📗 Cite This Artifact
Lipid Droplet Dynamics as Phenotype Switches
🧪 Overview
Mechanistic Overview
Lipid Droplet Dynamics as Phenotype Switches starts from the claim that modulating DGAT1 and SOAT1 within the disease context of neurodegeneration can redirect a disease-relevant process. The original description reads: "Molecular Mechanism and Rationale The hypothesis centers on the differential regulation of lipid droplet composition between A1 and A2 astrocyte phenotypes through the enzymatic balance of diacylglycerol O-acyltransferase 1 (DGAT1) and sterol O-acyltransferase 1 (SOAT1). DGAT1 catalyzes the final step in triglyceride synthesis by transferring acyl-CoA to diacylglycerol, while SOAT1 (also known as ACAT1) esterifies cholesterol to form cholesteryl esters. In A2 astrocytes, elevated SOAT1 activity relative to DGAT1 promotes the formation of cholesteryl ester-enriched lipid droplets that sequester inflammatory lipid mediators and serve as reservoirs for membrane repair components. These cholesteryl ester-rich droplets interact with perilipin-2 (PLIN2) and comparative gene identification-58 (CGI-58) to maintain a stable, anti-inflammatory lipid environment....
Mechanistic Overview
Lipid Droplet Dynamics as Phenotype Switches starts from the claim that modulating DGAT1 and SOAT1 within the disease context of neurodegeneration can redirect a disease-relevant process. The original description reads: "Molecular Mechanism and Rationale The hypothesis centers on the differential regulation of lipid droplet composition between A1 and A2 astrocyte phenotypes through the enzymatic balance of diacylglycerol O-acyltransferase 1 (DGAT1) and sterol O-acyltransferase 1 (SOAT1). DGAT1 catalyzes the final step in triglyceride synthesis by transferring acyl-CoA to diacylglycerol, while SOAT1 (also known as ACAT1) esterifies cholesterol to form cholesteryl esters. In A2 astrocytes, elevated SOAT1 activity relative to DGAT1 promotes the formation of cholesteryl ester-enriched lipid droplets that sequester inflammatory lipid mediators and serve as reservoirs for membrane repair components. These cholesteryl ester-rich droplets interact with perilipin-2 (PLIN2) and comparative gene identification-58 (CGI-58) to maintain a stable, anti-inflammatory lipid environment. Conversely, A1 astrocytes exhibit increased DGAT1/SOAT1 ratios, leading to triglyceride-predominant lipid droplets that associate with different peridroplet proteins, including perilipin-3 (PLIN3) and adipose triglyceride lipase (ATGL). This composition facilitates rapid lipolysis and the release of arachidonic acid and other pro-inflammatory fatty acids. The liberated arachidonic acid is subsequently metabolized by cyclooxygenase-2 (COX-2) and 5-lipoxygenase (5-LOX) to generate prostaglandins and leukotrienes, respectively. Additionally, triglyceride-rich droplets in A1 astrocytes show enhanced interaction with the endoplasmic reticulum stress sensor IRE1α, promoting the unfolded protein response and inflammatory cytokine production through NF-κB activation. The molecular switch between these phenotypes involves the transcriptional regulation of DGAT1 and SOAT1 by sterol regulatory element-binding protein 1c (SREBP-1c) and liver X receptors (LXRα/β). In the presence of neuroinflammatory signals like TNF-α or amyloid-β, SREBP-1c activation preferentially upregulates DGAT1 expression, while LXR-mediated SOAT1 transcription is suppressed through competing cofactor availability. This regulatory mechanism is further modulated by the metabolic sensor AMP-activated protein kinase (AMPK), which phosphorylates and inactivates DGAT1 while promoting SOAT1 activity through enhanced substrate availability. Preclinical Evidence Extensive preclinical validation supports this hypothesis across multiple model systems. In 5xFAD mice, immunofluorescence microscopy reveals distinct lipid droplet compositions in cortical astrocytes, with A1-like astrocytes showing 3.2-fold higher triglyceride content compared to A2-like astrocytes, while A2 astrocytes contain 4.7-fold more cholesteryl esters. Lipidomic analysis using mass spectrometry demonstrates that DGAT1 knockdown in primary mouse astrocytes reduces triglyceride accumulation by 65% and decreases IL-1β secretion by 45% following LPS stimulation. Conversely, SOAT1 overexpression increases cholesteryl ester content by 80% and reduces TNF-α production by 38%. In vitro studies using human iPSC-derived astrocytes confirm these findings, showing that pharmacological inhibition of DGAT1 with A-922500 (10 μM) shifts astrocytes toward an A2-like phenotype, evidenced by increased GFAP and S100β expression and