Novel Propofol Analogs: Design, Synthesis and Evaluation of Dihydrobenzofuran Derivatives as General Anesthetics

Abstract

Background: Propofol is used worldwide as a short-acting intravenous anesthetic in clinical practice; however, side effects such as injection pain and respiratory depression remain clinically relevant. Therefore, identification of safer propofol analogs is required. Method: In response to the urgent need for optimized potency and reduced side effects, a series of dihydrobenzofuran derivatives were designed as expectedly better propofol analogs through conformational restriction. A loss of righting reflex assay was conducted to evaluate the sedative/anesthetic properties of the synthesized compounds, and a respiratory depression test was performed for safety assessment. Results: Most of the designed compounds were shown to possess promising anesthetic properties as propofol analogs. The represented 53A had higher potency and a wider safety margin (ED₅₀:3.898 vs. 8.040 mg/kg in mice; 2.985 vs. 5.894 mg/kg in rats; TI (therapeutic index): 6.172 vs. 5.061 in mice; 4.362 vs. 2.580 in rats) than propofol, and fast onset and recovery times were maintained. The phosphate prodrug 56A also exhibited better efficiency and safety than fospropofol, along with a longer duration and faster recovery time in sedative profiles. Furthermore, alleviation of the adverse effects of respiratory depression has been demonstrated. Conclusions: 53A has the potential to be selected as a preclinical candidate for clinical development.

1. Introduction

Propofol (1), also known as 2,6-diisopropylphenol, was first reported in the 1980s as a structure-symmetrical small-molecule anesthetic [1]. It functions as a positive allosteric modulator of the γ-aminobutyric acid type A (GABAA) receptor, the allosteric mechanism of which can be characterized by potentiation of the receptor’s response to γ-aminobutyric acid at low concentrations and direct agonism at high concentrations. These changes lead to the opening of chloride ion channels, inducing neuronal hyperpolarization and ultimately inhibiting neurotransmission [2,3,4]. Nearly four decades after its initial approval, propofol remains the mainstay of short-acting intravenous anesthesia in clinical practice owing to its potent efficacy, rapid onset, and favorable recovery [2]. However, the inherent hydrophobicity of the scaffold leads to poor water solubility and requires formulation of lipid emulsions for clinical administration, which is closely related to injection pain. In addition, propofol inhibits the sensitivity of central chemo-sensors and reduces minute ventilation, leading to serious side effects, including hypotension, respiratory depression, and even apnea [5,6]. Consequently, structural optimization of propofol to mitigate these disadvantages has considerable clinical value.

Previous studies have indicated that the replacement of isopropyl groups with other alkyl substituents can enhance activity by increasing lipophilicity; however, excessive steric bulk leads to unexpected loss of activity. This effect may stem from the altered position of the phenolic hydroxyl group, which affects the formation or strength of the critical hydrogen bonds [7,8]. Based on the above findings, Pfizer and Haisco independently developed PF-0713 (2) and HSK3486 (3), respectively. Although the former achieved increased potency, unfavorable sedative effects and extended post-anesthesia recovery time were considered drawbacks, whereas compound 3 effectively alleviated injection pain and exhibited a superior cardiac safety profile regarding QT interval prolongation compared with propofol [9,10]. We hypothesized that the mitigation of injection pain derives from the use of only one-third or even lower dosages compared to propofol, which reduces the injection volume of the lipid-based emulsion. Significant research efforts have also focused on enhancing the aqueous solubility to circumvent the need for lipid emulsions. The incorporation of water-solubilizing groups is a clear strategy, as exemplified by the introduction of morpholine (4-1) and 1,4-oxazepine hexahydro rings (4-2) [11,12]. Prodrug strategies have also been extensively explored. Fospropofol (5), the earliest prodrug of propofol, was approved by the FDA in 2008 for endoscopic painless diagnosis and treatment. However, its advantages over propofol in short-duration endoscopic anesthesia are marginal, and due to safety concerns, the inconveniences resulting from the necessity for anesthesiologists are obvious, both leading to subsequent withdrawal [13]. Notably, fospropofol has recently been reintroduced in China for general anesthesia induction, which raises the question of whether its modified indications could create new opportunities. In addition, a series of propofol prodrugs designated as HX has also been developed by Sichuan University. HX0969w (6-1), similar to a sodium phosphate ester salt, exhibited slightly superior activity and therapeutic index compared to fospropofol [14,15]. However, both exhibited prolonged onset times; hence, carbonate-based prodrugs HX0921 (6-2) and compound 6-3 were further developed. Both feature a faster propofol release rate and reduced propofol exposure, resulting in a faster onset and shorter duration of action [16,17] (shown in Figure 1).

