English

• 1. Introduction

  In 2024, the U.S. Food and Drug Administration (FDA) approved a total of 50 drug applications, including 19 biologics and 31 novel molecular entities [1,2]. Notably, the recently approved small molecule drugs are widely used in various therapeutic areas such as infectious diseases, central nervous system (CNS) disorders, genetic ailments, and malignancies. Many of these approved small molecule drugs are optimized based on the structure of existing drugs. For example, introducing a methyl group into triazole based on the structure of tazobactam results in enmetazobactam with significantly improved antibacterial activity [3]. Ceftobiprole medocaril sodium, with a wider antibacterial spectrum, is obtained by optimizing the substituents of ceftaroline [4]. Sulopenem is transformed into sulopenem etzadroxil by introducing ester bond to improve metabolic stability and drug administration [5]. Sofpironium bromide replaces the methyl group of glycopyrronium bromide with an ester group to reduce clinical side effects through improved absorption and metabolism [6]. Deuruxolitinib deuterates ruxolitinib’s cyclopentane, resulting in significant improvement in pharmacokinetic properties [7]. Elafibranor improves specificity and therapeutic efficacy by replacing bezafibrate’s amide linker with a chalcone structure [8]. The examples further validate that utilizing existing drug structures in developing new pharmaceuticals not only speeds up discovery but also enhances hit and success rates in development.

  A comprehensive analysis of approved drug molecules’ research and development experience not only helps discover new skeleton structures and simplify the synthesis steps of active pharmaceutical ingredients (APIs), but also aids in understanding cutting-edge drug development strategies, such as structure-based drug design (SBDD) and computer-aided drug design (CADD) [9-11]. The covalent drugs have gained attention due to their advantages in selectivity and affinity [12]. The innovative technologies like proteolysis targeting chimeras (PROTACs), prodrug strategies, deuterated drug approaches, and pharmacokinetics-driven drug design show great potential in reducing resistance, enhancing specificity, and minimizing toxicity [13-15]. The breakthroughs in artificial intelligence (AI) and deep learning have created opportunities at the intersection of machine learning and drug discovery, applied to protein structure prediction, drug-target interaction prediction, and drug synthesis route design [16,17]. As stated in our previous review, the approved drugs are valuable resources for advanced drug discovery ideas [18-20]. Understanding the design concept and binding model of recently marketed drugs can enhance the efficiency of me-too drugs’ research and development while expediting the identification of novel compounds within the same therapeutic class with improved selectivity, efficacy, and oral bioavailability. Additionally, it can inspire the design of structurally similar pharmacophores targeting alternative molecular targets. Therefore, this review presents a comprehensive analysis of approved small molecule drugs, including drug design, molecular docking, and mechanism of action investigations. The aim is to provide valuable insights for pharmaceutical chemists and expedite the new drug discovery process.

  2. Anti-infective drugs

  2.1 Cefepime and enmetazobactam (Exblifep^Ⓡ)

  In February 2024, Allecra Therapeutics’ antibiotic combination therapy Exblifep^Ⓡ (cefepime/enmetazobactam) received FDA approval for treating complex urinary tract infections (cUTI) [21]. Cefepime, a fourth-generation cephalosporin antibiotic developed by Bristol-Myers Squibb and approved in 1994, effectively penetrates the cell walls of both Gram-positive and Gram-negative bacteria and exerts bactericidal activity by binding to penicillin-binding protein (PBP) targets and inhibiting cell wall synthesis [22]. Enmetazobactam acts as an extended-spectrum β-lactamase (ESBL) inhibitor, safeguarding cefepime against ESBL degradation and preventing the emergence of antibiotic resistance [23]. The synergistic action of different antibiotics targeting distinct cellular structures or processes within bacteria enhances efficacy compared to individual drugs alone, resulting in a potentiated effect. The notable examples include Xacduro^Ⓡ (sulbactam/durlobactam), an approved antibiotic combination last year, as well as Pfizer’s aztreonam/avibactam currently undergoing phase 3 clinical trials.

  Given cefepime’s mechanism comprehensive understanding and its predominant efficacy against drug-resistant bacteria, this study primarily focuses on enmetazobactam, an active compound that effectively reverses drug resistance [24]. Enmetazabactam was initially discovered by Orchid Pharma in 2008 and subsequently licensed to Allecra Therapeutics. Structurally, enmetazobactam is synthesized through the introduction of methyl groups to tazobactam triazoles (Fig. 1A). Upon analyzing the crystal structure data, we found that adding a single methyl group significantly changed how the active molecule binds to the target protein (Fig. 1B). Tazobactam primarily interacted with Ser-91 through its carbonyl group, while its sulfone formed hydrogen bonds with Ala-345, Thr-346, and Gly-347 (PDB code: 6XFS). The carboxyl group did not contribute to any interaction. Enmetazobactam preserved these hydrogen bond interactions of sulfone with Gly-320 and carbonyl with Ser-64 and Ala-318 while also enabling synergistic interaction between the carboxyl group and α-methyl group for enhanced binding mode (PDB code: 6T35). The inhibitory activity of enmetazobactam was enhanced by hydrogen bond interaction and σ-π hyperconjugation with Gln-120 and Tyr-221, improving its overall efficacy. The triazole also increased the biological activity of enmetazobactam through ionic interactions with Asp-123.

  Figure 1

  

  Figure 1.  (A) The potential discovery of enmetazobactam based on the structure of tazobactam. (B) The X-ray crystal structures of β-lactamases bind to tazobactam (PDB code: 6XFS) and enmetazobactam (PDB code: 6T35).

  DownLoad: Full-Size Img PowerPoint

  2.2 Ceftobiprole medocaril sodium (Zevtera^Ⓡ)

  In April 2024, Basilea Pharmaceutica’s Zevtera^Ⓡ (ceftobiprole medocaril sodium) gained FDA approval for treating Staphylococcus aureus bacteremia (SAB), acute bacterial skin and skin structure infection (ABSSSI), and community-acquired bacterial pneumonia (CABP) [25]. Ceftobiprole achieves its bactericidal activity by inhibiting bacterial cell wall synthesis through binding to essential PBPs and suppressing their transpeptidase activity. Zevtera^Ⓡ provides an additional therapeutic option for a wide range of severe bacterial infections.

  Ceftobiprole medocaril was successfully synthesized from ceftaroline by removing the phosphate group and exposing the hydroxylamine moiety, substituting the thiazole with a lactam ring, and introducing a prodrug moiety (Fig. 2A) [26]. The X-ray crystal structure (PDB code: 3ZG0) of ceftaroline revealed that the active ring-opening substrate formed multiple hydrogen bonds with Ser-403, Ser-462, Asn-464, Gly-520, Lys-597, Ser-598 and Thr-600 while also interacting ionically with Glu-447. Additionally, there was a σ-π hyperconjugation with Thr-582 (Fig. 2B). The strong interactions strongly supported its exceptional effectiveness. The active metabolites of ceftobiprole medocaril sodium maintained these interactions, with the tetrahydropyrrole ring system forming ionic interactions with Met-641 and the primary amine weakly interacting with Tyr-519 through water (PDB code: 4DKI). These additional interactions likely contributed to its significantly increased activity.

  Figure 2

  

  Figure 2.  (A) The potential discovery of ceftobiprole medocaril sodium based on the structure of ceftaroline. (B) The X-ray crystal structures of ceftaroline and ceftobiprole medocaril bind to PBP (PDB codes: 3ZG0 and 4DKI).

  DownLoad: Full-Size Img PowerPoint

  2.3 Sulopenem etzadroxil and probenecid (Orlynvah^Ⓡ)

  In October 2024, Orlynvah^Ⓡ (sulopenem etzadroxil and probenecid) was approved by FDA as a treatment for uncomplicated urinary tract infection (uUTI) [27]. Sulopenem etzadroxil is an innovative antibiotic prodrug that can be transformed into the active sulopenem by hydrolysis. It has potent activity against various Gram-negative bacteria, Gram-positive bacteria, and anaerobic bacteria [28]. Sulopenem works by binding to PBPs and inhibiting bacterial cell wall synthesis. Probenecid, a sulfonamide anti-gout medication, promotes uric acid excretion while inhibiting penicillin excretion [29]. Particularly, probenecid competitively inhibits the secretion of weak organic acids (such as penicillin and cephalosporin) in renal tubules, thereby increasing blood concentrations of antibiotics and prolonging their effectiveness. Probenecid also hinders the elimination of sulopenem through organic anion transporter 3 (OAT3), leading to higher plasma concentrations and enhanced antimicrobial effect.

  Sulopenema etzadroxil was an ester prodrug that could be hydrolyzed by enteric esterase to release active sulopenema (Fig. 3A). In terms of skeleton structure, sulopenema etzadroxil shared similarities with other penem drugs like imipenem, meropenem, and tebipenem. To enhance efficacy and metabolism, OAT inhibitors could be obtained through deuterated ethylenediamine modification of probenecid and structural alteration of the carboxyl terminal [30]. The molecular docking analysis (PDB code: 6I1H) revealed that sulfoxide, carboxyl group, and ring-opening carbonyl group in sulopenem formed multiple hydrogen bond interactions with Asn-378, Ser-376, Asp-359, and Thr-309 (Fig. 3B). The X-ray structure analysis (PDB code: 8SDZ) demonstrated diverse hydrogen bond interactions between the carboxyl and sulfonyl groups of probenecid and Lys-382 and Tyr-354.

  Figure 3

  

  Figure 3.  (A) The structures of sulopenem etzadroxil and probenecid. (B) The X-ray crystal structures of sulopenem and probenecid bind to PBP and OAT (PDB codes: 6I1H and 8SDZ).