reduced complement C3 levels. Time-lapse imaging reveals that A1 astrocytes exhibit more dynamic lipid droplets with 2.8-fold faster fusion/fission rates compared to A2 astrocytes, consistent with active lipolysis in triglyceride-rich droplets. C. elegans models expressing human amyloid-β show that dgat-1 knockdown extends lifespan by 18% and reduces neuronal death by 32%, while soat-1 overexpression provides similar neuroprotective effects. Quantitative RT-PCR analysis demonstrates that the DGAT1/SOAT1 mRNA ratio correlates inversely with astrocyte anti-inflammatory gene expression (r = -0.73, p < 0.001) across multiple mouse models of neurodegeneration, including APP/PS1, SOD1G93A, and rotenone-induced Parkinson's models. Electron microscopy studies reveal ultrastructural differences in lipid droplets between A1 and A2 astrocytes, with A1 droplets showing irregular morphology and frequent ER contact sites, while A2 droplets appear more spherical and associate with mitochondria. Functional assays demonstrate that conditioned media from A2 astrocytes with high cholesteryl ester content promotes oligodendrocyte survival by 42% and enhances neurite outgrowth by 28% compared to media from A1 astrocytes. Therapeutic Strategy and Delivery The therapeutic approach involves selective modulation of DGAT1 and SOAT1 activity using a combination of small molecule inhibitors and activators. The lead compound, designated LD-Switch-001, is a dual-action molecule that simultaneously inhibits DGAT1 (IC50 = 15 nM) while activating SOAT1 (EC50 = 23 nM) through allosteric mechanisms. This compound demonstrates excellent CNS penetration with a brain-to-plasma ratio of 0.8 and minimal off-target effects on hepatic lipid metabolism at therapeutic doses. Alternative approaches include antisense oligonucleotides (ASOs) targeting DGAT1 mRNA, designed with 2'-O-methoxyethyl modifications for enhanced stability and CNS delivery. The ASO formulation uses conjugation to the transferrin receptor-binding peptide THR-002 for improved blood-brain barrier penetration and astrocyte-specific uptake. Pharmacokinetic studies show that intrathecal administration of the ASO achieves 70% DGAT1 knockdown in cortical astrocytes within 48 hours, with effects lasting 3-4 weeks. For chronic administration, an adeno-associated virus (AAV-PHP.eB) vector expressing a SOAT1-specific transcriptional activator under the GFAP promoter provides astrocyte-targeted gene therapy. The vector includes a tet-ON system for temporal control of SOAT1 expression, allowing dose titration through oral doxycycline administration. Biodistribution studies demonstrate preferential transduction of astrocytes over neurons (15:1 ratio) with minimal peripheral organ exposure. Delivery considerations include the use of focused ultrasound with microbubbles to enhance drug penetration across the blood-brain barrier, particularly for larger molecules like therapeutic antibodies targeting DGAT1. The optimal dosing regimen involves weekly intravenous infusions of 5 mg/kg for the dual-action small molecule, with therapeutic drug monitoring based on CSF lipid profiles and astrocyte activation markers. Evidence for Disease Modification Disease-modifying effects are evidenced through multiple biomarker and functional outcome measures that distinguish symptomatic improvement from underlying pathology modification. CSF analysis reveals that successful DGAT1/SOAT1 modulation increases cholesteryl ester levels by 45% while reducing inflammatory lipid mediators including leukotriene B4 (60% reduction) and prostaglandin E2 (38% reduction). These changes correlate with decreased CSF levels of complement C3 (42% reduction) and increased anti-inflammatory cytokines IL-4 and IL-10 (2.1-fold and 1.8-fold increases, respectively). Advanced imaging biomarkers include positron emission tomography (PET) using [18F]-FEPPA to measure translocator protein (TSPO) expression as an indicator of microglial activation. Treated animals show 35% reduction in TSPO binding potential in cortical regions, indicating decreased neuroinflammation. Magnetic resonance spectroscopy demonstrates increased N-acetylaspartate (NAA) levels by 22%, suggesting improved neuronal viability and function. Functional outcomes include improved performance in Morris water maze testing (25% reduction in escape latency) and novel object recognition tasks (40% improvement in discrimination index). These cognitive improvements persist for at least 8 weeks after treatment cessation, indicating lasting disease modification rather than temporary symptomatic relief. Electrophysiological recordings show enhanced long-term