The circadian transcription factor ARNTL2 is regulated by weight-loss interventions in human white adipose tissue and inhibits adipogenesis

Obesity is linked to disrupted circadian rhythms and expansion of white adipose tissue (WAT) via increased adipocyte size and number. Adipogenesis from adipose stem/progenitor cells (ASCs) is driven by insulin/IGF-1 via PI3K/Akt/mTOR and RAS/MAPK cascades and transcriptional regulators such as C/EBPβ and PPARγ. Circadian clocks, governed by bHLH-PAS transcription factors, integrate light and feeding cues and modulate adipogenic pathways. ARNTL1 (BMAL1) positively regulates adipogenesis, while the role of its paralog ARNTL2 (BMAL2) in human ASCs remains unclear, with evidence suggesting both compensatory and distinct functions. Weight-loss (WL) interventions like caloric restriction modulate peripheral clocks and ASC function. The study aims to define ARNTL2’s regulation by WL and its functional role in human ASC adipogenesis, and to delineate its interaction with ARNTL1 and key signaling pathways.

Prior work shows: (1) Circadian disruption contributes to obesity and aging; peripheral clocks respond to feeding and regulate metabolic tissues, including WAT where ~25% of transcripts are diurnal. (2) ARNTL1/CLOCK complexes drive clock-controlled genes; PER/CRY provide negative feedback to generate rhythms. (3) ARNTL1 regulates adipogenesis; ARNTL2 can compensate for ARNTL1 in knockout mice but may have distinct activities. (4) WL/caloric restriction influences circadian mechanisms and improves stem cell function; WL targets in ASCs include inhibitors of PI3K/Akt/mTOR (DIRAS3) and MAPK (SPRY1), preserving ASC fitness. (5) The circadian-adipogenic crosstalk involves PER3→KLF15 pathway and early adipogenic factors like C/EBPβ. However, ARNTL2’s specific regulation and function in human adipogenesis had been incompletely characterized.

Human subcutaneous WAT (sWAT) was obtained from patients undergoing routine elective plastic abdominal surgery with informed consent and ethics approval. Donors included normal-weight (NWD), obese (OD), and weight-loss (WLD) groups; a total of n=32 donors (Caucasian) were used. ASCs were isolated and cultured per established protocols. Adipogenic differentiation was induced using differentiation medium (DM) as described previously; short-term stimulations employed DM or individual components (insulin, dexamethasone, FCS, transferrin, IBMX) to assess signaling and ARNTL responses. Circadian synchronization: Confluent ASCs were serum-starved for 2 days, treated with 30% FCS for 2 h (t=0), then maintained serum-free. Cells were harvested at multiple time points up to 72 h for RNA and protein analyses. Cosinor regression (web tool) was used to estimate circadian parameters for core clock genes. Signaling perturbations: Pharmacological inhibitors included Akt Inhibitor VIII (10 μM), U0126 (MEK inhibitor; 50 μM), and Rapamycin (mTORC1 inhibitor; 0.5 μM). Serum-starved ASCs were pretreated 30 min before DM stimulation, then harvested at 30 min and 4 h to assess effects on ARNTL2 stability and pathway activity. ARNTL2 overexpression: ARNTL2 cDNA (pENTR223 clone HsCD00514112) was Gateway-cloned into pLenti6/V5-DEST to generate pLenti6-ARNTL2-V5; empty vector served as control. Lentiviral particles were produced and ASCs infected; blasticidin selection was applied as appropriate. Overexpression was confirmed via V5 tag and immunoblotting. Protein analyses: Whole sWAT lysates (0.2–1 g tissue) were prepared in NP-40 buffer for Western blotting. For cell lysates, standard SDS-PAGE/Western blotting was performed; densitometry used ImageJ/Image Lab. Antibody specificity for ARNTL2 was validated by 2D gel electrophoresis. Protein stability was assessed by cycloheximide chase; proteasome involvement was tested with MG132 and LLnL. Gene expression: RNA was isolated; cDNA synthesized; RT-qPCR performed with SYBR Green on a QuantStudio 7. Relative expression was calculated using ΔΔCT normalized to ACTB. Microarray data (Affymetrix U133 Plus 2.0) from freshly isolated ASCs of NWD (n=3), OD (n=3), and WLD (n=4) were from prior work to identify WL-regulated genes. Adipogenesis readouts: Protein and mRNA levels of adipogenic markers (C/EBPβ, PPARγ2, FABP4, ADIPOQ) and clock-linked factors (ARNTL1, ARNTL2, KLF15, PER3) were measured across differentiation. Lipid accumulation was quantified by Oil Red O staining and spectrophotometric measurement. Statistics: Experiments used at least n=3 biological replicates (donors) with technical triplicates. Analyses employed paired/unpaired two-tailed t tests and ANOVA with appropriate multiple-comparison corrections; significance at p≤0.05.

