Fig 1 T → E2 → ERα in a neuron hover or focus a part · space: pause aromatase active · E2 low
blood vessel cell membrane endoplasmic reticulum aromatase nucleus ERα dimer on ERE axon → circuit downstream neuron OH O OH OH C19 methyl testosterone (T) 17β-estradiol (E2) aromatase (CYP19A1) loses C19, A ring aromatised Schematic of the mechanism proposed for aromatase-expressing neurons in rodent aggression circuits. Not to scale; the human picture is far weaker (see sources).
Block aromatase Testosterone supply low medium high Pause motion
Pulse size on the axon is a schematic of E2–ERα effect, not a measured firing rate.
Hover or focus a labelled part for a one-line explanation.
What the figure shows, and how sure we are
Chemistry and enzymology (well established). Aromatase (CYP19A1) converts testosterone to estradiol by three oxidations that cleave the C10–C19 bond and aromatise ring A; it sits in the endoplasmic reticulum membrane and the brain is one tissue where it acts (Structural and functional characterization of aromatase, estrogen receptor, and their genes ; The essential role of the aromatase/p450arom ).Receptor action (well established). Estradiol-occupied ERα forms a homodimer that binds estrogen response elements and regulates transcription; ERα can also act by tethering to other transcription factors (Estrogen receptor interaction with estrogen response elements ; Estrogen receptor alpha: molecular mechanisms and emerging insights ).Male mice (good evidence, species-specific). Aromatase-expressing neurons in the posterodorsal medial amygdala regulate aggression (Unger et al. 2015 ); ERα knock-down in the medial amygdala before puberty reduces adult male social behaviour including aggression (PNAS 2016 ); ERα and AR both contribute (Scordalakes & Rissman 2004 ). Individual differences in ERα in select nuclei track individual differences in aggression (Trainor, Greiwe & Nelson 2006 ). Direction and receptor subtype matter: ERα and ERβ can differ, and estradiol effects depend on day length (Trainor, Rowland & Nelson 2007 ; Trainor, Finy & Nelson 2008 ). A systematic review of animal work finds estrogen signalling usually contributes to aggression, varying by species, sex and paradigm (Hapres JPBS review ).Men (weak and mixed). A 2023 meta-analysis of 14 studies (1,820 participants) found a weak positive correlation between estradiol and human aggression, pooled Fisher z = 0.16 (95% CI 0.05–0.26; I² = 73%), with sex, age and the aggression measure contributing to heterogeneity (Wang et al., Psychosomatic Medicine 2023 ; PubMed ). Fisher z is a transformed correlation, not a percentage of variance, and it does not show causation.
What was simplified. One neuron stands for a circuit; membrane and other non-genomic ER signalling, co-regulators, other brain regions, and local versus circulating hormone sources are omitted. Positions, sizes and particle speeds are illustrative. The circuit and receptor details come from rodents; they have not been shown to explain aggression in people. “Aromatase-blocked” is a teaching toggle, not a statement about any drug or dose.
For education only; this figure is not medical advice and says nothing about any individual’s hormones or behaviour.