potentiation (LTP) in hippocampal slices from treated animals, with 1.6-fold greater synaptic strength compared to controls. Neuropathological analysis reveals 48% reduction in amyloid plaque burden and 35% decrease in phospho-tau accumulation in relevant disease models. Importantly, these changes occur independently of cognitive improvements, suggesting that lipid droplet modulation affects multiple pathological processes. Astrocyte morphology analysis shows increased process complexity and territory coverage in treated animals, consistent with enhanced neuroprotective function. Clinical Translation Considerations Patient selection criteria focus on individuals with early-stage neurodegeneration and evidence of astrocyte activation, identified through CSF biomarkers including elevated GFAP (>300 pg/mL) and reduced cholesteryl ester/triglyceride ratios (<0.8). Genetic stratification considers APOE genotype, as APOE4 carriers show enhanced response to cholesteryl ester modulation due to baseline deficits in lipid homeostasis. Age-related considerations include dose adjustments for patients over 75 years due to altered drug metabolism and increased sensitivity to lipid modifications. The regulatory pathway follows FDA guidelines for neurodegenerative disease therapeutics, with Phase I safety trials focusing on maximum tolerated dose and pharmacokinetic characterization. Phase II efficacy trials employ adaptive designs with interim analyses at 6 and 12 months, using composite endpoints combining cognitive assessments, biomarker changes, and imaging outcomes. The primary endpoint includes a 30% reduction in cognitive decline rate measured by CDR-SB progression over 18 months. Safety considerations include monitoring for hepatic lipid accumulation through regular liver function tests and imaging, as systemic DGAT1 inhibition may affect hepatic triglyceride metabolism. Cardiovascular monitoring addresses potential effects on plasma lipid profiles, though preclinical studies suggest minimal systemic exposure at therapeutic CNS doses. Drug-drug interactions require careful evaluation, particularly with statins and other lipid-modifying agents that may interfere with the therapeutic mechanism. The competitive landscape includes other neuroinflammation-targeting therapies, but the specific focus on astrocyte lipid metabolism provides a differentiated mechanism of action. Intellectual property protection covers the dual DGAT1/SOAT1 modulation approach and specific delivery methods for CNS targeting. Future Directions and Combination Approaches Future research directions include investigation of additional enzymes in lipid droplet metabolism, particularly hormone-sensitive lipase (HSL) and monoacylglycerol lipase (MAGL), which may provide additional therapeutic targets for fine-tuning astrocyte phenotypes. Single-cell RNA sequencing studies aim to identify astrocyte subpopulations with distinct lipid metabolic profiles and their specific contributions to neurodegeneration progression. Combination therapeutic approaches show particular promise when pairing lipid droplet modulation with complementary mechanisms. Co-administration with PPARγ agonists enhances the anti-inflammatory effects of cholesteryl ester accumulation, while combination with autophagy modulators may improve clearance of damaged lipid droplets and associated inflammatory mediators. Synergistic effects are observed with omega-3 fatty acid supplementation, which provides anti-inflammatory substrates for incorporation into cholesteryl ester stores. Broader applications extend to other neurodegenerative conditions including Parkinson's disease, where α-synuclein aggregation is influenced by lipid droplet dynamics, and Huntington's disease, where mutant huntingtin affects lipid metabolism. Multiple sclerosis represents another target indication, as oligodendrocyte lipid requirements may be supported through astrocyte-derived cholesteryl ester mobilization. Technology development focuses on improved delivery systems, including engineered extracellular vesicles from astrocytes that naturally accumulate cholesteryl esters, providing a biomimetic approach to therapeutic delivery. Advanced imaging techniques using hyperpolarized MRI may enable real-time monitoring of lipid metabolism in living patients, facilitating personalized dosing and treatment monitoring. The integration of artificial intelligence and machine learning approaches enables prediction of optimal DGAT1/SOAT1 ratios for individual patients based on genetic, metabolic, and clinical parameters, moving toward precision medicine applications in neurodegeneration treatment.