  DownLoad: Full-Size Img PowerPoint

  3. Antitumor drugs

  3.1 Tovorafenib (Ojemda^Ⓡ)

  In April 2024, the U.S. FDA granted accelerated approval tovorafenib for treating pediatric low-grade glioma (pLGG), a type of brain tumor in children [31]. Tovorafenib is a highly specific inhibitor of pan-Raf kinase with brain permeability. It works by regulating the mitogen-activated protein kinase/extracellular-signal-regulated kinase (MAPK/ERK) pathway to inhibit tumor growth in cases with BRAF fusion or BRAF V600 mutations. Notably, tovorafenib is the first and only FDA-approved treatment option for pediatric brain tumors characterized by BRAF fusion or rearrangement.

  The structural characteristics of Raf inhibitors were identical, and tovorafenib could potentially be derived from naporafenib, enhancing development efficiency (Fig. 4A) [32]. The X-ray crystal structure analysis revealed that naporafenib (PDB code: 8F7O) and tovorafenib (PDB code: 8F7P) had similar binding modes, both engaging in multiple hydrogen bond interactions with Glu-501, Cys-532, and Asp-594 (Fig. 4B). Additionally, the chlorine atom of tovorafenib formed a hydrogen bond interaction with Ile-573, while the newly introduced amide linker established a hydrogen bond interaction with Thr-529. These additional interactions might contribute to its enhanced activity.

  Figure 4

  

  Figure 4.  (A) The potential discovery of tovorafenib. (B) The binding models of naporafenib and tovorafenib with BRAF (PDB codes: 8F7O and 8F7P).

  DownLoad: Full-Size Img PowerPoint

  3.2 Vorasidenib (Voranigo^Ⓡ)

  In August 2024, vorasidenib gained FDA approval for treating grade 2 astrocytoma or oligodendroglioma [33]. Vorasidenib is a dual-target inhibitor of isocitrate dehydrogenase 1/2 (IDH1/2) that effectively reduces in vivo production of 2-hydroxyglutaric acid and partially restores cellular differentiation. Its ability to penetrate the blood-brain barrier makes it valuable for brain tumor treatment. Vorasidenib is the first inhibitor to simultaneously target two mutation sites of IDH1 and IDH2, addressing drug resistance caused by single-target mutations.

  The researchers assessed the inhibitory activity of AGI-12026 against IDH1/2 and observed its favorable brain permeability after removing the hydroxyl group (Fig. 5A) [34]. Subsequently, an extensive screening of triazine compounds led to the discovery of AGI-15056 as a potent inhibitor with high brain exposure. The X-ray crystal structure analysis (PDB code: 6VG0) revealed two potential pseudo-symmetric binding conformations for AGI-15056, ultimately leading to the discovery of vorasidenib through further structure-activity relationship (SAR) studies.

  Figure 5

  

  Figure 5.  (A) The original discovery of vorasidenib. (B) The binding models of enasidenib and vorasidenib with IDH1/2 (PDB codes: 5I96, 6VFZ and 6VEI).

  DownLoad: Full-Size Img PowerPoint

  The X-ray crystal structure analysis (PDB codes: 5I96 and 6VFZ) revealed that enasidenib and vorasidenib had similar binding patterns with IDH2 (Fig. 5B) [35]. In this model, aminotriazine formed multiple hydrogen bonds with Gln-316 and engages in σ-π hyperconjugation with Val-315. Additionally, the amino triazine moiety of vorasidenib established multiple hydrogen bond interactions with Gln-277, while the chlorine atom participated in hydrogen bond with Asp-273.

  3.3 Lazertinib (Lazcluze^Ⓡ)

  The U.S. FDA approved the marketing application for lazertinib, a third-generation epidermal growth factor receptor (EGFR) inhibitor [36]. This drug is used in combination with amivantamab-vmjw to treat non-small cell lung cancer (NSCLC). Amivantamab-vmjw is a humanized bispecific antibody that targets EGFR/c-mesenchymal-epithelial transition factor (c-Met) and has multiple anti-cancer mechanisms, including inhibiting signaling pathways and recruiting immune cells to target tumors with active and drug-resistant EGFR/c-Met mutations and amplification [37]. Lazertinib, an oral tyrosine kinase inhibitor targeting EGFR, effectively crosses the blood-brain barrier and irreversibly inhibits the T790 M mutation while remaining effective against activated Ex19del and L858R mutations.

  The structure of lazertinib likely derived from osimertinib, with the substitution of indole by benzene-pyrazole biaryl and N,N,N’-trimethylethylenediamine by a morpholine ring (Fig. 6A) [38]. The X-ray crystallography analysis (PDB codes: 7JXM and 7UKW) revealed that both osimertinib and lazertinib exhibited similar binding modes, where amino pyrimidine and acrylamide formed hydrogen bond and covalent interactions with Met-793 and Cys-797, respectively (Fig. 6B). Despite their structural disparities, the introduction of morpholine eliminated the hydrogen-bond interaction with Asp-800. However, incorporating dimethylamine at the biaryl site enhanced hydrogen-bond interaction with Asp-855. Consequently, both compounds demonstrated potent inhibitory activity against EGFR.

  Figure 6

  

  Figure 6.  (A) The potential discovery of lazertinib based on the structure of osimertinib. (B) The binding models of osimertinib and lazertinib with EGFR (PDB codes: 7JXM and 7UKW).

  DownLoad: Full-Size Img PowerPoint

  3.4 Inavolisib (Itovebi^Ⓡ)

  The FDA approved inavolisib, an oral tablet, in October 2024 for treating locally advanced or metastatic breast cancer when combined with palbocent and fulvestrant [39]. As a highly selective inhibitor of phosphatidylinositol 3-kinase α (PI3Kα), inavolisib specifically binds to the ATP binding site of PI3Kα, inhibiting its activity and degrading its catalytic subunit p110α. This further suppresses downstream phosphorylation events involving protein kinase B (Akt), reducing cell proliferation and inducing apoptosis for therapeutic effects.

  Inavolisib, a potential best-in-class PI3Kα inhibitor, was under development since 2011. A hit compound called INA-01 was discovered through high-throughput screening (Fig. 7A) [40], featuring a novel parent scaffold. Due to the poor physical and chemical properties of INA-01 and its challenging production process for medicine, researchers screened the amide substitution group and obtained compound INA-02 with improved activity. To address the issue of benzyl site oxidation in thiophene leading to ring opening and difficulty in maintaining the conformation of the parent nucleus, researchers employed a ring expansion strategy resulting in seven-membered ring compound INA-03. Based on findings that revealed further expansion space at the 8-position of benzoxacyclic heptane from eutectic structure analysis, structural modification was performed by researchers leading to compound INA-04. Further optimization led to the synthesis of compound INA-05 (GNE-614). However, due to the various conformations of amide bonds and their tendency for N-demethylation and hydrolysis reactions that produced toxic substances like aniline, researchers replaced the amide structure with a bioelectron isosteric strategy to obtain compound INA-06. By optimizing both the 8-position side chain and replacing the thiophene ring while considering selectivity and metabolic stability aspects, further improvement was achieved in compound INA-07 (taselisib). Finally, by combining the eutectic structure with structural characteristics and optimizing both the triazole moiety and 8-position side chain, drug inavolisib was obtained.

  Figure 7

  

  Figure 7.  (A) The original discovery of inavolisib. (B) The binding model of inavolisib with PI3Kα (PDB code: 8EXV).

  DownLoad: Full-Size Img PowerPoint

  The X-ray structure analysis (PDB code: 8EXV) revealed that oxazolidinone, oxygen atom, aniline, and amide of inavolisib formed hydrogen bond interactions with Tyr-836, Val-851, Ser-854, and Gln-859 respectively (Fig. 7B). The protein pocket binding model suggested focusing on modifying the oxazolidinone and amide components.

  3.5 Revumenib (Revuforj^Ⓡ)

  Revumenib was approved by the FDA in November 2024 for the treatment of relapsed or refractory acute leukemia with lysine methyltransferase 2A (KMT2A) translocation [41]. As a menin inhibitor, revumenib specifically blocks the interaction of wild-type and fusion protein of KMT2A with menin, inhibiting abnormal signal transduction caused by KMT2A gene mutation for therapeutic effect.

  The hit compound with pyrimidine-fused thiophene was initially identified through high-throughput screening and cultured eutectic crystal (PDB code: 4GQ3). Based on the protein structure, researchers optimized the chemical structure using REV-01 (Fig. 8A) [42]. By introducing p-fluoro aniline to the pyrimidine ring in REV-01, a hydrogen bond interaction with Tyr-276 side chain was improved, resulting in compound REV-02. The addition of isopropyl phenyl in the ortho-position of p-fluoro aniline better occupied the hydrophobic pocket. Furthermore, replacing piperazine with a nitrogen-containing spiroid to form a hydrogen bond with a distal amino acid residue led to compound REV-03, which exhibited a 30-fold increase in affinity. However, REV-03 exhibited poor pharmacokinetics and off-target effects. To address these issues, a sulfonyl group was introduced at the distal end and the isopropyl phenyl group was replaced with an amide structure, resulting in compound REV-04 which displayed significantly improved pharmacokinetic properties. Further refinement around REV-04 ultimately led to the development of revumenib as a first-in-class drug.

  Figure 8

  

  Figure 8.  (A) The original discovery of revumenib. (B) The binding model of revumenib with human menin (PDB code: 7UJ4).

  DownLoad: Full-Size Img PowerPoint

  The X-ray crystal structure (PDB code: 7UJ4) revealed that revumenib’s pyrimidine and sulfonamide formed multiple hydrogen bond interactions with Tyr-276, Met-322, and Trp-341 respectively (Fig. 8B) [43]. The protein pocket model demonstrated good compatibility between the benzoyl group and the protein cavity.

