Comparing LJ-4378 and tirzepatide effects on obesity and weight regain in mice

Animal experiments were conducted in accordance with the guidelines approved by the Institutional Animal Care and Use Committees (IACUC) of Seoul National University (SNU-230113–5-2, SNU-250410–1, and SNU-250811–5). C57BL/6 J male mice and Sprague–Dawley (SD) male rats were raised in a controlled animal facility with a temperature of 22 ± 1 °C and subjected to a 12-h light/12-h dark cycle. Mice were randomly assigned to experimental groups and fed either a normal chow diet (NCD, protein: 24.52% calories, carbohydrates: 63.07% calories, fat: 12.41% calories, 38,057, Purina Lab, Seongnam, Republic of Korea) or a high-fat diet (HFD, protein: 20% kcal, carbohydrate: 20% kcal, fat: 60% kcal, New Brunswick, D12492, Research Diets, NJ, USA) for the specified durations, according to group assignment. Both diet and water were provided ad libitum. SD rats (230–240 g) were provided standard rat chow and had free access to water. Prior to the experiment, the rats underwent a 12-h fasting period.

To evaluate the anti-obesity effect in vivo, mice fed either NCD or HFD for 10 weeks were subsequently treated for up to 4 weeks. To control for injection route-related variables, all mice received both intraperitoneal (I.P.) and subcutaneous (S.C.) injections. The administration were as follows: LJ-4378 treated mice received LJ-4378, which was synthesized in our laboratory at > 95% purity (1 mg/kg/day, I.P.) (Hou et al. 2010; Kim et al. 2022) along with the vehicle for TZP (40 mM Tris–HCl, pH 8, S.C.); TZP treated mice received TZP (Samms et al. 2021; Kim et al. 2022) (LY3298176, MedChemExpress, Monmouth Junction, NJ, USA) (10 nmol/kg/day, S.C.) along with the vehicle for LJ-4378 (0.2% DMSO, I.P.); vehicle control mice received both vehicles (0.2% DMSO, I.P. and 40 mM Tris–HCl, pH 8, S.C.)

To assess post-treatment weight rebound (WR), the mice previously treated with vehicle, LJ-4378, or TZP were maintained on HFD for an additional 4 weeks following treatment cessation. Treatment protocols for the WR groups were identical to those of the HFD groups. In total, nine experimental groups were established: NCD + CTL, NCD + LJ, NCD + TZP; HFD + CTL, HFD + LJ, HFD + TZP; and WR + CTL, WR + LJ, WR + TZP. The volume of oxygen consumed (VO₂), carbon dioxide produced (VCO₂), respiratory exchange ratio (RER), energy expenditure, activity, and food intake were measured with PhenoMaster (TSE Systems, Bad Homburg, Germany). Indirect calorimetry was performed under controlled conditions, with a stable ambient temperature of 24 °C and a 12 h light/12-h dark cycle. Mice fed HFD were administered vehicle, LJ-4378, or TZP for 14 days, and the last 2 days of treatment were conducted during indirect calorimetry analysis. The WR model was examined by indirect calorimetry after 2 and 4 weeks of the weight regaining phase.

Additionally, food consumption was measured every 2 days during the 14-day treatment period and for an additional 16 days post-administration.

Cytochrome 450 (CYP) enzyme inhibition was evaluated using pooled human liver microsomes (Corning, #452117). The assay included the following reagents: ketoconazole (CYP3A4 selective inhibitor, Sigma Aldrich, K1003), Nicotinamide adenine dinucleotide phosphate (NADPH) regeneration system (Promega, V9510), terfenadine as the internal standard (Sigma Aldrich, T9562), and 0.1 M potassium phosphate buffer (pH 7.4, Corning, #451201). Substrates and corresponding metabolites for each CYP isoform were as follows: phenacetin and acetaminophen for CYP1A2 (Sigma Aldrich, #77440 and A7085, respectively), diclofenac and 4'-hydroxydiclofenac for CYP2C9 (Sigma Aldrich, D6899; TRC, H825225), S-mephenytoin and 4'-hydroxymephenytoin for CYP2C19 (Sigma Aldrich, UC175; Santacruz, sc210197), dextromethorphan and dextrorphan for CYP2D6 (Sigma Aldrich, #81091 and UC205), and midazolam and 1'-hydroxymidazolam for CYP3A4 (KFDA; Sigma Aldrich, UC430).

