The refrigerators against the back wall ofJacopo Annese’s lab are filled with humanbrains. Some float in tubs of formaldehyde;others have been sliced into ultrathin sec-tions to be treated with stains that will revealvarious anatomical features. One of the tubsis labeled HM, after Henry Molaison, themost studied human being in the history ofpsychology. In 1953, an experimental brainsurgery intended to correct a devastatingseizure disorder left Molaison unable to formnew memories. At age 27, he became frozenin time. He could remember facts he’d learnedand names of people he’d encountered beforethe surgery but virtually nothing after it. Forhalf a century, until his death last December,Molaison gamely cooperated with psycholo-gists and neuroscientists, and his casereshaped scientific thinking on the neuralbasis of memory. Annese, 42, is a neuroanatomist at theUniversity of California, San Diego (UCSD),and he has been charged with preservingMolaison’s brain for perpetuity. He takes thisresponsibility seriously. Among other pre-cautions, the fridge containing Molaison’sbrain has an alarm system that calls Annese’soffice, home, and cell phones if the tempera-ture deviates too far from the set point. He hasalso taken certain security measures he’sloath to discuss on the record. “It probablysounds a bit paranoid,” he says, “but we’vetried to think of anything that could go wrongand eliminate the risk.”With funding from the Dana Foundation,a New York City nonprofit that supportsresearch in neuroscience, and the NationalScience Foundation, Annese plans to createa digital, zoomable atlas of Molaison’s brainand make it freely available online, the firstentry in what he hopes will become an open-access brain library to be used by scientists,students, or anyone with an Internet connec-tion and an interest in neuroanatomy. Ulti-mately, the library will include donatedbrains from people with Alzheimer’s diseaseand other neurological disorders, Annesesays, as well as a collection of healthy brainsof different ages. But for now, Molaison’sbrain is the star attraction. “More was learned about memory byresearch with just that one patient than waslearned in the previous 100 years of researchon memory,” says Vilayanur Ramachandran,a behavioral neurologist at UCSD who is notdirectly involved with Annese’s project.Postmortem anatomical studies could revealnew clues about thecause of Molaison’samnesia by revealingmore precisely theextent of his surgicallesions, as well assubsequent degener-ation, Ramachandran says. Preserving thedonated brains of such unique patients hasboth historical and scientific value, he says.A long historyMolaison’s story begins decades beforeAnnese got involved. Henry Gustav Molaisonwas born in 1926 and grew up in and aroundHartford, Connecticut. His seizures began atage 10. The cause is uncertain, but his father’sfamily had a history of epilepsy, and a colli-sion with a passing bicycle rider had onceknocked Henry unconscious for several min-utes. The seizures worsened through his teensand 20s until frequent blackouts and increas-ingly severe convulsions left him unable tocontinue his work repairing electric motors,despite high doses of anticonvulsant drugs.He ended up in the care of William BeecherScoville, a neurosurgeon at Hartford Hospital.Because of the incapacitating nature ofMolaison’s seizures, Scoville decided to trywhat he later described as a “frankly experi-mental operation.” He removed, via suction, afinger-sized piece of the temporal lobes onboth sides of the brain, including most of thehippocampus, amygdala, and nearby para-hippocampal gyrus. Scoville had previouslyperformed such bilateral medial temporallobotomies on dozens of psychiatric patients,hoping to calm their psychosis without thepersonality changes associated with the moredrastic frontal lobotomies that were begin-ning to fall out of favor at the time. By coinci-dence, two of Scoville’s previous patients hadbeen prone to seizures, and those diminishedpostsurgery, says Suzanne Corkin, a neuro-scientist at the Massachusetts Institute ofTechnology in Cambridge who worked withMolaison for 46 years and is writing his biog-raphy. As a last resort, Scoville thought thesurgery was worth a try with Molaison. In a landmark 1957 paper, Scoville andCanadian psychologist Brenda Milnerreported that although Molaison’s seizureshad diminished greatly, he now exhibitedprofound amnesia. He easily recalledevents prior to his operation but had virtu-ally no lasting memories of anything thathad happened since. A half-hour after hav-26 JUNE 2009 VOL 324 SCIENCE www.sciencemag.org1634CREDIT: (BANNER) JACOPO ANNESE, PETER J. CHON, AND ALEX GHATAN The Brain CollectorJacopo Annese plans to create an open-access digital brain library—starting with the most famousamnesic patient of all timeNEWSFOCUSOnlinePodcast interviewwith author Greg Miller. sciencemag.orgPublished by AAAS on July 12, 2009 www.sciencemag.orgDownloaded froming lunch with Milner in the hospital cafete-ria, for example, he couldn’t name a singlething he had eaten and did not even remem-ber eating. The paper also described moder-ate to severe memory deficits in seven otherpatients with similar surgeries, mainly forschizophrenia. In the coming years, Scovilleurged other surgeons not to perform this sur-gery on both sides of the brain.Memory lessonsMolaison’s case startled mem-ory researchers. Conventionalwisdom at the time held thatmemory could not be pinnedto any specific location in thebrain. This view sprang largelyfrom the work of Karl Lashley,an American physiologist who in the 1930s and ’40shad searched in vain for the“engram,” or neural trace ofmemory, by removing differentregions of the cerebral cortex ofrats to see whether this woulddestroy a previously learnedmemory, such as the locationof food in a maze. Becauseremoving even large swathsof cortex did not render therodents amnesic, Lashley con-cluded that memory tracesmust be widely distributed.Molaison’s case suggestedotherwise. His ability to learn and retain newfacts, what researchers now call declarativememory, had been devastated by removing arelatively small part of his brain. (Lashley, cer-tain he would find the engram in the cerebralcortex, never tried removing the hippocampus.) Further studies provided additional cluesabout the biology of memory. AlthoughMolaison was unable to create long-lasting newmemories, his recollection