  3.6 Ensartinib (Ensacove^Ⓡ)

  In December 2024, the FDA approved the marketing application for ensacove for the treatment of NSCLC. Ensartinib is a potent, selective anaplastic lymphoma kinase (ALK) inhibitor that blocks the phosphorylation of ALK and its downstream signaling proteins Akt, ERK, and S6, inhibiting ALK-mediated pathways and mutant cell proliferation. It also targets other kinases like MET and ROS1. Ensartinib demonstrates promising efficacy in crizotinib-resistant and other ALK-resistant patients, offering new hope for NSCLC treatment.

  As a new generation of ALK inhibitors, ensartinib overcomed crizotinib-caused drug resistance by modifying its structure (Fig. 9A). Replacing the pyrazole ring with bis-benzamide significantly enhanced its inhibitory activity against multiple ALK mutations. The molecular docking analysis (PDB code: 2XP2) showed similar binding patterns, with aromatic amines forming hydrogen bonds with Met-1199 and Glu-1197, and thiazole and benzene rings forming σ-π hyperconjugations with Gly-1202 and Leu-1122 (Fig. 9B). The key difference was that ensartinib’s methylpiperazine formed ionic bond with Gly-1210, whereas crizotinib’s piperazine bound to Ala-1200, potentially altering the drug’s orientation and binding pattern.

  Figure 9

  

  Figure 9.  (A) The potential discovery of ensartinib. (B) The binding models of crizotinib and ensartinib with ALK (PDB code: 2XP2).

  DownLoad: Full-Size Img PowerPoint

  4. Cardiovascular and cerebrovascular drugs

  4.1 Aprocitentan (Tryvio^Ⓡ)

  The FDA approved aprocitentan, an oral tablet for refractory hypertension management [44]. Aprocitentan is a dual antagonist that targets endothelin-type A and endothelin-type B (ET_A/ET_B) receptors, effectively inhibiting the binding of endothelin-1 (ET-1) to these receptors and counteracting vasoconstrictive effects caused by ET-1. This mechanism inhibits vasoconstriction and fibrosis, while also mitigating vascular permeability and reducing water retention. Aprocitentan is a novel antihypertensive drug with a unique mechanism that complements Renin-angiotensin-aldosterone system (RAAS) blockers.

  The researchers found that T-0201, another ET antagonist, had 10 times greater activity in vitro compared to bosentan. This might be due to the presence of a pyrimidine ring at the end of glycol moiety (Fig. 10A) [45]. As a result, adding a 5‑bromo-pyrimidine to the ethylene glycol side chain of bosentan significantly improved its activity (BOS-01). Screening sulfonamide side chains for enhanced selectivity revealed that benzylamine side chains greatly improved selectivity (BOS-02). Further investigation into the impact of C2 and C5 substituents on pyrimidine ring activity showed that removing C2 substituents and incorporating a biaryl skeleton at C5 further improved selectivity (BOS-03). To meet oral availability requirements, replacing the benzylamine side chain with propylamine led to the development and commercialization of macitentan. Subsequent pharmacokinetic studies indicated that aprocitentan was the primary metabolite in vivo for macitentan, resulting in significantly increased plasma exposure and half-life (T_1/2: 48 h vs. 14 h).

  Figure 10

  

  Figure 10.  (A) The original discovery of aprocitentan. (B) The binding models of bosentan, macitentan and aprocitentan with ET_A receptor (PDB code: 8XVJ).

  DownLoad: Full-Size Img PowerPoint

  Since both macitentan and aprocitentan were derived from structural optimization of bosentan, we analyzed the differences between them using the X-ray crystal structure (PDB code: 8XVJ) of macitentan and ET_A receptor (Fig. 10B) [46]. The ethylene glycol side chain of bosentan exclusively formed hydrogen bonds with Thr-359, while the aromatic ring of anisole interacted through σ-π hyperconjugation with Lys-166. Macitentan, a selective ETA antagonist, enhanced hydrogen bond interactions with Asp-126 by introducing 5-bromopyrimidine from T-0201. Additionally, its key sulfonamide groups established multiple hydrogen bond interactions with Lys-166, Lys-255, and Arg-326. Aprocitentan, the active metabolite, exhibited a similar binding pattern to macitentan.

  4.2 Landiolol (Rapiblyk^Ⓡ)

  In November 2024, landiolol gained FDA approval for treating supraventricular tachycardia [47]. It is an ultra-short-acting adrenergic receptor antagonist with a β1/β2 selectivity ratio of 255. Landiolol quickly reduces heart rate without significantly affecting blood pressure.

  The introduction of chiral dioxolane into the phenyl group and piperazine into the amino group in propranolol’s structure resulted in the discovery of landiolol (Fig. 11A) [48]. Due to the limited research on landiolol, we speculated that these side chain modifications altered its conformation, thereby enhancing its selectivity. The eutectic structure (PDB code: 1DY4) revealed that hydroxyl and amino groups served as primary binding sites, forming hydrogen bonds with Gly-175, Glu-212, Asp-214, and Glu-217. Additionally, the naphthalene ring engaged in face-to-face π-π conjugation with Trp-376 (Fig. 11B). The molecular docking simulations also demonstrated that hydroxyl and carbonyl groups interacted via hydrogen bond with Asn-310 and Asn-329 respectively. These distinct binding modes contributed to differences in selectivity between the two drug molecules.

  Figure 11

  

  Figure 11.  (A) The potential discovery of landiolol based on the structure of propranolol. (B) The binding models of propranolol and landiolol with β adrenergic receptor (PDB code: 1DY4 and 4AMI).

  DownLoad: Full-Size Img PowerPoint

  5. Endocrine/metabolic drugs

  5.1 Resmetirom (Rezdiffra^Ⓡ)

  The FDA approved the marketing application for resmetirom in March 2024, making it the first therapeutic drug for non-alcoholic steatohepatitis (NASH) with liver fibrosis [49]. Resmetirom is a highly selective agonist of thyroid hormone receptor β (THR-β), which specifically activates THR-β expression in the liver to promote lipolysis, reduce levels of low-density cholesterol and triglycerides, and lower other atherogenic lipoproteins. Additionally, resmetirom stimulates mitochondrial biogenesis, effectively reduces hepatic lipid accumulation, alleviates liver fibrosis, and improves overall liver function. The approval of resmetirom marks the end of a 40-year period without significant advancements in the NASH field.

  Resmetirom was structurally modified based on triiodothyronine (T3), an endogenous active substance. The initial modification primarily focused on the carboxyl group, which enhanced selectivity by establishing hydrogen bonds with Asn-331 of THR-β. Despite safety concerns leading to discontinuation of most drugs, they provided valuable insights for subsequent investigations, particularly highlighting the possibility of replacing the phenolic ring with a nitrogen-containing heterocyclic ring. Consequently, resmetirom was discovered through cooperative modification of two functional groups: the phenolic ring and carboxyl group (Fig. 12A) [50]. The active molecule ALG-055009, derived from T3, had a binding model similar to that of T3 and resmetirom (Fig. 12B). The X-ray crystal structure analysis showed that the modified phenol ring maintained its H-bond interaction with His-435 while enhancing σ-π hyperconjugation with Leu-346 (PDB codes: 3GWS and 1N46). Additionally, the substitution of the carboxyl group enabled direct hydrogen bond between azapyrimidinone and Arg-320 instead of synergistic interaction with water, reducing binding desolvation loss.

  Figure 12

  

  Figure 12.  (A) The original discovery of resmetirom based on the structure of thyroid hormone (T3). (B) The binding models of T3, ALG-05509 and resmetirom with THR-β (PDB codes: 3GWS and 1N46).

  DownLoad: Full-Size Img PowerPoint

  5.2 Sofpironium (Sofdra^Ⓡ)

  In June 2024, the FDA approved sofpironium gel for treating primary axillary hyperhidrosis [51]. Sofpironium acts as a competitive inhibitor of the M3 muscarinic acetylcholine receptor (mAChR), effectively blocking sweat signals and inhibiting glandular sweating. This soft drug, based on glycopyrronium bromide, is specifically designed to produce its desired effect at the application site before quickly converting into an inactive, non-toxic metabolite upon entering systemic circulation. Therefore, it avoids typical anticholinergic side effects that may occur from off-site effects.

  Sofpironium was synthesized from the glycopyrronium structure by introducing an ester bond at the N-methyl of the quaternary ammonium salt, which enhanced drug absorption and metabolism while reducing side effects (Fig. 13A) [52]. The molecular docking simulation results (PDB code: 4DAJ) showed that both glycopyrronium and sofpironium had a similar binding mode. The hydroxyl and carbonyl groups on the chiral carbon formed multiple hydrogen bonds with Ala-235 and Asn-507, respectively, while methylene formed covalent interactions with Cys-532 (Fig. 13B).

  Figure 13

  

  Figure 13.  (A) The original discovery of sofpironium bromide based on the structure of glycopyrronium bromide. (B) The binding model of sofpironium with mAChR3 (PDB code: 4DAJ).

  DownLoad: Full-Size Img PowerPoint

  6. Genetic drugs

  6.1 Givinostat (Duvyzat^Ⓡ)

  The U.S. FDA approved Italfarmaco’s histone deacetylase (HDAC) inhibitor givinostat for treating Duchenne muscular dystrophy (DMD) in March 2024 [53]. Givinostat alters the structure of cellular DNA, preventing gene translation and muscle degeneration by inhibiting enzymes. It also promotes muscle repair, reduces inflammation, and prevents muscle fibrosis and fat accumulation by inhibiting HDAC activity. This approval establishes givinostat as a groundbreaking non-steroidal treatment option for all DMD gene variants, bringing new vitality to DMD therapy.