The reaction mixture, containing human liver microsomes (0.25 mg/mL), phosphate buffer, a cocktail of the five CYP substrates, and the test compound LJ-4378 (0 or 10 μM). Samples were preincubated at 37 °C for 5 min, and the reaction was initiated by the addition of the NADPH regeneration system, followed by incubation at 37 °C for 15 min. The reaction was terminated by adding acetonitrile containing the internal standard, and centrifuged at 15,000 rpm for 5 min at 4 °C. The supernatant was analyzed by LC–MS/MS using a Nexera XR system (Shimadzu, Japan) coupled with a TSQ Vantage mass spectrometer (Thermo, USA). Chromatographic separation was performed on a Kinetex C18 column (2.1 × 100 mm, 2.6 μm; Phenomenex, USA) with a gradient elution of water (A) and acetonitrile (B), both containing 0.1% formic acid. Metabolites were quantified in multiple reaction monitoring (MRM) mode with Xcalibur software (version 4.4).

All experiments were performed in duplicate with variability ≤ 15%. The percent activity of each CYP enzyme was calculated based on metabolite formation relative to the no-inhibitor control. Ketoconazole (0.1 μM) was used as a positive control for CYP3A4 inhibition, and the resulting activity (25% ± 10%) was consistent with internal validation criteria.

CHO cells (ATCC #CRL-61) stably expressing hERG potassium channels from Sophion Biosciences were used for this test. The cells were cultured in a humidified incubator with 5% CO₂ at 37 °C. CHO cells that were at least two days post-plating and more than 75% confluent were used for experiments. Before testing, cells were harvested using TrypLE and resuspended in physiological solution at RT. The physiological and external solutions were prepared at least one month in advance. The intracellular solution was prepared in batches, aliquoted, and stored at 4 °C until use.

Stock solutions of the test compound were prepared in 100% DMSO at 10 mM. These stock solutions were then diluted into the external solution to achieve the final concentrations of 1, 10, and 30 μM. A visual inspection for precipitation was performed before testing. The final DMSO concentration in the extracellular solution did not exceed 0.30%.

The hERG SyncroPatch (384PE) assay was conducted at RT. The setup, prime chip, cell catch and seal, amplifier settings, voltage, and application protocols were controlled using Biomek software (Nanion). A 40 μL addition of vehicle was first applied, followed by a 300-s baseline period. Then, 40 μL of the test compound at the desired concentration was added. Exposure time for each test concentration was no less than 300 s. The entire recording process was subjected to quality control, and wells failing to meet criteria were automatically excluded and retested using PatchControl software.

Each compound was tested at three concentrations (1, 3, and 10 μM), with a minimum of two replicates per concentration.

Serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), blood urea nitrogen (BUN), Creatinine, Creatine kinase MB (CKMB), and lactate dehydrogenase (LDH) levels were analyzed with Fujifilm Dri-Chem 3500 s (Fujifilm, Kanagawa, Japan) according to the manufacturer's instructions. The slides GOT/AST-P III, GPT/ALT-P III, BUN-P III, CRE-P III, CKMB-P, and LDH-P III were used.

Serum free fatty acid (FFA) levels were quantification using HR Series NEFA-HR (2) (Color Reagent A (#999–34691), Solvent A (#995–34791), Color Reagent B (#991–34891), Solvent B (#993–35191), and NEFA Standard Solution (#276–76491) (FUJIFILM Wako Pure Chemicals, Osaka, Japan), following the manufacturer's instruction.

The serum TNF-α level was measured using TNF-alpha ELISA MTA00B (R&D system) in accordance with the manufacturer's protocol.