Mechanistic Pathway Diagram
" Framed more explicitly, the hypothesis centers DGAT1 and SOAT1 within the broader disease setting of neurodegeneration. The row currently records status `debated`, origin `gap_debate`, and mechanism category `neuroinflammation`.
SciDEX scoring currently records confidence 0.35, novelty 0.80, feasibility 0.50, impact 0.55, mechanistic plausibility 0.40, and clinical relevance 0.44.
Molecular and Cellular Rationale
The nominated target genes are `DGAT1 and SOAT1` and the pathway label is `Astrocyte reactivity signaling`. Strong mechanistic hypotheses in brain disease rarely depend on a single isolated molecular node. Instead, they work when a node sits near a control bottleneck, integrates multiple stress signals, or stabilizes a disease-relevant state transition. That is the standard this hypothesis should be held to. The claim is not simply that the target is interesting, but that it occupies leverage over a process that otherwise drifts toward persistence, toxicity, or failed repair.
Gene-expression context on the row adds an important constraint: # Gene Expression Context
DGAT1
- Primary Function: Catalyzes the final committed step of triglyceride synthesis by esterifying diacylglycerol with long-chain fatty acyl-CoA; rate-limiting enzyme for neutral lipid accumulation in lipid droplets
- Brain Regional Expression:
- Highest expression in hypothalamus and midbrain regions (Allen Brain Atlas)
- Moderate expression in cortex, hippocampus, and cerebellum
- Lower basal expression in white matter tracts
- Cell Type Expression:
- Primary expression in astrocytes, particularly A1 phenotype astrocytes
- Present in oligodendrocytes and microglia at lower levels
- Neuronal expression minimal under homeostatic conditions
- Disease State Changes:
- DGAT1 expression increases 1.5-2.2 fold in A1 astrocytes during neuroinflammatory states
- Upregulation correlates with triglyceride-enriched lipid droplet accumulation in acute neurodegeneration models
- In Alzheimer's disease, DGAT1 elevation observed in activated astrocytes surrounding amyloid plaques
- Chronic neuroinflammation (LPS, TNF-α stimulation) increases DGAT1 expression via NF-κB pathway activation
- Relevance to Hypothesis: Elevated DGAT1 in A1 astrocytes promotes triglyceride-dominant lipid droplets that accumulate pro-inflammatory lipid mediators (arachidonic acid-derived eicosanoids), supporting the proinflammatory phenotype and metabolic inflexibility characteristic of neurotoxic astrocytes
- Quantitative Details:
- Baseline expression ~2-3 fold lower in primary astrocytes compared to hepatocytes
- TNF-α stimulation increases DGAT1 mRNA ~2.8 fold within 4-6 hours
- Lipid droplet triglyceride content increases proportionally with DGAT1 activity
SOAT1
- Primary Function: Esterifies cholesterol to cholesteryl esters via acyl-CoA transferase activity; regulates free cholesterol homeostasis and cholesteryl ester accumulation in lipid droplets
- Brain Regional Expression:
- High expression in cortex, hippocampus, and striatum (Allen Brain Atlas)
- Enriched in gray matter regions with high synaptic density
- Moderate expression in cerebellum and midbrain
- White matter expression relatively lower
- Cell Type Expression:
- Predominantly astrocytic expression, with enrichment in A2 phenotype astrocytes
- Significant expression in microglia and perivascular pericytes
- Oligodendrocytes express SOAT1 at moderate levels for myelin maintenance
- Neuronal soma and dendrites express SOAT1 at lower baseline levels
- Disease State Changes:
- SOAT1 expression increases 1.3-1.8 fold in resting/A2 astrocytes during recovery phase of neuroinflammation
- Downregulated 0.4-0.6 fold in acute A1 astrocytes (48 hours post-LPS/TNF-α)
- In Alzheimer's disease models, SOAT1 increases in perilesional astrocytes suggesting neuroprotective response
- Chronic neurodegeneration shows biphasic SOAT1 expression: early suppression followed by late upregulation in remote astrocytes
- Cholesteryl ester accumulation increases 2-3 fold in A2 astrocytes relative to A1 phenotype
- Relevance to Hypothesis: Elevated SOAT1 in A2 astrocytes generates cholesteryl ester-enriched lipid droplets that sequester oxidized cholesterol and inflammatory lipid mediators, supporting neuroprotective functions and facilitating myelin repair through cholesterol redistribution; SOAT1/DGAT1 ratio serves as critical switch determining astrocyte phenotype and metabolic state
- Quantitative Details:
- SOAT1/DGAT1 expression ratio ~0.3-0.5 in A1 astrocytes vs ~1.5-2.1 in A2 astrocytes
- IL-4 stimulation (A2 induction) increases SOAT1 mRNA ~2.5 fold over 12-24 hours
- Cholesteryl ester content correlates with SOAT1 expression (R² ~0.78 in primary astrocytes)
- Perilipin-2 (PLIN2) co-expression with SOAT1 increases ~1.9 fold in A2 phenotype, stabilizing lipid droplet surface
Evidence Supporting the Hypothesis
Contradictory Evidence, Caveats, and Failure Modes
Clinical and Translational Relevance
From a translational perspective, this hypothesis only matters if it can be turned into a selection rule for experiments, biomarkers, or patient stratification. The row currently records market price `0.6967`, debate count `2`, citations `21`, predictions `1`, and falsifiability flag `1`. Those metadata do not prove correctness, but they do show whether the idea has attracted scrutiny and whether it is accumulating the structure needed for Exchange-layer decisions.