Clearly, the scaffold of the aforementioned derivatives retains the characteristic structure of 1,2,6-trisubstituted benzene. In this study, we propose a novel conformational restriction strategy involving the generation of a novel [5,6] fused bicyclic scaffold by cyclizing one of the alkyl substituents (see Figure 2). Restricted Conformation reduces the number of rotatable bonds, which may be advantageous for BBB permeability. On the other hand, the resulting dihydrobenzofuran scaffold has been documented in various compounds to exhibit anticancer, antioxidant and antibacterial activities, demonstrating its favorable drug-like properties [18,19]. Next, the scaffold was then partitioned into three zones (A, B, and C); group R¹ was incorporated into the 4-position in zone A to suppress phenol oxidation, while diverse hydrophobic substituents R²/R³ (zones B and C) were investigated to identify the optimal candidate compound.

2. Results and Discussion

2.1. Chemistry

The syntheses described herein can be referenced to the previously published patents [20], and the relevant operations and characterizations can also be obtained from the cited patents or the supporting information. As shown in Scheme 1, the synthesis of the target compound 14 commenced with 7-hydroxy-2,3-dihydro-1H-inden-1-one 7 as the starting material. The 4-position was selectively chlorinated using N-chlorosuccinimide (NCS) to block this site, thus enabling subsequent site-specific bromination at the 6-position to afford compound 9. The phenolic hydroxyl group was then protected as methoxymethyl (MOM) ether 10, in which the ketone was converted to the corresponding alkene via Wittig reaction, followed by Suzuki cross-coupling to yield diene intermediate 12. Finally, the diene was catalytically reduced using palladium under hydrogen, and the MOM protecting group was removed under acidic aqueous conditions to furnish the final compound, 14.

The synthesis of 27—29 and 33 began with key intermediate 20 (shown in Scheme 2). The route began with 2′,6′-Dihydroxyacetophenone (15), which was monoalkylated using ethyl bromoacetate. The resulting ester was hydrolyzed using lithium hydroxide to yield the corresponding carboxylic acid, which was then cyclized to construct the dihydrobenzofuran core via the intramolecular ketene cycloaddition [21].

Benzofuran 18 was hydrogenated and brominated using N-bromosuccinimide (NBS) to provide key intermediate 20. Then, 20 was subjected to MOM protection, followed by Suzuki cross-coupling, hydrogenation, and deprotection using a method similar to the synthesis of 14 to produce the target compound 27. Chlorination of 27 provided compound 28. Select-fluor was used to provide intermediate 24 by subjecting 20 to 4-fluorination, which was then processed using the same four-step sequence as described for 27, affording the final compound, 29. In addition, the MOM-protected intermediate 21 underwent Stille coupling with 1-ethoxyvinyl tri-n-butyltin to yield vinyl ether 30. Hydrolysis of the enol ether moiety yielded the corresponding methyl ketone intermediate 31, which was treated with cyclopropyl magnesium bromide and then underwent water elimination and hydrogenated to afford the target compound, 33.