Body composition and micro-CT analysis were performed after the treatment with vehicle, LJ-4378, or TZP for the indicated duration in HFD-fed mice and WR mice.

A nuclear magnetic resonance (NMR) technology system with EchoMRI-700 (Echo Medical Systems, Nanded, India) was used to analyze body composition.

To assess the abdominal fat volume, we employed the Quantum GX2 micro-CT imaging system (PerkinElmer, Waltham, MA, USA). Mice were anesthetized with isoflurane and positioned on the micro-CT bed for imaging of the abdominal region, defined from the proximal end of the L1 vertebra to the distal end of the L5 vertebra. Imaging was performed at 90 kV/88 µA with a field of view of 72 mm, a pixel size of 90 µm, and a 2-min acquisition time at standard resolution. Visceral, subcutaneous, and total abdominal fat volumes were quantified using AnalyzeDirect software (ver. 12.0, Overland Park, KS, USA).

The glucose tolerance test was conducted on mice fasted for 12 h with ad libitum access to water. Mice were intraperitoneally injected with 20% D-( +)-glucose (2 g/kg body weight, #49139, Sigma), and blood glucose levels were measured at each time point using a Gluco Doctor Top meter and strip (Allmedicus, South Korea).

Adipose tissues were fixed in 10% formalin (HT50128, Sigma) for 2 days and made into paraffin slides. The deparaffinized sections were stained with ClearView Staining Hematoxylin (MA010081, BBC Biochemical, Mount Vernon, WA, USA) and Eosin Y Alcoholic (#3610, BBC Biochemical) solution. The crown-like structure formation of gonadal white adipose tissue (gWAT) and triglyceride (TG) accumulation in the liver are observed using hematoxylin and eosin staining on paraffin slides. The images were captured using Nikon Elements (NIS BR Analysis ver 5.10.00).

Immunohistochemistry was performed on deparaffinized gWAT sections using an anti-F4/80 antibody (1:500, MCA497GA, Bio-rad). After overnight incubation with the primary antibody, sections were incubated for 1 h at RT with secondary antibody: Donkey anti-Rat 488 Alexa Fluor™ 488 (1:500, A21208, Invitrogen). Lastly, DAPI (1:1000, D9542, Sigma) was incubated for 4 min at RT to stain nuclei. Images were acquired using a Leica THUNDER Imager 3D (Leica Microsystems, Buffalo Grove, IL, Germany). Image analysis and F4/80 signal quantification were performed using LAS X software (ver. 3.6.0) and ImageJ.

Hepatic TG content was measured following a protocol (Lee et al. 2024). Briefly, 100 mg of liver tissue was digested overnight at 55 °C in 300 μL of ethanolic KOH solution (Ethanol: 30% KOH = 2:1). The digested sample was then adjusted to a total volume of 1000 μL with 50% ethanol and centrifuged at 1000 rpm for 5 min. The supernatants were transferred to a new tube, and 50% Ethanol was added to bring the final volume to 12,000 μL. The 200 μL of the sample was then mixed with 215 μL of 1 M MgCl₂, incubated on ice for 10 min, and centrifuged. The supernatant was collected for TG quantification. Triolein level was measured using glycerol reagent (F6428, Sigma) according to the manufacturer's instructions. Hepatic TG content was expressed as triolein equivalents.