  Givinostat, patented in 1997, had the typical structural characteristics of traditional HDAC inhibitors with a cap, linker, and zinc binding group (ZBG). The X-ray crystal structure (PDB: 6UOC) revealed that the primary active site featured a hydroxylamine structure which chelated with zinc ions and formed diverse hydrogen bond interactions with His-192, His-193, and Tyr-363 (Fig. 14) [54]. This observation was further supported by the protein pocket binding model where the cap moiety faced the solvent region through the linker. Due to its early development, givinostat shared a SAR with conventional HDAC inhibitors. Incorporating nitrogen-containing heterocycles like pyridine and pyrimidine or hydrophilic groups such as dimethylamine, diethylamine, and piperazine at the cap site enhanced solubility and bioavailability. The linker site typically adopted an elongated chain structure that might include aromatic rings and amide bonds. The ZBG site predominantly employed classical hydroxylamine and o-phenylenediamine structures.

  Figure 14

  

  Figure 14.  The structure and binding models of givinostat with HDAC6 (PDB code: 6UOC).

  DownLoad: Full-Size Img PowerPoint

  6.2 Arimoclomol (Miplyffa^Ⓡ)

  In September 2024, the FDA approved arimoclomol and miglustat for treating Niemann-Pick disease type C (NPC) [55,56]. While the actual mechanism of action of arimoclomol is still unclear, the CytRx believes that it stimulates natural cellular protein repair and blocks interactions between non-native polypeptides through a unique ’molecular chaperone’ co-induction mechanism. In addition, arimoclomol boosts the activation of transcription factors EB (TFEB) and E3 (TFE3), which increases the expression and regulation of genes involved in lysosomal function. It is a small molecule that induces heat shock stress response, leading to increased production of heat shock proteins (HSPs). The heat shock response protects against cellular stress and helps maintain proper protein folding. Additionally, HSPs can rescue misfolded proteins, eliminate protein aggregates, and enhance lysosome functionality. Arimoclomol is also used in clinical treatment for Gaucher disease, sporadic inclusion body myositis, and amyotrophic lateral sclerosis. Miglustat is an oral small-molecule drug that inhibits glucosylceramide synthase through substrate reduction therapy.

  Miglustat, initially developed in the 1990s as a D-glucose analogue for HIV treatment, was later approved for Gaucher disease, Pompe disease, and NPC. Arimoclomol was first identified by Hungarian researchers in 2002 for treating insulin resistance and diabetes complications like retinopathy, neuropathy, and kidney disease. It was later screened for further development by the Hungarian company Biorex. Arimoclomol was commercially available as a citrate and initially developed as an HSP70 agonist (Fig. 15A). The molecular docking of arimoclomol with the Niemann-Pick C1 protein (PDB code: 5U74) revealed hydrogen bond interactions involving pyridine nitrogen atoms, Phe-1087, and Gln-1090 (Fig. 15B) [57]. In contrast, arimoclomol showed hydrogen bond interactions with Asp-40, Asp-79, and Asn-92 of HSP, along with a covalent interaction involving Met-173 (PDB code: 7KS9) [58]. These findings suggested that arimoclomol might modulate HSPs. While many HSP inhibitors and degraders have been developed, there were few reported HSP activators [59]. Interestingly, iroxanadine, an HSP activator derived from conformational restriction strategy by cyclizing the linker of arimoclomol.

  Figure 15

  

  Figure 15.  (A) The structures of arimoclomol and miglustat. (B) The binding models of arimoclomol with Niemann-Pick C1 protein and HSP (PDB codes: 5U74 and 7K9S).

  DownLoad: Full-Size Img PowerPoint

  6.3 Levacetylleucine (Aqneursa^Ⓡ)

  In September 2024, the U.S. FDA approved levacetylleucine for treating neurological symptoms associated with NPC (Fig. 16) [60]. Levacetylleucine, a modified amino acid, crosses the blood-brain barrier via monocarboxylate transporters and is then distributed to various tissues. It regulates ATP production in the enzyme control pathway, normalizes energy metabolism, improves lysosomal dysfunction, and reduces lipid accumulation within lysosomes. Additionally, it has potential to restore neuronal membrane potential, enhance cell signaling processes, preserve neuronal circuits, and alleviate neurological symptoms. Levacetylleucine was first discovered in 1901 by Nobel Prize winner Emil Ficher, in its racemic form (tanganil^Ⓡ), for the treatment of vertigo. Structurally, levacetylleucine is an acetylated derivative of leucine.

  Figure 16

  

  Figure 16.  The structure of levacetylleucine.

  DownLoad: Full-Size Img PowerPoint

  6.4 Acoramidis (Attruby^Ⓡ)

  In November 2024, acoramidis was approved by the U.S. FDA for treating transthyretin amyloid cardiomyopathy (ATTR-CM) [61]. Acoramidis selectively stabilizes transthyretin (TTR) by binding to its thyroxine binding site, slowing down the decomposition of TTR tetramer into monomers and preventing toxic amyloidosis.

  TTR stabilizer development is in its early stages with limited research reports available. Only acoramidis and tafamidis have been approved for the market (Fig. 17A) [62]. Acoramidis, discovered and developed by Eidos co-founders Isabella Graef and Mamoun Alhamadsheh, was designed to stabilize TTR effectively by preventing the formation of amyloid fibrils. Its selective binding mimicked the stabilizing effect of the T119 M mutation, which protected some individuals from ATTR. The X-ray crystal structure analysis of acoramidis (PDB code: 4HIQ) and tafamidis (PDB code: 3TCT) showed that the carboxyl group primarily bound to TTR through hydrogen bonds with Lys-15 (Fig. 17B). Acoramidis’ remarkable capability to achieve almost complete TTR stabilization might be due to the hydrogen bond interaction between pyrazole and Ser-117. The exceptional ability of acoramidis to achieve near-complete TTR stabilization may be attributed to the hydrogen bond interaction between pyrazole and Ser-117.

  Figure 17

  

  Figure 17.  (A) The structures of acoramidis and tafamidis. (B) The binding models of acoramidis and tafamidis with transthyretin (PDB codes: 4HIQ and 3TCT).

  DownLoad: Full-Size Img PowerPoint

  6.5 Crinecerfont (Crenessity^Ⓡ)

  In December 2024, the FDA approved crinecerfont for treating classical congenital adrenal hyperplasia (CAH), a rare genetic disorder caused by 21-hydroxylase deficiency. This deficiency leads to insufficient cortisol and aldosterone production and excess androgens. Crinecerfont, a selective corticotropin-releasing factor (CRF) 1 receptor antagonist, reduces adrenocorticotropic hormone (ACTH)-mediated adrenal androgen production by blocking CRF binding to the CRF1 receptor, thereby lowering the required glucocorticoid dose. This is the first major breakthrough in the field of CAH treatment in 70 years, providing patients with an entirely new non-hormone-dependent treatment option.

  Although no clear literature disclosed the development process of crinecerfont, it could be seen as using a scaffold hopping strategy since it originated from the same patent as NBI35965. As shown in Fig. 18A, opening the ternary fused ring of NBI35965 and replacing pyrazole with thiazole, pyridine with a benzene ring, and alkane with alkyne resulted in desired crinecerfont with significantly improved selectivity and activity. The molecular docking analysis (PDB code: 8GTI) revealed that NBI35965 and crinecerfont shareed similar binding patterns, primarily due to hydrogen bond interactions between the nitrogen heteroatoms on their pyrazole and thiazole rings and Asn-283 (Fig. 18B). The key difference was that NBI35965 formed σ-π hyperconjugation with Leu-280, while crinecerfont interacted with Met-206. This difference likely explained their distinct pharmacological activities.

  Figure 18

  

  Figure 18.  (A) The potential discovery of crinecerfont based on the structure of NBI35965. (B) The binding models of NBI35965 and crinecerfont with CRF1 receptor (PDB code: 8GTI).

  DownLoad: Full-Size Img PowerPoint

  6.6 Vanzacaftor, tezacaftor and deutivacaftor (Alyftrek^Ⓡ)

  In December 2024, the FDA approved Alyftrek^Ⓡ for marketing as a triple combination cystic fibrosis transmembrane conduction regulator (CFTR) modulator to treat cystic fibrosis. Vanzacaftor and tezacaftor bind to different sites on the CFTR protein, promoting cell processing and transport of mutant CFTR like F508del-CFTR, increasing surface CFTR levels compared to single molecule. Deutivacaftor enhances the probability of CFTR channel opening on the cell surface. Together, these components increase both the amount and function of CFTR on the cell surface, boosting overall CFTR activity.

  Tezacaftor, an active ingredient in cystic fibrosis drugs Symdeko^Ⓡ and Trikafta^Ⓡ, was approved by the FDA in 2018 and 2019, respectively (Fig. 19A). Deutivacaftor significantly improved the metabolic activity of ivacaftor through tert‑butyl deuteration. Vanzacaftor was an active molecule based on the structure of the marketed drug elexacaftor by closing the loop of chiral tetrahydropyrrole and sulfonamide using macrocyclic strategy (Fig. 19B). The molecular docking analysis (PDB code: 8EIG) showed that elexacaftor and vanzacaftor had similar binding models, mainly forming hydrogen bond interactions and σ-π hyperconjugation with Arg-21, Arg-25, Trp-1098 and Arg-1102 around the sulfonamide site (Fig. 19C).

  Figure 19

  

  Figure 19.  (A) The structures of vanzacaftor, tezacaftor and deutivacaftor. (B) The potential discovery of vanzacaftor. (C) The binding models of elexacaftor and vanzacaftor with CFTR (PDB code: 8EIG).