Compounds 49–54 were synthesized using an identical route, which optimized the previously reported 10-step synthesis of compound 29. Starting with 4-fluoro-1,3-benzenediol 34, Friedel–Crafts acylation was employed to construct acetophenone, followed by bromination using NBS. The 3-hydroxy group of 36 underwent selective allylation with the relevant bromoalkene using silver carbonate. Compounds 37–39 were then subjected to radical cyclization under the action of AIBN and tributyltin hydride, and the resulting dihydrobenzofuran derivative was treated with the selected alkyl magnesium bromide. The tertiary alcohol functionality generated by the Grignard addition was deprotected by a sequential reaction with triethylsilane and TBAF to furnish the final products, 49–54 (see Scheme 3).

To determine the absolute configuration of the optimized compound, intermediate 41 was resolved to yield two enantiomers. The enantiomer with the shorter retention time is labeled with an asterisk (*), and the enantiomer with the longer retention time is labeled with two asterisks (**). Subsequent Grignard addition was performed, followed by dehydration of the tertiary alcohol, and then the substitution of the 53AB and 53CD’s phenolic hydroxyl with dibenzyl chloromethyl phosphate to afford compounds 55AB and 55CD, which were separately resolved to provide the four stereoisomers A, B, C, and D. Finally, reduction followed by conversion to sodium salt gave prodrugs 56A–56D, while acidic hydrolysis of the phosphate ester provided compounds 53A–53D (see Scheme 4). Further single-crystal X-ray diffraction (SCXRD) analysis was performed on 53C, which was selected because of the lowest activity and thus no further biological test value. Because neither the parent compound nor its phosphate ester readily formed suitable single crystals, the phenyl ethyl carbamate derivative was synthesized with the relevant isocyanates. These results unambiguously confirm the absolute configuration of 53C as (S, S). Consequently, the configuration of diastereomer 53D was determined to be (R, S). Furthermore, because the ¹H-NMR spectrum of 53A was identical to that of 53C, the configuration of 53A was determined to be (R, R), whereas 53B was (S, R) (shown in Figure 3).

2.2. Biological Evaluation

2.2.1. Anesthetic Properties Evaluations: Loss of Righting Reflex

The loss of righting reflex (LORR) assay is a classical experimental paradigm for evaluating the sedative/anesthetic properties of compounds [22]. Measuring key pharmacodynamic endpoints, such as onset time, duration of anesthesia, and recovery time, could facilitate preclinical characterization of anesthetic agents. All dihydrobenzofuran derivatives evaluated in this study were administered in solutions, and pharmacological testing was conducted in ICR mice using 5–8 distinct dose levels per compound. For each dosage, a minimum of three mice were independently dosed and monitored to ensure accurate observation (n ≥ 3). Immediately post-injection, the mice were positioned in dorsal recumbency, and the following temporal parameters were recorded: 1. Onset latency refers to the time interval from administration to the loss of righting reflex; 2. The duration of anesthesia represents the time interval from LORR onset to recovery of the righting reflex; 3. Recovery time refers to the time interval from righting reflex recovery to resumption of spontaneous ambulation, which signifies fully functional motor coordination characterized by stable, unassisted walking with gait and velocity within normal limits. The above data were tabulated under the minimal effective dose that induced LORR in 100% of the tested animals. Concurrently, the median effective dose (ED₅₀), defined as the dose producing LORR in 50% of the cohort, was derived through dose–response relationship modeling (see

Table 1. Evaluation studies on anesthetic profiles of compounds 14 and 27–29.

Cpd No.	X	Y	CLogP	ED50 (mg/kg)	Onset (s)	Duration (s)	Recovery (s)
1 a	/	/	3.93	16.0	16.2	94.7	32.3
14	H	C	4.23	56.0	25.5	232	77.0
27	H	O	3.46	13.0	12.4	65.4	35.6
28	Cl	O	4.39	52.5	18.0	662	326
29	F	O	3.82	26.5	11.7	71.0	9.00

^a The initial dose was based on preliminary test data. The dose was reduced by a fixed ratio (e.g., 1.25) if unexpected mortality occurred or increased by 1.25-fold in the absence of anesthesia. A total of 5–8 distinct dosages were conducted for each compound, and a minimum of three mice were independently dosed and monitored to ensure accurate observation for each dosage (n ≥ 3).