The plasma protein binding of LJ-4378 was evaluated using the Rapid Equilibrium Dialysis (RED) method with 100% plasma from human (Biochemed, BC22089NCP), rat (Innovative, IRT-N), and mouse (Biochemed, 029-APSC-MP), all treated with sodium citrate as an anticoagulant. The RED device (ThermoFisher Scientific, #89809) and reusable base plate (ThermoFisher Scientific, #89811) were used with an 8 K MWCO membrane. Test compounds were added to plasma at a final concentration of 10 μM and dialyzed against phosphate-buffered saline (PBS, pH 7.4; Hyclone, #SH30256.01) at 37 °C for 4 h. After incubation, equal volumes of plasma and PBS samples were mixed with blank PBS and blank plasma, respectively, followed by the addition of acetonitrile containing chlorpropamide (TRC, C424800) as the internal standard. The mixtures were centrifuged at 15,000 rpm for 5 min at 4 °C, and the supernatants were analyzed using LC–MS/MS. Chromatographic separation was performed on a Kinetex C18 column (2.1 × 100 mm, 2.6 μm; Phenomenex) using a gradient elution of water with 0.1% formic acid (A) and acetonitrile with 0.1% formic acid (B), on a Nexera XR system (Shimadzu, Japan) coupled to a TSQ Vantage mass spectrometer (Thermo, USA) operating in MRM mode with Xcalibur 4.4 software. All measurements were performed in duplicate with a deviation of less than 15% between replicates. The free fraction (% Free) was calculated as (concentration in buffer chamber/concentration in plasma chamber) × 100, and the bound fraction (% Bound) was obtained by subtracting % Free from 100. In cases where measurements were unstable, deviations greater than 15% were allowed.

The plasma stability of LJ-4378 was evaluated using 100% plasma obtained from human (Biochemed, BC22089NCP, sodium citrate), rat (Innovative, IRT-N, sodium citrate), and mouse (Biochemed, 029-APSC-MP, sodium citrate). 10 Μm of LJ-4378 was added to the plasma, and samples were incubated at 37 °C for 0, 30, and 120 min. At each time point, an aliquot of the plasma was removed and mixed with acetonitrile containing the internal standard, chlorpropamide (TRC, C424800). The mixture was vortexed for 5 min and centrifuged at 15,000 rpm for 5 min at 4 °C. The supernatant was injected into the LC–MS/MS system for quantification of the remaining compound. Analysis was performed using a Nexera XR system (Shimadzu, Japan) with a TSQ Vantage mass spectrometer (Thermo, USA). Chromatographic separation was achieved using a Kinetex C18 column (2.1 × 100 mm, 2.6 μm particle size; Phenomenex, USA) with a gradient elution of 0.1% formic acid in water (mobile phase A) and 0.1% formic acid in acetonitrile (mobile phase B). Data acquisition was conducted in MRM mode using Xcalibur software (version 4.4). All experiments were performed in duplicate with a deviation within 15%, and the percentage of compound remaining (% Remaining) at each time point was calculated relative to the non-incubated control sample. For quality control, procaine and enalapril (each at 10 μM) were used as reference compounds and were confirmed to fall within acceptable ranges.

The pharmacokinetics of LJ-4378 were evaluated in male C57BL/6 J mice following a single administration by either the oral (P.O.) or intravenous (I.V.) route. For P.O., LJ-4378 was administered at 10, 20, or 50 mg/kg in a vehicle composed of 0.5 mg/mL L-ascorbic acid, 5% DMSO, 20% PEG400, 20% Tween 80, and 55% of 0.1% copovidone. For I.V. administration, a 2 mg/kg dose was delivered via tail vein injection in a vehicle consisting of 5% DMSO, 25% PEG400, and 70% sterile water. Blood samples (200 μL) were collected at 5, 15, and 30 min and 1, 2, 4, 8, 12, and 24 h post-dose for the P.O. groups, and at 2, 5, 15, and 30 min and 1, 2, 4, 8, and 12 h post-dose for the I.V. group. Plasma was separated by centrifugation, and the concentrations of LJ-4378 were determined using a validated LC–MS/MS method. Pharmacokinetic parameters were calculated using non-compartmental analysis with Phoenix WinNonlin (v8.3), and included C_max, T_max, AUC_last, AUC_inf, t_1/2, and clearance (Maclean et al. 2011). Oral bioavailability was calculated using dose-normalized AUC_inf values.