  DownLoad: Full-Size Img PowerPoint

  7. Hematological drugs

  7.1 Vadadustat (Vafseo^Ⓡ)

  In March 2024, the FDA approved Akebia Therapeutics’ vadadustat tablets for treating anemia caused by chronic kidney disease (CKD) [63]. Vadadustat is a reversible hypoxia-inducing factor prolyl hydroxylase inhibitor (HIF-PHI), which stabilizes and accumulates HIF-lα and HIF-2α transcription factors, increasing erythropoietin (EPO) production to stimulate red blood cell production and improve oxygen transport.

  The activity of HIF-pH was regulated by iron and 2-oxoglutaric acid (2-OG). Since the discovery of N-oxalylglycine (NOG), subsequent HIF-PHIs roxadustat, daprodustat, and vadadustat all utilized NOG as the primary active binding site (Fig. 20A) [64]. The X-ray crystal structure (PDB code: 5OPC) revealed that NOG and pyridine chelated with the iron ion, forming hydrogen bond interactions with Tyr-145, Thr-196, and Lys-214 (Fig. 20B). Additionally, the phenol hydroxyl group formed a hydrogen bond interaction with Gln-147, while the biaryl group engaged in σ-π hyperconjugation with Trp-296 to enhance compound activity. The protein pocket binding model confirmed this theory once again as NOG resided deep within the pocket while its biaryl structure faced towards the solvent region.

  Figure 20

  

  Figure 20.  (A) The potential discovery of vadadustat. (B) The binding model of vadadustat and HIF-pH (PDB code: 5OPC).

  DownLoad: Full-Size Img PowerPoint

  7.2 Danicopan (Voydeya^Ⓡ)

  In April 2024, Astrazeneca’s oral small molecule danicopan received FDA approval as an adjunct to standard therapy for treating extravascular hemolysis (EVH) in patients with paroxysmal nocturnal hemoglobinuria (PNH), alongside complement factor C5 inhibitors ravulizumab and eculizumab [65]. Danicopan selectively inhibits complement factor D in the alternative pathway, blocking C3 invertase production and inhibiting replacement pathway activity. Unlike C5 inhibitors, danicopan effectively prevents deposition of C3b fragments on erythrocytes, controlling extravascular hemolysis and red blood cell breakdown.

  Through high-throughput screening, researchers identified a series of L-prolinamide-based compounds that inhibited complement factor D (Fig. 21A) [66]. Subsequent modifications to the substituents of L-prolinamide led to the discovery of DAN-01, an indazole-containing compound. Finally, employing SBDD and fragment-based drug design (FBDD) approaches, the substituents in each site were optimized and screened, resulting in the development of danicopan. The X-ray crystal structure (PDB code: 8DG6) revealed that L-prolinamide-indazole formed multiple hydrogen bond interactions with Leu-25, Trp-128, Gly-181, Thr198, and Arg-202 while also exhibiting σ-π hyperconjugation with Lys-180 (Fig. 21B). The molecular docking simulations demonstrated that danicopan further enhanced its interaction with Lys-180 and Arg-202 by introducing pyridine rings without compromising original interactions.

  Figure 21

  

  Figure 21.  (A) The original discovery of danicopan. (B) The binding models of DAN-01 and danicopan with complement factor D (PDB code: 8DG6).

  DownLoad: Full-Size Img PowerPoint

  8. Immunological drugs

  8.1 Elafibranor (Iqirvo^Ⓡ)

  In June 2024, elafibranor received accelerated FDA approval for treating primary biliary cholangitis (PBC) [67]. As a peroxisome proliferator-activated receptor (PPAR) α/δ agonist, elafibranor effectively reduces fibrosis and bile acid production. Its receptor binding capability also suppresses inflammatory factors, alleviating PBC inflammation. Being the first small molecule drug in nearly a decade to address PBC, elafibranor offers patients a new therapeutic option.

  The spacious ligand-binding domain of PPAR accommodated various ligands, leading to significant variations in the molecular structure of PPAR agonists. This allowed for the development of drugs with distinct specificity, potency, and therapeutic effects. Elafibranor was derived from bezafibrate by substituting its amide linker between two benzene rings with an α,β-unsaturated carbon group (Fig. 22A) [68,69]. Because they shared similar structures, bezafibrate and elafibranor showed comparable binding patterns (PDB codes: 7BPZ and 8HUQ). Both compounds bound primarily at their carboxyl groups which formed strong hydrogen bonds with Ser-280, Tyr-314, His-440, and Tyr-464 (Fig. 22B). Furthermore, Cys276 formed covalent interactions with the unsaturated carbonyl linker.

  Figure 22

  

  Figure 22.  (A) The potential discovery of elafibranor based on the structure of bezafibrate. (B) The X-ray crystal structures of bezafibrate and elafibranor bind to PPAR (PDB codes: 7BPZ and 8HUQ).

  DownLoad: Full-Size Img PowerPoint

  8.2 Seladelpar (Livdelzi^Ⓡ)

  The FDA approved seladelpar in August 2024 for treating PBC. As a selective agonist of PPAR-δ, it activates PPAR-δ to reduce bile acid synthesis by downregulating CYP7A1 expression [70]. Additionally, seladelpar has anti-cholesterol and anti-inflammatory properties, potentially possessing anti-fibrotic characteristics. Compared to placebo, seladelpar demonstrated significant therapeutic efficacy in normalizing alkaline phosphatase levels, key biomarkers, and controlling itch.

  Initially, CymaBay discovered that the PPAR-δ agonist GW501516 had dual effects on normal men’s systemic circulation: it increased high density lipoprotein cholesterol (HDL-C) and decreased triglyceride levels, while also reducing plasma total cholesterol and apolipoprotein B in men with dyslipidemia and metabolic syndrome [71]. These findings suggested the potential of PPAR-δ agonists for treating dyslipidemia, obesity, and diabetes. In contrast to PPAR-α/γ agonists, the selective PPAR-δ agonists have been rarely reported. Leveraging existing drugs GW501516 and L-165041, the researchers used a molecular hybridization strategy to construct seladelpar through group substitution based on traditional drug design principles (Fig. 23A). The X-ray crystal structure (PDB code: 8HUO) revealed that seladelpar exhibited a binding pattern similar to other PPAR inhibitors, interacting with His-287, His-413, and Tyr-437 at the carboxyl terminal through hydrogen bonds (Fig. 23B). The only difference was the substitution of an unsaturated linker with a long chain of chiral ethers. This enhanced molecular flexibility and precisely achieved selectivity in PPAR-δ inhibitors.

  Figure 23

  

  Figure 23.  (A) The potential discovery of seladelpar based on the structures of GW501516 and L-165,041. (B) The X-ray crystal structure of seladelpar binds to PPAR (PDB code: 8HUP).

  DownLoad: Full-Size Img PowerPoint

  8.3 Mavorixafor (Xolremdi^Ⓡ)

  The oral capsule formulation of mavorixafor received FDA approval in April 2024 for treating warts, hypogammaglobulinemia, immunodeficiency, myelokathexis (WHIM) syndrome [72]. By selectively blocking the interaction between C-X-C chemokine receptor 4 (CXCR4) and its ligand CXCL12, mavorixafor effectively inhibits the overactivation of the CXCR4/CXCL12 signaling pathway. This inhibition ultimately increases mature neutrophils and lymphocytes in the blood, impacting their normal circulation between the bone marrow and bloodstream.

  Mavorixafor was initially developed as an inhibitor against human immunodeficiency virus 1 (HIV-1). The primary screening identified MAV-01 (JM1657), which possessed inhibitory activity due to its cyclam structure (Fig. 24A) [73,74]. However, the synthesis posed challenges due to two chiral centers. The initial optimization efforts focused on enhancing the linker, leading to the development of a compound with enhanced activity known as MAV-02 (AMD3100). Subsequently, screening of cyclam rings and alternative groups led to the identification of the active compound MAV-03 (AMD3465). However, its oral bioavailability was found to be poor. To address this issue, a skeleton transition technique replaced a nitrogen atom on cyclam with a pyridine ring, resulting in the synthesis of MAV-04 as a hit compound. Considering the challenges associated with macrocycle synthesis, the researchers opted for opening the macrocycle and screening various side chain substituents. Ultimately, they successfully obtained MAV-05 by incorporating a benzimidazole moiety into the structure which exhibited significant inhibitory activity. To further enhance both bioactivity and pharmacokinetic properties, substitution of the initial benzylamine side chain with n-butylamine led to mavorixafor.

  Figure 24

  

  Figure 24.  (A) The original discovery of mavorixafor. (B) The binding models of MVA-02 and mavorixafor with CXCR4 (PDB code: 8U4P).

  DownLoad: Full-Size Img PowerPoint

  The X-ray crystal structure analysis (PDB code: 8U4P) revealed that MAV-02′s two cyclam rings formed multiple ionic bond interactions with Gly-288 and Asp-262, respectively (Fig. 24B) [75]. The modified mavorixafor maintained its interaction with Glu-288 through benzimidazole, while the primary amine and pyridine exhibited ionic bond interactions and face-to-face π-π stacking with Asp-97, Cys-186, and Trp-94. By integrating research and development data, we discovered that benzimidazole was an essential pharmacophore for activity maintenance. The tetrahydroquinoline obtained through scaffold hopping preserved the hydrogen bond receptor and rigid structure, thereby enhancing biological activity similar to n-butylamino which improved flexibility and water solubility.

