Transcriptomics Reveals Adipocyte Beiging as a Metabolic Disease Drug Target

Metabolic diseases remain among the most complex and heterogeneous conditions in modern medicine, presenting a global health challenge that requires innovative therapeutic strategies. 

Promoting adipocyte beiging offers an avenue to enhance energy expenditure in these tissues and potentially combat metabolic diseases. During beiging, white adipose tissue—which predominantly stores fat as energy—becomes more like brown adipose tissue, which expends the stored energy via lipid oxidation. 

Transcriptomic technologies are increasingly being used to interrogate adipocyte biology and identify therapeutic mechanisms with translational potential.  

Ahead of her talk at the Society for Laboratory Automation and Screening (SLAS) Europe Conference 2026, Technology Networks spoke with Dr. Maren Feist, a principal scientist at Evotec, to learn more about applying high‑throughput transcriptomic phenotyping to decode adipocyte function, uncover disease‑relevant pathways, and guide compound prioritization. 

Transcriptomics as a window into adipocyte biology 

“Metabolic diseases are highly complex and require multidimensional readouts to capture the full range of cellular changes,” Feist told Technology Networks. “While other omics approaches, such as proteomics or metabolomics, also capture these changes, high-throughput transcriptomic protocols tend to be more cost-efficient and scalable.” 

Widely available reference and compound-induced transcriptomic data, alongside well-characterized cellular pathways and transcription factor regulators, enable transcriptomics to deliver valuable mechanistic information. The rise of single‑nucleus and spatial RNA sequencing has further expanded the field, enabling researchers to identify disease‑relevant gene expression changes at single‑cell resolution.

  

“Together, these resources support higher translatability from in vitro to in vivo settings while enabling a level of mechanistic insight that is currently not covered to the same extent by other omics technologies,” Feist explained.

 

The high lipid content of adipocytes can strongly interfere with transcriptomic sample processing, she noted.

However, Evotec’s high-throughput transcriptomic protocol, ScreenSeq, which has been validated across more than 150 different cell models, can be adapted slightly to deliver high-quality data from these cells. 

Mechanistic advantages of transcriptomics: 

• Extensive reference datasets support pathway‑level interpretation 

• High scalability enables large compound screens 

• Single‑cell and spatial data improve in vitro to in vivo translatability 

Defining a biologically-grounded adipocyte beiging signature 

“Adipocyte beiging has consistently been shown to exert beneficial effects on obesity, insulin resistance, and metabolic syndrome,” explained Feist. “These effects are not driven by increased energy expenditure alone, but also by improvements in lipid and glucose handling and endocrine effects, which represent core cellular dysfunctions in metabolic disease.” 

Conventional assays have so far been unable to capture the complex cellular changes behind this transition. Feist and her team therefore sought to identify a gene signature that reflects the full spectrum of beiging‑associated changes. 

They began by establishing tightly controlled white-to-brown adipocyte differentiation in a model to generate a reliable reference. Multiple functional assays confirmed expected metabolic shifts, and the team refined the gene signature to genes belonging to common regulatory networks found in publicly available in vivo and in vitro datasets. 

“We arrived at a transcriptional signature that balances experimental robustness with biological relevance across model systems and patients,” said Feist.

How the beiging signature was refined: 

• Started with a controlled beiging differentiation model 

• Validated metabolic changes using functional assays 

• Integrated in vitro and in vivo datasets from rodent and human studies 

• Selected genes belonging to conserved regulatory networks 

Screening for beiging drivers in a translatable adipocyte model 

Using an immortalized human white adipocyte (iHWA) model, Feist and her team studied how compounds influence adipose cell phenotype. 

“The core of our model is the differentiation of immortalized pre‑adipocytes into functional white adipocytes,” said Feist. “From the outset, it was very important to us to demonstrate that these cells are not just generic lipid-storing cells, but closely resemble adipocytes found in vivo.”  

To address this, they compared the basal transcriptome of the differentiated cells against single‑nucleus RNA‑seq data from human adipose tissue, finding a strong alignment in adipocyte marker expression, which confirmed the model’s translatability. 

With the model established, the researchers screened a focused library of 320 target‑selective tool compounds. The transcriptomic profile induced by each compound was compared to the transcriptomic beiging signature using three complementary scoring metrics: 

• Concordance: How many signature genes were regulated in the expected direction 

• Strength: The magnitude of this regulation 

• Specificity: How genes outside of the beiging signature were regulated 

Several pathways emerged as potentially disease-relevant from the screen. While expected regulators such as the cAMP–AMPK axis appeared, the strongest and most selective hits included modulators of PDGFR and Notch signaling—pathways implicated in brown adipocyte development and white‑to‑brown transdifferentiation in mice. 