In this study, steady state tissue partition coefficients (, i.e., the ratio of tissue to plasma concentrations in steady state) Eq. () were determined for LJ-4378 in SD rats. In our preliminary study involving I.V. bolus administration of LJ-4378 to mice, the terminal phase half-life was determined to be approximately 0.29 h. Therefore, to ensure a steady state condition, rats received I.V. infusion (i.e., infusion solution, 5% DMSO in normal saline; delivery rate, 10 µL/min; infusion rate, 0.31 µg/µL; total dose of LJ-4378, 0.837 mg per rat) for 270 min (i.e., the time exceeding five times the half-life). Blood samples (~ 150 µL) were collected from the animal at every 10 min from 240 to 270 min for the confirmation of the steady state of the compound. At completion of the last blood collection, rats were sacrificed by decapitation, and four representative tissues (i.e., gonadal adipose, brain, lung, and liver) were harvested. Tissues were then washed four times with ice-Dulbecco's phosphate-buffered saline (DPBS), weighed, and homogenized in two volumes of DPBS using an Ultra Turrax homogenizer (IKA, Staufen, Germany). Aliquots (50 µL) of plasma and tissue homogenate were mixed with 200 µL ice-cold acetonitrile containing propranolol (100 ng/mL, internal standard). After vortex mixing for 10 min, the mixture was centrifuged at 16,100 × g for 5 min at 4 °C, and 10 µL of the supernatant was injected directly onto the HPLC–MS/MS system. The concentration of LJ-4378 in the biological sample was quantified on an Agilent 1200 HPLC coupled to an API 3200 mass spectrometer with a Phenomenex Luna C18(2) column (3 µm, 100 Å, 50 × 2 mm). Chromatography was performed under isocratic conditions with 80% acetonitrile containing 0.1% formic acid (mobile phase A) and 20% water containing 0.1% formic acid (mobile phase B) at a flow rate of 0.4 mL/min. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T}$$\end{document} K P T , 1

When calculating \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T}$$\end{document}KP,T for LJ-4378 in SE rats, Eq. (1) was used:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T} = \frac{{C_{T,ss} }}{{C_{p,ss} }}$$\end{document}KP,T=CT,ssCp,sswhere \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{T,ss}$$\end{document}CT,ss and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{p,ss}$$\end{document}Cp,ss represented the tissue and plasma concentrations at 270 min after the initiation of LJ-4378 infusion. In addition, under the free-drug hypothesis, the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T}$$\end{document}KP,T can be expressed as the ratio of the unbound fractions in plasma (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$f_{up}$$\end{document}fup) and tissue (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$f_{u,T}$$\end{document}fu,T) [Eq. (2)]. Given similar interspecies values of tissue unbound fraction (Poulin and Theil 2000), the rat \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T}$$\end{document}KP,T values were extrapolated to those in mice and humans using the following Eq. (3):2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T} = \frac{{f_{up} }}{{f_{u,T} }}$$\end{document}KP,T=fupfu,T3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T,extrapolated} = K_{P,T,rat} \times \frac{{f_{up} }}{{f_{up,rat} }}$$\end{document}KP,T,extrapolated=KP,T,rat×fupfup,rat

We conducted the statistical analysis using GraphPad Prism 8 (Software Inc., San Diego, CA, USA). Data are presented as mean ± standard errors of the means (SEMs). Statistical significance between two groups was determined using an unpaired two-tailed t-test. For comparisons involving more than two groups, one-way or two-way ANOVA was applied, followed by Bonferroni post hoc tests to adjust for multiple comparisons. The specific statistical tests used for each dataset are indicated in the corresponding figure legends.

Before conducting in vivo preclinical studies, we evaluated the safety profile of LJ-4378. First, we assessed cytochrome 450 (CYP450) inhibition and human ether-à-go-go-related gene (hERG) channel blockade. LJ-4378 exhibited minimal inhibitory activity against the major human CYP isoforms when tested at a concentration of 10 μM. Residual enzymatic activities were maintained at 82.2% (CYP1A2), 95.8% (CYP2C9), 85.3% (CYP2C19), 89.3% (CYP2D6), and 96.9% (CYP3A4), indicating low potential for drug-drug interactions (Table). The hERG inhibition assay revealed a half-maximal inhibitory concentration (IC50) value greater than 30 μM, suggesting a low risk of cardiotoxicity. Furthermore, in vivo analyses of hepatic, renal, and cardiac injury markers revealed no evidence of organ toxicity, supporting the favorable safety profile of LJ-4378 (Fig.). S1 S1

