that 3 did not suppress IκBα phosphorylation in co-stimulated T cells, indicating that the kinase inhibition is neither direct nor does it affect all NF-κB activation pathways. Moreover, acetyl-11-keto-β-boswellic acid (31) and acetyl-α-boswellic acid (33) inhibits NF-κB signaling by specifically inhibiting IκBα kinase (IKK), which is pivotal for the degradation of the NF-κB inhibitor IκB, as well as the phosphorylation of p65, two steps essential for NF-κB activation and the subsequent cytokine expression. Syrovets et al. (2005) proved that acetyl-boswellic acids inhibits LPS-mediated TNFα induction in monocytes by direct interaction with IκB kinase (Syrovets et al., 2005). Our docking results also indicate that these compounds have less of a potential to directly bind with NF-κB.
IL-17 is a pro-inflammatory cytokine produced by Th-17 cell in response to their stimulation by IL-23. IL-17 interacts with its receptor to mediate proinflammatory and allergic responses, and, as a result, production of several cytokines (IL-1β, TGF-β, TNF-α, IL-6, GM-CSF, and G-CSF), chemokines (GRO-α, IL-8, and MCP-1), and prostaglandins (PGE2) occurs via macrophages, keratinocytes, fibroblasts, epithelial, and endothelial cells. Activation of IL-17 signaling is observed in the pathogenesis of various autoimmune disorders, including rheumatoid arthritis, asthma, lupus, allograft rejection, antitumor immunity, psoriasis, and multiple sclerosis, and IL-17 inhibitors are being investigated as possible treatments for these disorders. The monoclonal antibodies (mAB) secukinumab and ixekizumab inhibit IL-17 while brodalumab blocks the IL-17 receptor (IL-17RA). By interfering with IL-17 signaling, these antibodies interrupt the inflammatory cycle of psoriasis. The mAB against IL-23 (ustekinumab) is also used to treat psoriasis by reducing IL-17, and, thus, the IL-23/IL-17 pathway plays a major role in psoriasis. Recently, Stürner et al. (2014) revealed that compound 29 (11-keto-β-Boswellic acid, KBA) reduces the differentiation of human CD4+ T cells to Th17 cells while slightly increasing Th2-and Treg-cell differentiation. Furthermore, KBA reduces the IL-1β-triggered IL-17A release of memory Th17 cells. KBA may affect IL-1β signaling by preventing IL-1Rassociated kinase 1 phosphorylation and subsequently decreasing STAT3 phosphorylation at Ser727, which is required for Th17cell differentiation. The effects of KBA on Th17 differentiation and IL-17A release make the compound a good candidate for potential treatment of Th17-driven diseases. Boswellic acids reduce Th17 differentiation via blockade of IL-1β-mediated IRAK1 signaling (Stürner et al., 2014). Our docking results also confirm that compound 29 (KBA) possesses excellent binding affinity for IL-17.
The cytokine IL-1α is produced by macrophages, neutrophils, and epithelial and endothelial cells, while IL-1β is produced by activated macrophages and mediates inflammatory response after binding with their receptor. Both the cytokines are excessively produced in psoriatic lesions (Tsai and Tsai, 2017). Both the cytokines were depicted as moderate to least binding proteins for this set of compounds. IL-22 is produced by T cells and is involved in the modulation of tissue responses during inflammation and found significantly higher in psoriatic patients (Fujita, 2013;Hofny et al., 2017;Voglis et al., 2018). IL-22 was also depicted as a least possible target for di-and triterpene derivatives.
PPAR-γ regulates lipid and glucose metabolism. Recently, the pathophysiological role of PPAR-γ was reported in the inflammatory immune response and identified as a major drug target for the treatment of psoriasis, benign epidermal tumors, and atopic dermatitis (Sertznig and Reichrath, 2011). Liu et al. (2013) described that AKBA (31) downregulates the expression of PPAR-γ2 and loss of phenotypic markers of mature human adipocytes and mobilizes lipolysis (Liu et al., 2013). In the current study, 31 was recognized as an excellent inhibitor of PPAR-γ by MVD.
In psoriatic patients, STAT3 is activated in lesional keratinocytes, which can lead to the development of regenerative epidermal phenotype observed in psoriasis (Sano et al., 2005). Incensole (2) is a potent inhibitor of STAT3 (Pollastro et al., 2016). It is reported that 29 (KBA) inhibits the phosphorylation of STAT3 (Stürner et al., 2014). In this study, both the compounds were identified as moderate inhibitors of STAT3 by MVD. Our insilico experiment correlates well with the reported experimental findings. The docking studies are used in the discovery of novel inhibitors against several therapeutic targets (Halim et al., 2013). However, docking must be combined with more sophisticated computational techniques like molecular dynamic simulation to enhance the efficiency of virtual screening. This work needs to be further explored by molecular dynamic simulation to study the dynamic behavior of protein upon protein-ligand complex formation.

CONCLUSION
The anti-psoriatic potential of several Boswellia diterpenoids and triterpenoids was explored by computational analysis. The compounds have displayed selective docking to human anti-inflammatory (JAK1/2/3, eNOS, iNOS, and TNF-α) and anti-psoriatic (IL-17) molecular targets. Moreover, several other proteins (MAPK2, PPAR-γ, IL-13, IL-23, IL-36γ, and IFNγ) were retrieved as leading biological targets for these compounds. Thus, these compounds have reflected excellent anti-psoriatic and anti-inflammatory potency in silico. The results indicate that Boswellia di-and tri-terpenoids can serve as promising chemical scaffolds for the development and improvement of inhibitors to treat psoriasis.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/Supplementary Materials.

AUTHOR CONTRIBUTIONS
AA-H, RC, and AA-R conceived and designed the study. SH performed all computational studies and analyzed the data. SH and AK wrote the manuscript with inputs and comments from all co-authors. All authors have read and approved the final version of the manuscript.

FUNDING
This project was supported by grant from The Oman Research Council (TRC) through the funded project (BFP/RGP/HSS/19/198).