,

Table 2. Evaluation studies on anesthetic profiles of compounds 33, 49–52 and 52–54.

Cpd No.	R1	R2	X	CLogP	ED50 (mg/kg)	Onset (s)	Duration (s)	Recovery (s)
1	/	/	/	3.93	16.0	16.2	94.7	32.3
49 a	Me	/	/	4.34	80.0	16.0	229	104
50	Et	/	/	4.87	112	129	632	257
51	iPr	/	/	5.27	45.8	39.0	216	142
52	Cyclopropyl	/	/	4.78	32.0	178	1948	0
	/	di-Me	F
53	/	Me	F	4.26	9.32	15.7	287	59.3
54	/	Et	F	4.79	28.6	86.8	460	297
33	/	Me	H	3.90	10.8	9.67	62.7	52.3

^a The initial dose was based on preliminary test data. The dose was reduced by a fixed ratio (e.g., 1.25) if unexpected mortality occurred or increased by 1.25-fold in the absence of anesthesia. A total of 5–8 distinct dosages were conducted for each compound, and a minimum of three mice were independently dosed and monitored to ensure accurate observation for each dosage (n ≥ 3).

and

Table 3. Evaluation studies on anesthetic profiles of compounds 53A-53D and 56A-56D.

Cpd No.	C1	C2	ED50 (mg/kg)	Onset (s)	Duration (s)	Recovery (s)
1	/	/	16.0	16.2	94.7	32.3
53	/	/	9.32	15.7	287	59.3
53A a	R	R	4.00	11.0	372	38.5
53B	S	R	10.7	17.7	385	13.7
53C	S	S	32.0	32.0	114	4.30
53D	R	S	17.9	11.3	211	37.3
56A	R	R	33.5	269	354	355
56B	S	R	60.0	220	235	347
56C	S	S	170	233	400	119
56D	R	S	76.8	174	405	297

^a The initial dose was based on preliminary test data. The dose was reduced by a fixed ratio (e.g., 1.25) if unexpected mortality occurred or increased by 1.25-fold in the absence of anesthesia. A total of 5–8 distinct dosages were conducted for each compound, and a minimum of three mice were independently dosed and monitored to ensure accurate observation for each dosage (n ≥ 3).

).

Compared to the rapid onset and short recovery time of propofol, these two parameters were significantly prolonged in the cyclized derivative 14 despite an increased duration of action, which may be attributable to the higher administered dose required for efficacy. The replacement of the carbon atom with an oxygen atom to form a dihydrobenzofuran scaffold 27 restored propofol-like properties (rapid onset and short recovery) while achieving a lower ED₅₀, indicating superior potency. Considering the susceptibility of the phenolic moiety to potentially undesirable quinones during air oxidation, halogens were strategically introduced to block oxidation-prone positions. The resulting chloro-analog, 28, exhibited a pharmacological profile similar to 14, characterized by an increased dose. Conversely, the fluoro-substituted derivative 29 retained a favorable anesthetic profile, characterized by significantly shorter onset and recovery times than propofol. Furthermore, the introduction of halogens exerts an electron-withdrawing inductive effect, which facilitates the formation of hydrogen bonds at the phenolic hydroxyl groups, while an elevated CLogP may exert adverse impacts on drug absorption and metabolism. Thus, compound 28 shows an inferior ED₅₀ and a prolonged onset time relative to 27, which is presumably attributable to the reduced drug absorption ratio and rate caused by the increased CLogP. In contrast, the longer anesthesia and recovery durations may arise from the combined effects of enhanced binding activity and slowed metabolic rates. Similarly, the change in CLogP was relatively milder for fluorine substitutes, so the differences in ED₅₀ and onset time are less pronounced between compounds 29 and 27.