Experimental Predictions and Validation Strategy
First, the hypothesis should be decomposed into a perturbation experiment that directly manipulates DGAT1 and SOAT1 in a model matched to neurodegeneration. The key readout should include pathway markers, cell-state markers, and at least one phenotype that maps onto "Lipid Droplet Dynamics as Phenotype Switches".
Second, the study design should include a rescue arm. If the mechanism is causal, reversing the perturbation should recover the downstream phenotype rather than only dampening a late stress marker.
Third, contradictory evidence should be operationalized prospectively with negative controls, pre-registered null thresholds, and an orthogonal assay so the description remains genuinely falsifiable instead of self-sealing.
Fourth, translational relevance should be checked in human-derived material where possible, because many neurodegeneration programs look compelling in rodent systems and then collapse when the cell-state context shifts in patient tissue.
Decision-Oriented Summary
In summary, the operational claim is that targeting DGAT1 and SOAT1 within the disease frame of neurodegeneration can produce a measurable change in mechanism rather than only a cosmetic change in a terminal biomarker. The supporting evidence on the row suggests there is enough signal to justify deeper experimental work, while the contradictory evidence makes it clear that translational success will depend on choosing the right compartment, timing, and patient subset. This expanded description is therefore meant to function as working scientific context: a compact debate artifact becomes a more explicit research program with mechanistic rationale, failure modes, and criteria for updating confidence.
🧬 Mechanism
Curated pathway from expert analysis
graph TD
A["Neuroinflammatory Stimuli"]
B["DGAT1 Expression"]
C["SOAT1 Expression"]
D["Triglyceride Synthesis"]
E["Cholesteryl Ester Formation"]
F["PLIN2/CGI-58 Interaction"]
G["A1 Astrocyte Phenotype"]
H["A2 Astrocyte Phenotype"]
I["Pro-inflammatory Lipid Release"]
J["Anti-inflammatory Lipid Storage"]
K["Membrane Repair Deficits"]
L["Neuroprotective Function"]
M["Neurodegeneration"]
N["Neuronal Survival"]
O["DGAT1 Modulators"]
P["SOAT1 Inhibitors"]
A -->|"induces"| B
A -->|"suppresses"| C
B -->|"catalyzes"| D
C -->|"catalyzes"| E
D -->|"promotes"| G
E -->|"facilitates"| F
F -->|"stabilizes"| H
G -->|"triggers"| I
H -->|"maintains"| J
I -->|"causes"| K
J -->|"supports"| L
K -->|"leads to"| M
L -->|"promotes"| N
O -->|"therapeutic target"| B
P -->|"therapeutic target"| C
classDef mechanism fill:#4fc3f7,color:#0d0d1a
classDef pathology fill:#ef5350,color:#0d0d1a
classDef therapy fill:#81c784,color:#0d0d1a
classDef outcome fill:#ffd54f,color:#0d0d1a
classDef genetics fill:#ce93d8,color:#0d0d1a
class A,B,C,D,E,F mechanism
class G,I,K,M pathology
class O,P therapy
class N outcome
class H,J,L genetics⚖️ Evidence
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract
Abstract


📙 Related Wiki Pages (15)
🏥 Translation
🧬 3D Protein Structure — DGAT1
No curated PDB or AlphaFold mapping for DGAT1 yet. Search RCSB →
🧠 GTEx v10 Brain ExpressionJSON
Median TPM across 13 brain regions for DGAT1 and SOAT1 from GTEx v10.