  8.4 Deuruxolitinib (Leqselvi^Ⓡ)

  In July 2024, deuruxolitinib gained FDA approval for treating severe alopecia areata [76]. As a selective Janus kinase (JAK) inhibitor, deuruxolitinib disrupts signal transduction pathways involved in autoimmune attack on hair follicles, preventing their destruction and promoting regeneration. Deuruxolitinib is a deuterated derivative of ruxolitinib specifically engineered to reduce the extensive oxidative metabolism associated with the cyclopentyl group. This advanced design significantly prolongs its pharmacological effects and enhances therapeutic potential.

  Deuruxolitinib was a deuterated derivative of cyclopentane methylene, designed to enhance the stability of its five-member ring and improve its pharmacokinetic properties based on the structure of ruxolitinib (Fig. 25A) [77]. The crucial binding site for deuruxolitinib, like ruxolitinib, involved pyrimidine pyrrole that formed hydrogen bond interactions with Glu-930 and Leu-932 (Fig. 25B). Additionally, the pyrimidine ring established σ-π hyperconjugation with Leu-855, while the covalent warhead cyano group interacted covalently with Met-929.

  Figure 25

  

  Figure 25.  (A) The potential discovery of deuruxolitinib based on the structure of ruxolitinib. (B) The binding model of ruxolitinib with JAK2 (PDB code: 6VGL).

  DownLoad: Full-Size Img PowerPoint

  9. Psychotropic drugs

  In September 2024, the combination drug xanomeline and trospium chloride gained official FDA approval for treating schizophrenia [78]. Xanomeline, a selective agonist of mAChR1 and mAChR4, has high blood-brain barrier permeability [79]. On the other hand, trospium chloride acts as a peripheral antagonist on M receptors due to its limited ability to penetrate the blood-brain barrier. Its primary purpose is to mitigate peripheral effects associated with xanomeline and minimize adverse reactions. Cobenfy^Ⓡ represents an innovative antipsychotic medication that targets cholinergic receptors instead of dopamine receptors in schizophrenia treatment, marking a significant therapeutic breakthrough.

  Both xanomeline and trospium were well-established drugs and developed over two decades ago (Fig. 26A). The development of xanomeline seemed to have been inspired by the structural characteristics of pilocarpine (Salagen Tablet) and the pharmacophore of arecoline [80]. The research and development strategy aimed to incorporate hydrophilic ring groups and lipophilic long chains into adjacent positions of five-membered nitrogen-containing heterocyclic rings. The notable mAChR agonists with similar structural features included ENS-163, tazomeline, alvameline, and vadaclidine. Trospium, a trodamine compound, exhibited structural characteristics similar to atropine, homatropine, tiotropium bromide, and scopolamine. These shared features included hydroxyl groups, phenylacetate moieties, and quaternary ammonium salts.

  Figure 26

  

  Figure 26.  (A) The structures of xanomeline tartrate and trospium chloride. (B) The binding models of xanomeline and trospium with mAChR (PDB codes: 8FX5 and 6WJC).

  DownLoad: Full-Size Img PowerPoint

  The X-ray crystal structure analysis revealed the crucial role of these structures in fulfilling their functions. For example, xanomeline (PDB code: 8FX5) had a hydrophilic side chain that formed hydrogen bonds with Asp-432, and its nitrogen-containing aromatic heterocyclic ring participated in σ-π hyperconjugation and formed hydrogen bonds with Trp-435 and Tyr-89 (Fig. 26B). Additionally, hydrophobic side chains establish σ-π hyperconjugation with Tyr-92. Trospium (PDB code: 6WJC) had ester and hydroxyl groups that interacted with Asn-382 and Thr-192 respectively, while its aromatic ring underwent side-to-face π-π stacking with Trp-378.

  10. Respiratory drugs

  In June 2024, ensifentrine gained FDA approval for maintaining chronic obstructive pulmonary disease (COPD) treatment [81]. As a dual inhibitor targeting phosphodiesterase 3/4 (PD3/4), it simultaneously inhibits PDE3 and PDE4 to reduce inflammation and relax airway muscles. Additionally, it activates cystic fibrosis transmembrane conductance regulators to improve mucus clearance. Notably, ensifentrine is the first inhalation therapy with a new mechanism of action for COPD maintenance in over two decades.

  The development of ensifentrine was based on trequinsin’s structural framework (Fig. 27A) [82]. By adding urea moieties to the N-methyl of amide, ensifentrine showed improved selectivity and activity. The X-ray crystal structure analysis (PDB code: 7L28) revealed that trequinsin and ensifentrine had similar binding patterns, with the terpolymer ring engaging in π-π and σ-π stacking interactions with pH-1004 and Ile-968 (Fig. 27B). The introduction of urea enhanced hydrogen bond interactions with Arg-843, significantly contributing to its improved activity and selectivity.

  Figure 27

  

  Figure 27.  (A) The potential discovery of ensifentrine based on the structure of trequinsin. (B) The X-ray crystal structures of trequinsin and ensifentrine bind to PDE3 (PDB code: 7L28).

  DownLoad: Full-Size Img PowerPoint

  11. Conclusion and prospect

  In 2024, the US FDA approved a total of 50 drug marketing applications, including 31 molecular entities, accounting for half of the country’s approvals. The approved biological products mainly included monoclonal antibodies, fusion proteins, parathyroid hormones, and oligonucleotides. These innovative molecules were outcomes of modern drug discovery techniques. For example, the introduction of a urea group on the N-methyl amide of trequinsin, facilitated by SBDD, enhanced hydrogen bond interaction with Arg-843 while preserving the original interaction relationship, resulting in improved PD3/4 inhibitor ensifentrine. The selective PARP-δ agonist seladelpar was obtained by combining pharmacophores from GW501516 and L-165041, along with a linker through molecular hybridization. The use of FBDD allowed for a comprehensive investigation into the SARs of lead compound DAN-01, resulting in an innovative complement factor D inhibitor with a novel mechanism of action. Deuteration was also an effective strategy, while deuruxolitinib with a cyclopentane moiety derived from ruxolitinib exhibited significantly improved pharmacokinetic properties through deuteration. Prodrug preparation was an effective approach in drug development that enhanced drug delivery and minimized toxic side effects. Sulopenem etzadroxil and sofpironium bromide exemplified this strategy by introducing ester bonds based on sulopenem and glycopyrronium bromide structures, respectively, to slow down metabolic rates in vivo. Therefore, the analysis of drug research and development experience is invaluable for understanding contemporary technology and principles of drug design.

  Upon reviewing recently approved drug molecules, it is evident that they often have similar or comparable skeletal structures. As a result, future drug development will require exploring pathogenesis, identifying new biomarkers, and diversifying drug classes. Additionally, there is an urgent need to embrace scientific and technological advancements such as machine learning and AI. The state-of-the-art technologies and drug scaffold structures will be the foundation for future drug development, providing valuable insights for emerging technologies and advancements in drug discovery. By reviewing the developmental strategies used for the latest approved drugs, we aim to enhance drug discovery technology and accelerate new drug development. With the increasing availability and advancement of emerging technologies, small molecule drugs show promise in addressing unresolved clinical needs while promoting greater alignment with biotechnological therapies for global patient benefit.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Yi-Ru Bai: Writing – review & editing, Writing – original draft, Conceptualization. Qing-Chuan Duan: Writing – review & editing, Writing – original draft. Dong-Jie Seng: Software, Data curation. Ying Xu: Software, Data curation. Hong-Bo Ren: Methodology, Investigation. Jie Zhang: Investigation, Data curation. Dan-Dan Shen: Software, Methodology. Li Yang: Writing – review & editing. Hong-Min Liu: Writing – review & editing, Conceptualization. Shuo Yuan: Writing – review & editing, Writing – original draft, Conceptualization.

  Acknowledgments

  The authors thank the National Natural Science Foundation of China (No. 82304286 by S. Yuan, Nos. U21A20416 and 82020108030 by H.-M. Liu) and the Scientific and Technological Project of Henan Province (No. 232102311165 by S. Yuan, No. 242102311236 by Y.-R. Bai) for financial support.

  
  
• 1. [1]

     P. Verdin, Nat. Rev. Drug Discov. 23 (2024) 331. doi: 10.1038/d41573-024-00063-x

  2. [2]

     P. Verdin, Nat. Rev. Drug Discov. 23 (2024) 573. doi: 10.1038/d41573-024-00118-z

  3. [3]

     I. Morrissey, S. Magnet, S. Hawser, et al., Antimicrob. Agents Chemother. 63 (2019) e00514-519.

  4. [4]

     S.S. Awad, A.H. Rodriguez, Y.C. Chuang, et al., Clin. Infect. Dis. 59 (2014) 51–61. doi: 10.1093/cid/ciu219

  5. [5]

     M.W. Dunne, S.I. Aronin, A.F. Das, et al., Clin. Infect. Dis. 76 (2023) 78–88. doi: 10.1093/cid/ciac704

  6. [6]

     S. Gregoriou, A. Tsiogka, G. Kontochristopoulos, et al., Expert Opin. Invest. Drugs 31 (2022) 15–21. doi: 10.1080/13543784.2022.2017880

  7. [7]

     B. King, M.M. Senna, N.A. Mesinkovska, et al., J. Am. Acad. Dermatol. 91 (2024) 880–888. doi: 10.1016/j.jaad.2024.06.097

  8. [8]

     K.V. Kowdley, C.L. Bowlus, C. Levy, et al., N. Engl. J. Med. 390 (2024) 795–805. doi: 10.1056/nejmoa2306185

  9. [9]

     B.D. Cons, D.G. Twigg, R. Kumar, et al., J. Med. Chem. 65 (2022) 7476–7488. doi: 10.1021/acs.jmedchem.2c00164

  10. [10]