Encouraged by the aforementioned results, the impact of diverse alkyl substitutions in regions B/C on anesthetic activity was further explored based on the retained fluorinated dihydrobenzofuran scaffold. To circumvent the uncertainty introduced by multiple chiral centers, 3-dimethylated analogues 49–52 were initially synthesized to investigate the SAR in region B. Compound 49 exhibited a significant reduction in activity compared to 29, and replacing the methyl group with ethyl resulted in suboptimal activity. Intriguingly, a further increase in the steric bulk by introducing a propyl group led to resurgent activity; derivative 52 with a cyclopropyl group demonstrated the most potent activity. Although its onset time was 10-fold worse than that of propofol, its duration was 20-fold longer, and the mice were able to immediately resume normal ambulation. To investigate whether this characteristic was an anomaly, additional analogues 53–54 and 33 were synthesized; however, the immediate recovery phenomenon was not replicated. We propose a hypothesis herein that the prolonged anesthesia duration causes animals to unconsciously adopt a non-righting posture during recovery, analogous to the fact that some patients choose to rest for an extended period after anesthesia to complete their recovery process. When animals regain the righting reflex and return to a normal body posture, they have actually gone through the recovery period we intended to monitor, thus immediately regaining the ability of autonomous motor function. Therefore, the time of the 1900s is likely to represent the total duration encompassing both anesthesia and recovery. For all this, compound 53 still exhibited superior anesthetic properties compared with propofol, featuring a more favorable ED₅₀, prolonged anesthetic duration, and acceptable onset and recovery times. Constrained conformation reduces the number of rotatable bonds, the cyclopropyl group enhances hydrophobic interactions, and the introduction of fluorine augments hydrogen bonding while blocking metabolic sites. These factors are presumably responsible for the superior properties of compound 53. Furthermore, the interchange between methyl and ethyl groups or the removal of the fluorine substituent conferred no beneficial effects on the pharmacological profile.

Optically pure stereoisomers corresponding to 53′s four absolute configurations of compound 53 and their respective phosphonooxymethyl prodrugs were further evaluated for their anesthetic profiles. Among these, stereoisomer 53A showed the best potency, 53C was the least active, and enantiomeric pairs 53B and 53D exhibited comparable properties. Compared to both propofol and the racemic mixture 53, compound 53A exhibited 4-fold and 2-fold greater potency than either propofol or a racemic mixture 53. Furthermore, mice treated with 53A displayed a shorter onset time and a prolonged duration of anesthesia. The corresponding phosphonooxymethyl prodrugs generally exhibit reduced activity, requiring higher doses and potentially longer onset and recovery times. However, the differences in activity between the stereoisomers remained nearly consistent with those observed for the free parent drugs, with prodrug 56A (derived from 53A) being the most potent. Consequently, compounds 53A and 56A were selected for more detailed preclinical evaluation.

2.2.2. Predicted Binding Mode of Compound 53A Compared to Propofol

Leveraging the publicly available propofol-binding pocket on GABAA receptors (PDB ID: 6X3T) [23], compound 53A was subjected to virtual molecular docking using GLIDE to compare the binding modes of the two distinct scaffolds [24] (shown in Figure 4). Its binding pocket overlapped almost entirely with that of propofol, forming a critical hydrogen bond interaction with ILE-228, which is the core receptor–ligand interaction characteristic of this class of general anesthetics. The conformational constraint in 53A leads to increased rigidity, and the introduced oxygen atom is oriented toward the hydrophilic region, whereas the methyl group inserted into the hydrophobic pocket is replaced by the cyclopropyl group. All the key amino residues were conserved. Overall, docking results validated the binding mode of compound 53 at the molecular level.

2.2.3. Head-to-Head Pharmacodynamic Evaluation

Clarified solutions of 53A and propofol were administered to mice and rats using a dose-escalation protocol, and dose-dependent curves for the duration of LORR were established for both species (see Figure 5 and

Table 4. Evaluation studies on anesthesia profiles of compound 53A in rodents.