💉 Clinical Trials (5)Relevance: 44%
Active
Completed
Total Enrolled
Highest Phase
No curated ClinVar variants loaded for this hypothesis.
Run scripts/backfill_clinvar_variants.py to fetch P/LP/VUS variants.
No DepMap CRISPR Chronos data found for DGAT1 and SOAT1.
Run python3 scripts/backfill_hypothesis_depmap.py to populate.
🏆 Tournament
🏆 Arenas / Elo
📊 Market Indicators
💾 Resource Usage
🧭 Related
🔍 Show all 50 edges across 5 relations
associated with (5)
co discussed (26)
🗺️ KG Entities (36)
🔗 Dependency Graph (1 upstream, 0 downstream)
🔮 Predictions
| Prediction | Predicted | Observed | Status | Conf |
|---|---|---|---|---|
| Modulation of DGAT1 and SOAT1 will affect the proposed pathway | DGAT1 and SOAT1 knockdown/overexpression shows measurable effect | — no observation — | pending | 0.35 |
📖 References (11)
- Inhibition of sterol O-acyltransferase 1 blocks Zika virus infection in cell lines and cerebral organoids.Schöbel A et al.. Commun Biol (2024)
- Lipid Metabolism in Glioblastoma: From De Novo Synthesis to Storage.Kou Y et al.. Biomedicines (2022)
- Lipid Stores and Lipid Metabolism Associated Gene Expression in Porcine and Bovine Parthenogenetic Embryos Revealed by Fluorescent Staining and RNA-seq.Kajdasz A et al.. Int J Mol Sci (2020)
- Effect of Carotenoids from Phaeodactylum tricornutum on Palmitate-Treated HepG2 Cells.Mayer C et al.. Molecules (Basel, Switzerland) (2020)
- Placental extract suppresses lipid droplet accumulation by autophagy during the differentiation of adipose-derived mesenchymal stromal/stem cells into mature adipocytes.Ando Y et al.. BMC research notes (2023)
- Inducible deletion of DGAT1 and 2 from microglia exacerbates neurodegeneration and endolysosomal lipid accumulation in male PS19 mice.["Tabor G" et al.. Cell reports (2026)
- AMPK protects proximal tubular epithelial cells from lysosomal dysfunction and dedifferentiation induced by lipotoxicity.["Pierre L" et al.. Autophagy (2025)
- Depalmitoylation of TEAD1 facilitates lipid droplet accumulation and resistance to oxidative stress by transactivating PP2Acα.["Sun X" et al.. Free radical biology & medicine (2025)
- Low Dose GLP-1 Therapy Attenuates Pathological Cardiac and Hepatic Remodelling in HFpEF Independent of Weight Loss.["Elbatreek M" et al.. bioRxiv : the preprint server for biology (2025)
- DGAT1 inhibitors protect pancreatic β-cells from palmitic acid-induced apoptosis.["Huang J" et al.. Acta pharmacologica Sinica (2021)
- Diacylglycerol acyltransferase 1/2 inhibition induces dysregulation of fatty acid metabolism and leads to intestinal barrier failure and diarrhea in mice.["Takemoto K" et al.. Physiological reports (2020)
▸Metadata
| status | proposed |
| _schema_version | 1 |
| hypothesis_type | None |
derives from (14)
▸ Show 9 more
supports (18)
▸ Show 13 more
🧬 Related Hypotheses — same target / disease (20)
Use ?embed=1 to load the artifact without SciDEX chrome — suitable for iframing into wiki pages or external sites.
<iframe src="http://scidex.ai/artifact/hypothesis-h-7d4a24d3?embed=1" width="100%" height="600" style="border:0;border-radius:8px"></iframe>
[Lipid Droplet Dynamics as Phenotype Switches](http://scidex.ai/artifact/hypothesis-h-7d4a24d3)
http://scidex.ai/artifact/hypothesis-h-7d4a24d3