      S. Zhao, X. Zhang, E.F. da Silva-Junior, et al., Drug Discov. Today 28 (2023) 103581. doi: 10.1016/j.drudis.2023.103581

  11. [11]

      A.V. Sadybekov, V. Katritch, Nature 616 (2023) 673–685. doi: 10.1038/s41586-023-05905-z

  12. [12]

      R.A. Bauer, Drug Discov. Today 20 (2015) 1061–1073. doi: 10.1016/j.drudis.2015.05.005

  13. [13]

      D.A. Nalawansha, C.M. Crews, Cell Chem. Biol. 27 (2020) 998–1014. doi: 10.1016/j.chembiol.2020.07.020

  14. [14]

      Y. Wang, A.G. Cheetham, G. Angacian, et al., Adv. Drug Deliv. Rev. 110-111 (2017) 112–126. doi: 10.1016/j.addr.2016.06.015

  15. [15]

      R.M.C. Di Martino, B.D. Maxwell, T. Pirali, Nat. Rev. Drug Discov. 22 (2023) 562–584. doi: 10.1038/s41573-023-00703-8

  16. [16]

      C. Sarkar, B. Das, V.S. Rawat, et al., Int. J. Mol. Sci. 24 (2023) 2026. doi: 10.3390/ijms24032026

  17. [17]

      C. Cerchia, A. Lavecchia, Drug Discov. Today 28 (2023) 103516. doi: 10.1016/j.drudis.2023.103516

  18. [18]

      Y.R. Bai, D.J. Seng, Y. Xu, et al., Eur. J. Med. Chem. 276 (2024) 116706. doi: 10.1016/j.ejmech.2024.116706

  19. [19]

      Y.R. Bai, X. Yang, K.T. Chen, et al., Eur. J. Med. Chem. 277 (2024) 116759. doi: 10.1016/j.ejmech.2024.116759

  20. [20]

      S. Yuan, D.D. Shen, R. Jia, et al., Med. Res. Rev. 43 (2023) 2352–2391. doi: 10.1002/med.21976

  21. [21]

      K.S. Kaye, A. Belley, P. Barth, et al., JAMA 328 (2022) 1304–1314. doi: 10.1001/jama.2022.17034

  22. [22]

      G.M. Pais, J. Chang, E.F. Barreto, et al., Clin. Pharmacokinet. 61 (2022) 929–953. doi: 10.1007/s40262-022-01137-y

  23. [23]

      R.A. Bonnin, K. Jeannot, A.S. Henriksen, et al., Clin. Microbiol. Infect. 31 (2025) 240–249. doi: 10.1016/j.cmi.2024.09.031

  24. [24]

      P.A. Lang, R. Raj, A. Tumber, et al., Proc. Natl. Acad. Sci. U. S. A. 119 (2022) e2117310119. doi: 10.1073/pnas.2117310119

  25. [25]

      W.H. Hsu, C.K. Hsu, C.C. Lai, Expert Rev. Anti-Infect. Ther. 21 (2023) 551–563. doi: 10.1080/14787210.2023.2202851

  26. [26]

      M.W. Shalaby, E.M.E. Dokla, R.A.T. Serya, et al., Eur. J. Med. Chem. 199 (2020) 112312. doi: 10.1016/j.ejmech.2020.112312

  27. [27]

      M.W. Dunne, S.I. Aronin, A.F. Das, et al., Clin. Infect. Dis. 76 (2023) 66–77. doi: 10.1093/cid/ciac738

  28. [28]

      G.G. Zhanel, M. Pozdirca, A.R. Golden, et al., Drugs 82 (2022) 533–557. doi: 10.1007/s40265-022-01688-1

  29. [29]

      R.J. Everts, S.J. Gardiner, M. Zhang, et al., J. Infect. 83 (2021) 182–189. doi: 10.1016/j.jinf.2021.05.037

  30. [30]

      A. Mollica, R. Costante, A. Akdemir, et al., Bioorg. Med. Chem. 23 (2015) 5311–5318. doi: 10.1016/j.bmc.2015.07.066

  31. [31]

      L.B. Kilburn, D.A. Khuong-Quang, J.R. Hansford, et al., Nat. Med. 30 (2024) 207–217. doi: 10.1038/s41591-023-02668-y

  32. [32]

      E. Tkacik, K. Li, G. Gonzalez-Del Pino, et al., J. Biol. Chem. 299 (2023) 104634. doi: 10.1016/j.jbc.2023.104634

  33. [33]

      I.K. Mellinghoff, M. Lu, P.Y. Wen, et al., Nat. Med. 29 (2023) 615–622. doi: 10.1038/s41591-022-02141-2

  34. [34]

      Z. Konteatis, E. Artin, B. Nicolay, et al., ACS Med. Chem. Lett. 11 (2020) 101–107. doi: 10.1021/acsmedchemlett.9b00509

  35. [35]

      K. Yen, J. Travins, F. Wang, et al., Cancer Discov. 7 (2017) 478–493. doi: 10.1158/2159-8290.CD-16-1034

  36. [36]

      B.C. Cho, S. Lu, E. Felip, et al., N. Engl. J. Med. 391 (2024) 1486–1498. doi: 10.1056/nejmoa2403614

  37. [37]

      J. Yun, S.H. Lee, S.Y. Kim, et al., Cancer Discov. 10 (2020) 1194–1209. doi: 10.1158/2159-8290.cd-20-0116

  38. [38]

      T.S. Beyett, C. To, D.E. Heppner, et al., Nat. Commun. 13 (2022) 2530. doi: 10.1038/s41467-022-30258-y

  39. [39]

      K.W. Song, K.A. Edgar, E.J. Hanan, et al., Cancer Discov. 12 (2022) 204–219. doi: 10.1158/2159-8290.cd-21-0072

  40. [40]

      E.J. Hanan, M.G. Braun, R.A. Heald, et al., J. Med. Chem. 65 (2022) 16589–16621. doi: 10.1021/acs.jmedchem.2c01422

  41. [41]

      P. Sidaway, Nat. Rev. Clin. Oncol. 20 (2023) 284.

  42. [42]

      H. Lei, S.Q. Zhang, S. Fan, et al., J. Med. Chem. 64 (2021) 15519–15533. doi: 10.1021/acs.jmedchem.1c00872

  43. [43]

      F. Perner, E.M. Stein, D.V. Wenge, et al., Nature 615 (2023) 913–919. doi: 10.1038/s41586-023-05755-9

  44. [44]

      M.P. Schlaich, M. Bellet, M.A. Weber, et al., Lancet 400 (2022) 1927–1937. doi: 10.1016/S0140-6736(22)02034-7

  45. [45]

      M.H. Bolli, C. Boss, C. Binkert, et al., J. Med. Chem. 55 (2012) 7849–7861. doi: 10.1021/jm3009103

  46. [46]

      J. Hou, S. Liu, X. Zhang, et al., Cell Discov. 10 (2024) 79.

  47. [47]

      G. Gelbenegger, B. Jilma, L.C. Horvath, et al., Resuscitation 201 (2024) 110273. doi: 10.1016/j.resuscitation.2024.110273

  48. [48]

      J. Stahlberg, H. Henriksson, C. Divne, et al., J. Mol. Biol. 305 (2001) 79–93. doi: 10.1006/jmbi.2000.4237

  49. [49]

      S.A. Harrison, R. Taub, G.W. Neff, et al., Nat. Med. 29 (2023) 2919–2928. doi: 10.1038/s41591-023-02603-1

  50. [50]

      M.J. Kelly, S. Pietranico-Cole, J.D. Larigan, et al., J. Med. Chem. 57 (2014) 3912–3923. doi: 10.1021/jm4019299

  51. [51]

      B. Kirsch, S. Smith, J. Cohen, et al., J. Am. Acad. Dermatol. 82 (2020) 1321–1327. doi: 10.1016/j.jaad.2020.02.016

  52. [52]

      A.C. Kruse, J. Hu, A.C. Pan, et al., Nature 482 (2012) 552–556. doi: 10.1038/nature10867

  53. [53]

      E. Mercuri, J.J. Vilchez, O. Boespflug-Tanguy, et al., Lancet Neurol. 23 (2024) 393–403. doi: 10.1016/S1474-4422(24)00036-X

  54. [54]

      J.D. Osko, D.W. Christianson, Biochemistry 58 (2019) 4912–4924. doi: 10.1021/acs.biochem.9b00934

  55. [55]

      M. Benatar, T. Hansen, D. Rom, et al., Lancet Neurol. 23 (2024) 687–699. doi: 10.1016/S1474-4422(24)00134-0

  56. [56]

      V. Mansouri, A.R. Tavasoli, M. Khodarahmi, et al., Eur. J. Neurol. 30 (2023) 2919–2945. doi: 10.1111/ene.15871

  57. [57]

      X. Li, F. Lu, M.N. Trinh, et al., Proc. Natl. Acad. Sci. U. S. A. 114 (2017) 9116–9121. doi: 10.1073/pnas.1711716114

  58. [58]

      P.T. Marcyk, E.V. LeBlanc, D.A. Kuntz, et al., J. Med. Chem. 64 (2021) 1139–1169. doi: 10.1021/acs.jmedchem.0c01777

  59. [59]

      P. Singh, B. Unik, A. Puri, et al., Database 2020 (2020) baaa003. doi: 10.1093/database/baaa003

  60. [60]

      P. Beninger, Clin. Ther. 46 (2024) 1091–1092. doi: 10.1016/j.clinthera.2024.10.011

  61. [61]

      J.H. Byun, S. Nguyen, N. Engl. J. Med. 390 (2024) 1343–1344. doi: 10.1056/NEJMc2402048

  62. [62]

      S.C. Penchala, S. Connelly, Y. Wang, et al., Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 9992–9997. doi: 10.1073/pnas.1300761110

  63. [63]

      M.J. Koury, R. Agarwal, G.M. Chertow, et al., Am. J. Hematol. 97 (2022) 1178–1188. doi: 10.1002/ajh.26644

  64. [64]

      T.L. Yeh, T.M. Leissing, M.I. Abboud, et al., Chem. Sci. 8 (2017) 7651–7668. doi: 10.1039/C7SC02103H

  65. [65]

      A.M. Risitano, A.G. Kulasekararaj, J.W. Lee, et al., Haematologica 106 (2021) 3188–3197.