Species	Cpd No.	ED50 (mg/kg)	LD50 (mg/kg)	TI	HD5min (mg/kg)	HD10min (mg/kg)	HD15min (mg/kg)
ICR mice	53A a	3.898	24.06	6.172 b	8.930 c	14.26	19.58
	1	8.040	40.69	5.061	17.01	26.32	35.63
SD rat	53A	2.985	13.02	4.362	3.29	5.07	6.84
	1	5.894	15.21	2.580	7.12	11.32	15.52

^a The initial dose was based on preliminary test data. ICR mice: 5% Cremophor EL + 95% saline; SD rats: 7% Cremophor EL + 7% Solutol HS-15 + 86% saline. Five rodents were independently dosed and monitored for each dosage (n = 5). ^b TI = LD₅₀/ED₅₀. ^c HD_x denotes the dose required to produce X minutes of anesthetic.

). The results demonstrated that the dose of 53A required to achieve an equivalent anesthetic duration was approximately half that of propofol in both species. Furthermore, the ED₅₀ was calculated using the criterion of duration persisting for ≥30 s as the efficacy endpoint; the therapeutic index (TI) was then determined as the ratio of the ED₅₀ to the median lethal dose (LD₅₀); HD_x derived from the dose–response curve for sedation duration denotes the dose required to produce X minutes of anesthetic. Consistent with the duration results, the ED₅₀ values for 53A were approximately half those for propofol in both species.

Notably, a marked reduction in TI was observed specifically in rats, which was attributed to a significantly decreased LD₅₀, indicating greater sensitivity of rats to this class of compounds. Given that the rats demonstrated fewer twitches and tremors during anesthesia induction, less interference with nociceptive reflex assessments and a more reliable determination of LORR onset was expected. Therefore, subsequent experiments were conducted in a rat model, thereby enhancing the stability and reliability of the experimental results to a certain extent.

In clinical practice, propofol is usually administered as a lipid emulsion formulation, which is associated with injection pain and hyperlipidemia and may also induce skeletal muscle injury [25]. Strategies such as lipid-based self-nanoemulsifying drug delivery systems (SNEDDS) and mixed polymeric micelles have been proposed to address formulation-related issues [26,27]. On the other hand, the probability of emulsion mitigating propofol’s side effects on heart rate and respiration has also been investigated [28]. Therefore, the dose–response experiment in rats using the lipid emulsion formulation was replicated (shown in Figure 6 and

Table 5. Evaluation studies on anesthesia profiles of compound 53A in rats administered via lipid emulsion.

Cpd No.	ED50 (mg/kg)	LD50 (mg/kg)	TI	HD5min (mg/kg)	HD10min (mg/kg)	HD15min (mg/kg)
53A a	4.119	20.77	5.042 b	5.18 c	8.04	10.09
1	7.105	31.59	4.446	11.07	16.30	21.54

^a The initial dose was based on preliminary test data. Five SD rats were independently dosed and monitored for each dosage (n = 5). ^b TI = LD₅₀/ED₅₀. ^c HD_x denotes the dose required to produce X minutes of anesthetic.

). The results demonstrated that the LD₅₀ was approximately double that obtained with the solution formulation, and the TI increased correspondingly. These changes were more pronounced with propofol than with compound 53A, suggesting that propofol’s adverse effect profile may be more serious, and further experiments are required to substantiate this.

As a prodrug, fospropofol has been reapproved for general anesthesia induction, which significantly mitigates the injection pain caused by emulsion formulations and offers a longer duration of anesthesia. However, it requires recommended doses up to 6.5–20 mg/kg, compared to 1.5–2.5 mg/kg for propofol, which may pose potential risks [29,30]. The efficacy and safety of the phosphate ester prodrug 56A, based on the parent compound 53A, were evaluated in a head-to-head comparison with fospropofol (see Figure 7 and

Table 6. Evaluation studies on anesthesia profiles of prodrug 56A in rats.