“Our transcriptomic approach identifies relevant beiging drivers while reliably filtering out non‑specific or nuisance effects,” said Feist. 

The specificity score also helped deprioritize compounds with broad or toxic signatures, ensuring that only mechanistically meaningful hits advanced. 

Insights from the iHWA compound screen: 

• Transcriptomic benchmarking confirmed the model had strong similarity to human adipocytes 

• Three‑metric scoring system improved hit prioritization 

• PDGFR and Notch modulators emerged as selective beiging drivers 

Unbiased discovery of disease‑relevant regulatory programs 

Rather than pre‑defining a narrow set of pathways, Feist emphasized the value of maintaining an unbiased discovery space. While early screens have already surfaced known regulators, the team aims to let the data reveal additional mechanisms that may be therapeutically relevant. 

“Our approach was designed to strike a balance between adding therapeutic relevance—by refining the signature using publicly available data—and maintaining an unbiased gene selection that allows a holistic view on beiging and the identification of novel regulatory programs and mechanisms,” said Feist. 

The team expects priority hits to affect lipid metabolism and mitochondrial pathways, but remains open to which regulatory mechanisms and targets emerge as the most meaningful. 

“This is one of the key advantages of transcriptomic screening,” explained Feist. “The range of detectable hits and mechanisms is far less constrained than with conventional single readout assays, which inherently limit the mechanistic space by focusing on only one predefined outcome.”  

Why an unbiased approach matters: 

• Allows discovery of novel regulatory mechanisms 

• Avoids constraints imposed by single‑readout assays 

• Preserves therapeutic relevance while enabling exploration 

• Supports identification of pathways beyond canonical beiging biology 

Integrating transcriptomic phenotypes with functional and translational data 

Functional assays—such as lipolysis, glucose uptake, and oxygen consumption—remain essential for validating the physiological impact of prioritized hits. However, Feist noted that on their own, these assays primarily serve as confirmation tools rather than discovery engines. 

“To go beyond this, we plan to apply advanced computational approaches that integrate the gene regulation patterns induced by our hits with proprietary patient‑derived transcriptomic data,” she said. “This allows us to further prioritize compounds based on human disease relevance and to identify potential targets that are more likely to translate into metabolic improvements in vivo.” 

In addition, the team will use CRISPR‑based gene knockdown followed by transcriptomic profiling to refine their models and validate mechanistic predictions. 

 “Transcriptomic phenotyping forms a natural bridge between functional assays, target identification, and translational validation.” — Dr. Maren Feist. 

Because RNA‑seq is applicable across experimental systems, in vitro gene expression patterns can be tracked for in vivo proof-of-concept studies. With systematic comparison to patients’ transcriptomic signatures, predictions can be cross-checked between model systems and human outcomes. 

How transcriptomics strengthens translational workflows: 

• Functional assays validate physiological impact 

• Patient‑derived datasets guide human‑relevant prioritization 

• Cross‑system transcriptomic comparison supports translational confidence
• Bridging functional assays, target identification, and translational validation is likely to be a key component of AI-driven drug discovery workflows

Expanding transcriptomic phenotyping across drug discovery 

“In my talk [at SLAS Europe 2026], I focus on adipocyte beiging as one example of how transcriptomic phenotyping can be applied in a high-throughput and translationally relevant way,” Feist said.  “Anyone interested in our screening strategy, the underlying concepts, or additional use cases is very welcome to connect with us during the poster session or at our booth.” 

She emphasized that this is just one of several ongoing applications of the platform, with additional use cases showcased in a poster session. 

 

Transcriptomic phenotyping is emerging as a powerful engine for metabolic disease drug discovery, offering scalability, mechanistic depth, and strong translational alignment. Feist’s work demonstrates how refined gene signatures, translatable adipocyte models, and multidimensional scoring approaches can uncover meaningful beiging drivers and disease‑relevant pathways. 

By integrating transcriptomic profiles with functional assays, CRISPR perturbation, and patient‑derived datasets, her team aims to identify compounds and targets with a higher likelihood of improving metabolic outcomes in vivo.  

 

Key takeaways 

• Transcriptomics offers mechanistic insight for metabolic disease research 
• A refined, cross‑validated beiging signature improves biological relevance and translatability for drug screening
• Unbiased transcriptomic screening enables the discovery of novel regulatory pathways 

This content includes text that has been created with the assistance of generative AI and has undergone editorial review before publishing. Technology Networks' AI policy can be found here.