  66. [66]

      W. Zhang, M. Wu, S. Vadlakonda, et al., Bioorg. Med. Chem. 74 (2022) 117034. doi: 10.1016/j.bmc.2022.117034

  67. [67]

      J.M. Schattenberg, A. Pares, K.V. Kowdley, et al., J. Hepatol. 74 (2021) 1344–1354. doi: 10.1016/j.jhep.2021.01.013

  68. [68]

      S. Kamata, A. Honda, R. Ishikawa, et al., Antioxidants 12 (2023) 1523. doi: 10.3390/antiox12081523

  69. [69]

      S. Kamata, T. Oyama, K. Saito, et al., iScience 23 (2020) 101727. doi: 10.1016/j.isci.2020.101727

  70. [70]

      G.M. Hirschfield, M.L. Shiffman, A. Gulamhusein, et al., Hepatology 78 (2023) 397–415. doi: 10.1097/hep.0000000000000395

  71. [71]

      R. Zhang, A. Wang, A. DeAngelis, et al., Bioorg. Med. Chem. Lett. 17 (2007) 3855–3859. doi: 10.1016/j.bmcl.2007.05.007

  72. [72]

      C.B. Geier, Blood 144 (2024) 1–2. doi: 10.1182/blood.2024024942

  73. [73]

      G.J. Bridger, R.T. Skerlj, P.E. Hernandez-Abad, et al., J. Med. Chem. 53 (2010) 1250–1260. doi: 10.1021/jm901530b

  74. [74]

      R.T. Skerlj, G.J. Bridger, A. Kaller, et al., J. Med. Chem. 53 (2010) 3376–3388. doi: 10.1021/jm100073m

  75. [75]

      K. Saotome, L.L. McGoldrick, J.H. Ho, et al., Nat. Struct. Mol. Biol. 32 (2025) 315–325. doi: 10.1038/s41594-024-01397-1

  76. [76]

      B.A. King, B.G. Craiglow, J. Am. Acad. Dermatol. 89 (2023) 29–32. doi: 10.12681/ayil.33038

  77. [77]

      R.R. Davis, B. Li, S.Y. Yun, et al., J. Med. Chem. 64 (2021) 2228–2241. doi: 10.1021/acs.jmedchem.0c01952

  78. [78]

      I. Kaul, S. Sawchak, D.P. Walling, et al., JAMA Psychiatry 81 (2024) 749–756. doi: 10.1001/jamapsychiatry.2024.0785

  79. [79]

      W.A.C. Burger, V. Pham, Z. Vuckovic, et al., Nat. Commun. 14 (2023) 5440. doi: 10.1038/s41467-023-41199-5

  80. [80]

      S.M. Paul, S.E. Yohn, M. Popiolek, et al., Am. J. Psychiatry. 179 (2022) 611–627. doi: 10.1176/appi.ajp.21101083

  81. [81]

      A. Anzueto, I.Z. Barjaktarevic, T.M. Siler, et al., Am. J. Respir. Crit. Care. Med. 208 (2023) 406–416. doi: 10.1164/rccm.202306-0944oc

  82. [82]

      C.W. Garvie, X. Wu, M. Papanastasiou, et al., Nat. Commun. 12 (2021) 4375. doi: 10.1038/s41467-021-24495-w

• 

  Figure 1  (A) The potential discovery of enmetazobactam based on the structure of tazobactam. (B) The X-ray crystal structures of β-lactamases bind to tazobactam (PDB code: 6XFS) and enmetazobactam (PDB code: 6T35).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) The potential discovery of ceftobiprole medocaril sodium based on the structure of ceftaroline. (B) The X-ray crystal structures of ceftaroline and ceftobiprole medocaril bind to PBP (PDB codes: 3ZG0 and 4DKI).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) The structures of sulopenem etzadroxil and probenecid. (B) The X-ray crystal structures of sulopenem and probenecid bind to PBP and OAT (PDB codes: 6I1H and 8SDZ).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (A) The potential discovery of tovorafenib. (B) The binding models of naporafenib and tovorafenib with BRAF (PDB codes: 8F7O and 8F7P).

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (A) The original discovery of vorasidenib. (B) The binding models of enasidenib and vorasidenib with IDH1/2 (PDB codes: 5I96, 6VFZ and 6VEI).

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (A) The potential discovery of lazertinib based on the structure of osimertinib. (B) The binding models of osimertinib and lazertinib with EGFR (PDB codes: 7JXM and 7UKW).

  下载: 全尺寸图片 幻灯片
  

  Figure 7  (A) The original discovery of inavolisib. (B) The binding model of inavolisib with PI3Kα (PDB code: 8EXV).

  下载: 全尺寸图片 幻灯片
  

  Figure 8  (A) The original discovery of revumenib. (B) The binding model of revumenib with human menin (PDB code: 7UJ4).

  下载: 全尺寸图片 幻灯片
  

  Figure 9  (A) The potential discovery of ensartinib. (B) The binding models of crizotinib and ensartinib with ALK (PDB code: 2XP2).

  下载: 全尺寸图片 幻灯片
  

  Figure 10  (A) The original discovery of aprocitentan. (B) The binding models of bosentan, macitentan and aprocitentan with ET_A receptor (PDB code: 8XVJ).

  下载: 全尺寸图片 幻灯片
  

  Figure 11  (A) The potential discovery of landiolol based on the structure of propranolol. (B) The binding models of propranolol and landiolol with β adrenergic receptor (PDB code: 1DY4 and 4AMI).

  下载: 全尺寸图片 幻灯片
  

  Figure 12  (A) The original discovery of resmetirom based on the structure of thyroid hormone (T3). (B) The binding models of T3, ALG-05509 and resmetirom with THR-β (PDB codes: 3GWS and 1N46).

  下载: 全尺寸图片 幻灯片
  

  Figure 13  (A) The original discovery of sofpironium bromide based on the structure of glycopyrronium bromide. (B) The binding model of sofpironium with mAChR3 (PDB code: 4DAJ).

  下载: 全尺寸图片 幻灯片
  

  Figure 14  The structure and binding models of givinostat with HDAC6 (PDB code: 6UOC).

  下载: 全尺寸图片 幻灯片
  

  Figure 15  (A) The structures of arimoclomol and miglustat. (B) The binding models of arimoclomol with Niemann-Pick C1 protein and HSP (PDB codes: 5U74 and 7K9S).

  下载: 全尺寸图片 幻灯片
  

  Figure 16  The structure of levacetylleucine.

  下载: 全尺寸图片 幻灯片
  

  Figure 17  (A) The structures of acoramidis and tafamidis. (B) The binding models of acoramidis and tafamidis with transthyretin (PDB codes: 4HIQ and 3TCT).

  下载: 全尺寸图片 幻灯片
  

  Figure 18  (A) The potential discovery of crinecerfont based on the structure of NBI35965. (B) The binding models of NBI35965 and crinecerfont with CRF1 receptor (PDB code: 8GTI).

  下载: 全尺寸图片 幻灯片
  

  Figure 19  (A) The structures of vanzacaftor, tezacaftor and deutivacaftor. (B) The potential discovery of vanzacaftor. (C) The binding models of elexacaftor and vanzacaftor with CFTR (PDB code: 8EIG).

  下载: 全尺寸图片 幻灯片
  

  Figure 20  (A) The potential discovery of vadadustat. (B) The binding model of vadadustat and HIF-pH (PDB code: 5OPC).

  下载: 全尺寸图片 幻灯片
  

  Figure 21  (A) The original discovery of danicopan. (B) The binding models of DAN-01 and danicopan with complement factor D (PDB code: 8DG6).

  下载: 全尺寸图片 幻灯片
  

  Figure 22  (A) The potential discovery of elafibranor based on the structure of bezafibrate. (B) The X-ray crystal structures of bezafibrate and elafibranor bind to PPAR (PDB codes: 7BPZ and 8HUQ).

  下载: 全尺寸图片 幻灯片
  

  Figure 23  (A) The potential discovery of seladelpar based on the structures of GW501516 and L-165,041. (B) The X-ray crystal structure of seladelpar binds to PPAR (PDB code: 8HUP).

  下载: 全尺寸图片 幻灯片
  

  Figure 24  (A) The original discovery of mavorixafor. (B) The binding models of MVA-02 and mavorixafor with CXCR4 (PDB code: 8U4P).

  下载: 全尺寸图片 幻灯片
  

  Figure 25  (A) The potential discovery of deuruxolitinib based on the structure of ruxolitinib. (B) The binding model of ruxolitinib with JAK2 (PDB code: 6VGL).

  下载: 全尺寸图片 幻灯片
  

  Figure 26  (A) The structures of xanomeline tartrate and trospium chloride. (B) The binding models of xanomeline and trospium with mAChR (PDB codes: 8FX5 and 6WJC).

  下载: 全尺寸图片 幻灯片
  

  Figure 27  (A) The potential discovery of ensifentrine based on the structure of trequinsin. (B) The X-ray crystal structures of trequinsin and ensifentrine bind to PDE3 (PDB code: 7L28).

  下载: 全尺寸图片 幻灯片
•