Cpd No.	ED50 (mg/kg)	LD50 (mg/kg)	TI	Latency (min)	HD20min (mg/kg)	HD40min (mg/kg)	HD60min (mg/kg)	Recovery (min)
56A a	13.6	90.0	6.62 b	2.05	15.73 c	34.14	52.54	8.82
5	37.4	178.8	4.78	2.02	61.16	110.38	159.61	10.53

^a The initial dose was based on preliminary test data. Five SD rats were independently dosed and monitored for each dosage (n = 5). ^b TI = LD₅₀/ED₅₀. ^c HD_x denotes the dose required to produce X minutes of anesthetic.

). The study demonstrated comparable latency times between the two agents; however, rats administered 56A exhibited slightly faster recovery times after anesthesia. Furthermore, the effective dose of 56A was only one-third to one-fourth of that of fospropofol, indicating superior efficacy and lower phosphate usage. Although transient phosphate intake does not impose a phosphate burden on healthy individuals, the majority of patients receiving anesthetics clinically are likely to have underlying diseases. For this population group, especially those with chronic kidney disease, hypoparathyroidism, or severe bone disorders, the reduced phosphate intake can greatly mitigate the risk of phosphate burden. Additionally, 56A revealed a better TI, signifying a wider safety margin, and its enhanced clinical utility was promising.

2.2.4. Safety Assessment on Respiratory Depression

The safety assessment focused on respiratory depression with a significant acute risk. To align with clinical application, the test compounds were administered as emulsions at a dose of 2x ED₅₀, with saline serving as the blank control and lipid emulsion as the negative control (shown in Figure 8). Administration of only lipid emulsion was found to increase the respiratory rate in rats, but both compound 53A and propofol induced depression. However, a distinct difference in recovery was observed: inhibition was immediately alleviated following recovery (~10 min after injection) in rats administered 53A, whereas propofol-induced respiratory rate depression persisted. The tidal volume was reduced after the administration of both the lipid emulsion and the tested drugs, but recovered in the drug groups after animal recovery. Notably, 53A was observed to potentially induce a mild compensatory effect, resulting in a post-recovery tidal volume higher than that in the emulsion group, and maintaining levels comparable to those of the saline control. Consequently, compound 53A effectively improved post-dose-minute ventilation, restoring it to saline control levels within 10 min of administration, which may be the result of a combination of reduced production of quinone metabolites and lower use of fat emulsion preparations. Additionally, respiratory depression is a major high-risk concern in the clinical use of anesthetics, particularly for patients with underlying cardiovascular diseases or the elderly. Compared to propofol, respiratory depression induced by 53A is less pronounced in intensity and shorter in duration, especially since the rapid recovery of ventilation exerts a prominent safety effect on this population and thus confers greater clinical value.

3. Materials and Methods

General Procedures in Animal Studies. Male Sprague Dawley rats (200–300 g) and ICR mice (20–30 g) involved in biology experiments were purchased from Shanghai Ji-hui Experimental Animal Breeding Co., Ltd. Animal housing, care protocols, and study procedures were conducted in accordance with Guide for the Care and Use of Laboratory Animals (8^thed., 2023). All experimental protocols were reviewed and approved by Shanghai Shujing Biopharma Co., Ltd. (protocol number: SHSJ-IACUC A-013). The rodents were housed at a suitable temperature (23 ± 3 °C) and humidity (60 ± 10%) under an automatically controlled 12/12 h light/dark cycle (8:00–20:00), and food and water were available ad libitum. Every effort was made to minimize animal suffering.

Loss of Righting Reflex Assay. The test compounds were formulated into solutions or emulsions and serially diluted to establish dose gradients. The initial dosing for each was based on the preliminary test data. The dose was reduced by a fixed ratio (e.g.,1.25) if mortality occurred or increased by 1.25-fold in the absence of anesthesia. The time at which the injection was completed was defined as T0. The animals were positioned in dorsal recumbency on a thermostatically controlled heating pad, and the time of failure to spontaneously regain a prone position (remaining supine or lateral) within 30 s was recorded as T1. Upon attenuation of the anesthetic effect, the time at which spontaneous self-righting occurred was recorded as the T2. Animals typically exhibited transient ataxia, and the time required for coordinated quadrupedal ambulation was recorded as T3. Key temporal parameters were calculated as follows: LORR latency = T1–T0, LORR duration = T2–T1, and full recovery time = T3–T2. The ED₅₀ or LD₅₀ was calculated based on the dose–response curve depicting the proportion of rodents exhibiting anesthetic effects or mortality per dose group. HD_x, derived from the dose–response curve for sedation duration, denotes the dose required to produce X minutes of anesthesia.

