After intracranial hemorrhage, heme is released from cell-free hemoglobin. This red blood cell component may drive secondary brain injury at the hematoma‒brain interface. This study aimed to generate a spatially resolved map of transcriptome-wide gene expression changes in the heme-exposed brain and to define the potential therapeutic activity of the heme-binding protein, hemopexin. We stereotactically injected saline, heme, or heme‒hemopexin into the striatum of C57BL/6J mice. After 24 h, we elucidated the two-dimensional spatial transcriptome by sequencing 21760 tissue-covered features, at a mean transcript coverage of 3849 genes per feature. In parallel, we studied the extravasation of systemically administered fluorescein isothiocyanate labeled (FITC)-dextran, magnetic resonance imaging features indicative of focal edema and perfusion, and neurological functions as translational correlates of heme toxicity. We defined a cerebral heme-response signature by performing bidimensional differential gene expression analysis, based on unsupervised clustering and manual segmentation of sequenced features. Heme exerted a consistent and dose-dependent proinflammatory activity in the brain, which occurred at minimal exposures, below the toxicity threshold for the induction of vascular leakage. We found dose-dependent regional divergence of proinflammatory heme signaling pathways, consistent with reactive astrocytosis and microglial activation. Co-injection of heme with hemopexin attenuated heme-induced gene expression changes and preserved the homeostatic microglia signature. Hemopexin also prevented heme-induced disruption of the blood‒brain barrier and radiological and functional signals of heme injury in the brain. In conclusion, we defined heme as a potent inflammatoxin that may drive secondary brain injury after intracerebral hemorrhage. Co-administration of hemopexin attenuated the heme-derived toxic effects on a molecular, cellular, and functional level, suggesting a translational therapeutic strategy.