Memory Mechanisms


by Guy McKhann, M.D.

July 16, 2013

As a neurologist, I spend my time with people who are losing their memories. Today, however, I want to talk about how the brain forms memories. I wish to highlight the work of two neuroscientists, Brenda Milner and Suzanne Corkin, and their studies of a famous patient, H.M.

Dr. Milner is 95 and still going strong. She had a very accomplished career before meeting the patient that cemented her legacy. The late Henry Molaison, known as H.M. in the medical literature, had severe seizures that were not responding to medications. A surgeon removed both of Molaison’s medial temporal structures. The operation reduced the numbers of seizures he experienced, but shortly after surgery it was noted that he had a profound memory deficit, unable to take in and retain new information.

Prior to the studies of H.M., memory was thought to be dispersed throughout the cerebral cortex. The finding that removal of specific areas, specifically medialtemporal lobe structures, could lead to memory deficits caught everyone by surprise. Further studies of H.M., by Milner and her graduate student, Suzanne Corkin, found that H.M.’s problems did not apply to all types of memory. Declarative memory, the conscious memory we use for facts, experiences, and people, was gone, never to return. However, non-declarative (or procedural) memory was retained. H.M. was able to learn complex motor acts and to improve with practice, even though he had no memory of having done the task previously. In other words, researchers now could distinguish two types of memory—declarative and procedural.

Dr. Corkin followed H.M. for 46 years, primarily at M.I.T. She and her colleagues examined him repeatedly in his assisted living facility in Hartford, Conn. H.M. did not recognize her from visit to visit; he said he thought he might have known her from high school. Dr. Corkin describes his lifetime of contribution to the understanding of memory in a new book, Permanent Present Tense: The Unforgettable Life of the Amnesic Patient, H.M.

If we think of the process of memory as having three broad steps—encoding, storing, and recall—H.M.’s modified brain could not reach step 2. His problem was in transforming short-term, immediate memory to more permanent, long-term memory. Short-term memory is what you use when you gather information for immediate use, like a telephone number or hotel room number. This information is stored for short periods of time; unless you re-enforce it, there is very little chance of recall.

This transformation requires medial temporal lobe structures, particularly the hippocampus and entorhinal cortex. Both sides have to be non-active or removed. One of Milner’s colleagues, Dr. Wilder Penfield, operated on the brains of hundreds of patients unilaterally; these patients did not show similar memory deficits. Milner and Penfield did report two patients, prior to H.M., who had similar memory deficits with unilateral surgery. It was assumed, and proven through autopsy in one, that the non-operated side in these subjects did not function.

Since H.M., there have been a number of reports of similar memory deficits, but not from surgery. Bilateral removal is not performed; for many patients, doctors make sure the non-operated side is normal before doing the surgery. For most patients today with memory problems similar to that of H.M., the damage was caused by lack of blood supply and oxygen to the brain as in stoppage of the heart (cardiac arrest) or from the focal effects of a viral infection, particularly encephalitis from the herpes simpex virus.

We had such a patient at Hopkins some years ago, a government intelligence operative who was involved with a coup in a foreign country. He developed focal encephalitis from the herpes simplex virus, involving both temporal lobes, and was left with a memory deficit similar to H.M. I’d enter his hospital room, introduce myself, and tell him where he was. I could get him to repeat this information several times. However, if I left the room for even a few minutes, he’d have forgotten everything by the time I returned, including have seen me, my name, or where he was.

In terms of his memory prior to the encephalitis, he remembered much of what he had been doing prior to the coup. But then everything stopped. He had no memory of what had happened during or after. He drove his intelligence colleagues nuts. At first they didn’t believe him. Only after showing them articles about H.M. were we were able to convince them that he was not faking.

As for H.M., he died in 2009 at age 82. His brain is being evaluated by Dr. Corkin, taking advantage of new sectioning techniques developed by Dr. Jacopo Annese at UCSD. The morphological and histological results will eventually be available online for a wide range of investigators. Grants from the Dana Foundation support this project.