Collectively, these findings demonstrated the weight-reducing effects of LJ-4378 and TZP. Notably, despite differences in total weight loss, similar reductions in fat mass suggest that LJ-4378 may preferentially target fat mass while preserving lean tissue. To compare the fat-specific effects of LJ-4378 with those of TZP under equivalent total weight loss, we extended the LJ-4378 treatment until the body weight matched that of the 2-week TZP group. Four weeks of LJ-4378 administration produced a comparable reduction (15.5%) in body weight compared to 2 weeks of TZP treatment (Fig. S2a, b, d). Notably, LJ-4378 exhibited a lower fat ratio (20.1% of the total body weight) compared to 2 weeks of TZP treatment (27.4% of the total body weight) (Fig. S2c, e).

Diet-induced obesity is characterized by chronic low-grade inflammation in adipose tissue, marked by increased macrophage infiltration in adipose tissue and the production of pro-inflammatory cytokines, which can be reduced by weight loss (Hildebrandt et al. 2023). Consistently, both LJ-4378 and TZP significantly reduced macrophage accumulation in gWAT, as indicated by decreased F4/80 protein expression and fewer crown-like structures (CLS) compared with HFD controls (Fig. 2c, d). This reduction was accompanied by a shift in gene expression towards an anti-inflammatory profile (Fig. 2e). Expression of adhesion G protein-coupled receptor E1 (Adgre1) and tumor necrosis factor-alpha (Tnf-α) was remarkably suppressed by both treatments, suggesting reduced pro-inflammatory signaling. Additionally, serum TNF-α levels were reduced in LJ-4378 and TZP-treated HFD-fed mice, indicating attenuation of both adipose tissue and systemic inflammation (Fig. S3). Notably, markers of M2-like macrophages, including arginase 1 (Arg1), C-type lectin domain containing 10A (Clec10a), chitinase-like 3 (Chil3), and interleukin 10 (Il-10), were selectively upregulated in the LJ-4378 group, indicating the promotion of anti-inflammatory macrophage polarization (Fig. 2e). Together, these data demonstrate that both LJ-4378 and TZP alleviate obesity-related metabolic dysfunctions, including glucose intolerance and gWAT inflammation.

To further elucidate the metabolic effects of LJ-4378 and TZP, we examined the expression of lipolytic and thermogenic regulators in the adipose tissue (Fig. 3c, d). In BAT, HFD-fed mice treated with either compound led to elevated levels of phosphorylated hormone-sensitive lipase (P-HSL), phosphorylated cAMP response element-binding protein (P-CREB), uncoupling protein 1 (UCP1), and cytochrome c oxidase subunit 4 (COXIV) (Fig. 3c). A similar pattern of upregulation was observed in iWAT, suggesting that both LJ-4378 and TZP stimulated adipose tissue metabolism under obese conditions (Fig. 3d). Notably, UCP1 expression in BAT was more strongly induced by LJ-4378 than TZP, suggesting a more pronounced thermogenic response (Fig. 3d). This effect was likely attributable to the direct activation of PKA signaling by LJ-4378 (Kim et al. 2022).

Consistent with these findings, the proteins involved in mitochondrial oxidative phosphorylation were upregulated in HFD-fed mice treated with either compound alone (Fig. S5a, b). Despite increased lipolysis, serum free fatty acid (FFA) levels remained unchanged, consistent with enhanced mitochondrial oxidative phosphorylation, suggesting the efficient utilization of lipolysis-derived substrates for energy metabolism. (Fig. S5c).

