Obesity and its cardiometabolic sequelae are among the foremost threats to human health. GLP-1 therapies are a genuine advance, producing previously unattainable weight loss, yet the benefit is not durable: most patients regain weight once the drug is withdrawn, as they do after bariatric surgery. This relapse is the hallmark of a biologically defended body-weight set point, raising the question: where, and in what form, does the body store a record of prior metabolic state? 

We propose that this record is physical, a metabolic engram encoded in the peripheral circuitry linking brain to adipose tissue and in the chromatin of the connected cells. The argument is one of timescales: a set point stable over years cannot be held by hormones or metabolites, which turn over within hours, but must reside in the system's slow variables. In the brain, long-term memory is stored in synaptic connectivity and chromatin; we hypothesize the periphery uses analogous substrates, making obesity, in part, a disorder of biological memory.

This hypothesis remains untested because no method resolves connectivity, transcription, and genome folding in one cell. Two platforms built independently in our laboratories make it possible. Connectome Scope (Sun) converts each neuron–adipocyte connection into a sequencing readout by barcoded trans-synaptic tracing; LiMCA with Easy Dip-C (Tan) resolves the full-length transcriptome and 3D genome of that same cell. Together they answer whether genome folding stabilizes a metabolic circuit. 

Aim 1 maps the autonomic and sensory innervation of defined adipocyte populations across baseline, cold, obese, and weight-reduced states. The weight-reduced state is decisive: any connection or enhancer–promoter contact that outlasts weight loss is, by definition, the memory driving regain. Aim 2 silences these connections, beginning with neurons innervating Aldh1a1+ thermogenic-brake adipocytes, to test whether the defended set point relaxes. A positive result defines a new therapeutic target, depot- and cell-type-specific and chromatinanchored, distinct from current GLP-1 and β-adrenergic agents.