- WL downregulates ARNTL2 mRNA in freshly isolated human ASCs: Microarray profiling of ASCs from NWD (n=3), OD (n=3), WLD (n=4) showed strong ARNTL2 mRNA reduction in WLD vs NWD/OD (Fig. 1). - Distinct circadian regulation of ARNTL paralogs in ASCs: After serum-shock, ARNTL2 protein displayed clear oscillations despite declining mRNA; ARNTL1 mRNA exhibited rhythmicity with peak at 4 h, while ARNTL1 protein peaked at ~6 h then declined. Cosinor regression fit was better for ARNTL1 mRNA (p≈0.054) than ARNTL2 mRNA (p≈0.144). - C/EBPβ mRNA oscillates after serum-shock but protein does not; KLF4 did not show rhythmicity in ASCs. - Adipogenic cues rapidly increase ARNTL2 protein: Short-term DM stimulation caused ARNTL2 protein to rise within 4 h and remain elevated, while ARNTL2 mRNA declined. ARNTL1 showed minimal protein change and a delayed mRNA increase. - FCS (2.5% v/v) alone robustly induced ARNTL2 protein and activated PI3K/Akt, mTOR (S6K), and MAPK pathways; ARNTL1 was not similarly induced (Fig. 4). - ARNTL2 stability requires both mTOR and MAPK activities: Short-term inhibition of PI3K/Akt/mTOR or MAPK alone did not prevent ARNTL2 induction at 30 min, but combined MEK (U0126) plus mTOR (Rapamycin) inhibition led to ARNTL2 protein depletion by 4 h, indicating cooperative pathway control of ARNTL2 stability. - Feedback: ARNTL2 overexpression (OE) dampened MAPK (reduced P-ERK/ERK) and mTOR signaling (reduced P-S6K/S6K), indicating ARNTL2 negatively feeds back on these pathways. - ARNTL2 OE inhibits adipogenesis: ARNTL2 OE reduced mRNA and protein of C/EBPβ, PPARγ2, FABP4, and Adiponectin; Oil Red O staining showed significantly decreased lipid accumulation, demonstrating impaired differentiation. - Mechanisms of inhibition: ARNTL2 OE decreased ARNTL1 protein (with unaffected ARNTL1 mRNA early), shortened ARNTL1 protein half-life, and proteasome inhibitor MG132 rescued ARNTL1 levels, indicating ubiquitin-proteasome-mediated degradation. ARNTL2 OE reduced KLF15 mRNA; PER3 was not detectable. - Temporal signaling dynamics: Despite early suppression, AktThr308 and later AktSer473/mTOR signaling reactivated during later differentiation, suggesting transient nature of ARNTL2’s suppressive effect. - In vivo tissue relevance: In whole sWAT lysates, ARNTL2 protein was significantly higher in WLDs than in NWDs and ODs; ARNTL1 protein was elevated in both ODs and WLDs vs NWDs; donor age did not correlate with ARNTL2 levels.

The study demonstrates that ARNTL2 is differentially regulated compared to ARNTL1 in human ASCs and acts as an inducible inhibitor of adipogenesis. Circadian synchronization and adipogenic stimuli elicit distinct ARNTL2 protein dynamics driven by post-transcriptional mechanisms and sustained by mTOR and MAPK signaling. ARNTL2, in turn, suppresses these pathways, establishing a feedback loop that stabilizes its own protein levels during early adipogenesis. Functionally, elevated ARNTL2 impairs adipocyte differentiation by promoting proteasomal degradation of the pro-adipogenic factor ARNTL1, attenuating the MAPK–C/EBPβ axis, and reducing KLF15 expression, a clock-linked driver of adipogenesis. These findings address the research question by linking WL-associated regulation of ARNTL2 to control of ASC differentiation and WAT homeostasis. The elevation of ARNTL2 protein in sWAT from WLDs underscores its physiological relevance and suggests ARNTL2 as a mediator of WL-induced adaptations that may protect ASCs from aberrant signaling and senescence while limiting adipose hyperplasia.

ARNTL2 is a weight-loss–regulated circadian transcription factor that opposes adipogenesis in human ASCs. It is rapidly induced at the protein level by adipogenic cues, stabilized via mTOR and MAPK pathways, and feeds back to inhibit these pathways. ARNTL2 impairs adipocyte differentiation by promoting ARNTL1 degradation, suppressing MAPK–C/EBPβ signaling, and reducing KLF15 expression. In human sWAT, ARNTL2 protein is elevated specifically after WL, highlighting distinct roles of ARNTL paralogs in adipose biology. These insights identify ARNTL2 as a potential therapeutic target to modulate adipogenesis and improve metabolic outcomes in obesity. Future work should elucidate the precise molecular mechanisms by which ARNTL2 regulates mTOR/MAPK crosstalk and ARNTL1 turnover, and assess in vivo functional consequences of modulating ARNTL2 in adipose tissue.

- The mechanistic basis of how ARNTL2 inhibits mTOR/MAPK signaling and promotes ARNTL1 proteasomal degradation remains to be fully elucidated and is noted as a subject for future studies. - Circadian analyses showed weaker Cosinor fit for ARNTL2 mRNA, suggesting complexities in its transcriptional rhythmicity versus protein oscillation. - Most functional data derive from in vitro human ASC models; in vivo causality and systemic metabolic effects were not directly tested. - Sample sizes per experiment were modest (typically n=3 donors for many assays), which may limit statistical power for some comparisons.