Molecular Docking Procedures. The crystal structure of the target protein in complex with its ligand was retrieved from the RCSB Protein Data Bank(PDB) and prepared using Schrödinger Maestro 10.1. The complex structure was preprocessed through steps including optimization of receptor connectivity, adjustment of protonation states at physiological pH, and removal of water molecules beyond 5 Å from the ligand. Subsequently, the receptor was separated from all bound ligands, and the specific protein–target ligand pair was selected. A binding site was defined around the geometric center of the target ligand using the Grid generation method, which yielded the final prepared ligand-binding pocket. In parallel, the structure of the test compound was imported into Maestro and subjected to appropriate preprocessing, followed by conformational sampling to generate five or more low-energy molecular conformations. Virtual docking was then performed between the prepared binding pocket and the test compound. The resulting poses were analyzed to identify key interaction forces and binding modes, with each pose assigned to a docking score. Higher absolute score values indicate greater pose stability and predicted binding affinity. The pose with the highest score was selected and visualized using PyMOL 3.0, followed by graphical refinement to produce a clear and intuitive representation of the receptor–ligand docking result.

Respiratory Evaluation. One day before the experiment, the animals were placed in whole-body plethysmograph boxes for 60 min of adaptation. Respiratory data were collected for grouping. During the experiment, the animals were placed in whole-body plethysmograph boxes, and basal data were collected for 60 min. Vehicles and test compounds (2×ED₅₀) were intravenously injected within 10s. The animals were put back into boxes, and respiratory data were collected for another 60 min. Respiratory frequency, tidal volume, and minute ventilation were analyzed every 5 or 10 min.

Synthetic Procedures in Chemistry. The reagents and starting materials were used as received without further purification. Preparative column chromatography was performed using 200−300 mesh silica. LC-MS data were recorded on Shimadzu LCMS-2020 instrument with a Waters Sunfire C18 (3.5 μm, 50 × 4.6 mm). NMR spectra were measured on Bruker Avance 400 and Avance 600 spectrometers (400 MHz or 600 MHz for ¹H NMR) with tetramethylsilane (TMS) as the internal standard, and chemical shifts were expressed in parts per million (ppm, δ units). All moisture- and air-sensitive reactions were conducted under a dry nitrogen atmosphere.

The detailed synthetic procedures and characterizations of the relevant compounds are available in the patent [20] and the Supporting Information.

4. Conclusions

While propofol is a cornerstone intravenous anesthetic, its clinical utility is limited by significant side effects, most notably injection pain and respiratory depression. This creates a clear and ongoing need for novel agents that retain their favorable pharmacokinetic profile while improving their therapeutic index. In this study, we address this need through a rational design strategy based on conformational restrictions. We report the synthesis and comprehensive pharmacological evaluation of a novel series of dihydrobenzofuran derivatives as propofol analogs. Our key findings demonstrate that this structural modification successfully decouples anesthetic efficacy from adverse effects. The lead compound, 53A, has higher potency while maintaining a faster onset and recovery time compared to propofol. The anesthesia evaluation of water-soluble prodrug 56A also demonstrated better efficiency than Fospropofol, along with a longer duration and faster recovery time in the sedative profiles. Collectively, compound 53A and its prodrug exhibited 2- to 4-fold greater anesthetic potency than the reference agent, accompanied by a wider margin of safety. Furthermore, a milder inhibitory effect was observed on the evaluation of respiratory depression. 53A could be developed as a potential preclinical candidate compound for application in general anesthetics.

5. Patent

A part of this work, containing syntheses and characterizations, has been filed as a patent application, WO_2023011634_A1.