To evaluate whether the observed effects extended beyond adipose tissue, we examined hepatic PKA signaling. Neither LJ-4378 nor TZP altered the phosphorylation of the downstream targets of PKA in the liver (Fig. S6). The metabolic improvement observed with treatment is likely attributable to its anti-obesity effects and consequent improvement in adipose tissue insulin sensitivity (Lee et al. 2023). However, the direct effects of LJ-4378 on hepatic lipid metabolism require further investigation.

To explore whether energy intake contributed to the observed weight regain, we monitored food consumption during the 14-day treatment and period and 16 days after cessation (Fig. 4d). During administration, TZP significantly suppressed food intake. However, upon withdrawal, the mice exhibited a sharp rebound in food consumption. By day 6 post-treatment, food intake in the TZP group exceeded that in both the WR controls and the LJ-4378 group. This post-treatment hyperphagia closely paralleled the body weight regain, suggesting that rebound feeding contributed substantially to the transient efficacy of TZP. In contrast, LJ-4378 did not alter food intake during or after treatment, which was consistent with its sustained effect on weight reduction.

The assessment of adiposity after treatment cessation revealed a more favorable profile in the LJ-4378 group than in the TZP group. The mice treated with LJ-4378 exhibited a significantly lower whole-body fat ratio than in those treated with TZP (Fig. 4e). Volumetric analysis of adipose depots revealed a reduction in gWAT and visceral abdominal fat in the LJ-4378-treated group, compared with the TZP group (Fig. 4f, g). These findings suggest that LJ-4378 not only limits post-treatment weight regain but also supports long-term reductions in adiposity.

Weight regain following dietary intervention is associated with macrophage infiltration and inflammatory responses in the visceral adipose tissue (Sougiannis et al. 2020). To investigate adipose tissue inflammation in the WR model, we assessed macrophage infiltration in the gWAT. In the TZP-treated group, the expression of the macrophage marker F4/80 (encoded by Adgre1) and the formation of CLS, which had been reduced by treatment, were re-elevated following treatment cessation (Fig. 5c–e, Fig. S7). In contrast, the LJ-4378-treated group maintained lower levels of macrophage accumulation after 4 weeks of the weight regaining phase (Fig. 5c–e, Fig. S7). Furthermore, reduced Tnf-α and increased Arg1 mRNA levels were maintained after 4 weeks of WR (Fig. S7).

Additionally, weight cycling in mice with a history of chronic diet-induced obesity leads to increased hepatic lipid storage despite weight loss compared to those maintained on a consistent diet (Bernecker et al. 2025). In our WR model, histological and biochemical analyses revealed significantly lower hepatic triglyceride (TG) deposition in the LJ-4378-treated group than in the TZP (Fig. 5f, g). Notably, the TZP-treated group, which experienced greater weight regain, also showed a resurgence of obesity-associated metabolic impairments compared with the LJ-4378 administration. These findings underscore the superior long-term metabolic protection conferred by LJ-4378 after treatment cessation.

The transient hyperphagia observed immediately after TZP withdrawal (Fig. 4d) appeared to be a short-term effect; however, it likely contributed to accelerated weight regain.

To evaluate the metabolic function of adipose tissue at the molecular level, we examined the expression of P-HSL and mitochondrial proteins in BAT and iWAT 2 weeks after treatment cessation (Fig. 6c, d). Mice previously treated with LJ-4378 expressed higher levels of P-HSL and mitochondrial proteins than those in both WR controls and the TZP-treated group, indicating sustained activation of lipolysis and mitochondrial function in BAT and iWAT. These findings suggest that LJ-4378 promotes prolonged adipose tissue metabolic activity compared with TZP, contributing to its capacity to maintain energy homeostasis and prevent post-treatment metabolic decline.

Extrapolation of thesevalues to mice and humans, using species-specific plasma unbound fractions and assuming the free drug hypothesis with consistent tissue unbound fractions, suggests that this distribution trend is likely to be maintained across species. Collectively, these characteristics indicate that LJ-4378 achieves sufficient adipose tissue exposure to mediate its pharmacological effects, while displaying favorable drug-like properties and translational potential for further development. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{P,T}$$\end{document} K P T ,