After the last day of the academic year, I had separated my school keys from my home set. But weeks later, when I went to the top drawer of my dresser, where I usually keep seldom-used keys, they were not there. The first thought that came to mind was that I was just tired; I was probably not looking carefully enough. But two systematic searches later and still no luck.
Then I began to worry about my silly, unrecognized streak of decades of never losing keys. Ever since my baseball scoreless streak was snapped at one and one third innings---thanks to my 65 mile per hour fast ball and slow-motion slider---I’ve always looked to set other kinds of records.
For example, I was never late or absent throughout high school. And although the police have never come knocking on my door to present me with a trophy, I have never been stopped for a traffic violation in 27 years.
Brushing such thoughts aside, I remembered that when I had put the keys away, a little voice in my head had warned, “Hmm... This is not where you usually place them. Won’t you forget?”
And naturally, the other little voice in my head said, “Of course not.”
Frustrated, I told my wife that she should have married the first little voice.
After cross-examining my daughter (to get her to cooperate, I was nice, not accusatory) and after we checked numerous possible and impossible drawers in various locations, I temporarily gave up, and started to clean house and put things away.
When I closed the zipper on the camera case, it all came back to me. The keys were in a zippered pocket of my school back pack. It was such a vivid memory that I did not even check the back pack until the next day.
My lost keys adventure got me wondering. On a molecular level, how do memories form? How do associations help retrieve memories? Does neuroscience have any clue? If I had never recovered the keys, there would have been a strong possibility that very little memory consolidation had occurred in my brain. Consolidation happens when an experience spins a web of molecular and cellular events resulting in a durable form of synaptic modification between neurons.
What does the web consist of? Not surprisingly, biochemistry's common second messenger, cyclic adenosine monophosphate ( cAMP) and the associated ion Ca2+ play a role. It is also not a shock to see the involvement of protein kinases (specifically mitogen-activated protein kinases and tyrosine kinases). After cell membrane receptors are activated, ATP is turned into cAMP inside the cell. This controls the passage of Ca2+ and activates the kinases, which through the simple attachement of a phosphate group, modify the chemistry and function of proteins.Neuroscientists have also elucidated the role of genes in memory consolidation. Some of the proteins involved in memory(C/EBPs, c-Fos and zinc finger protein 225--all transcription factors) bind to DNA and control the flow of information to messenger RNA. The immediate-early gene (IEG) family plays a role too. These are quickly transcribed in the presence of a protein inhibitor.
In what brain structures does all this happen? Damage to the hippocampus impairs multi-year old human memories related to factual, events and general knowledge. Also affected are animal contextual memories that are up to 30 days old. These observations have led to the idea that the hippocampus initially works with the neocortex to consolidate memory but gradually becomes less important. In contrast, as time passes, changes in the neocortex play an increasingly vital role in memory by networking various areas of the cortex.
Structures such as the hippocampus and amygdala have often been the subject of investigations into reconsolidation of memories. 
Retrieving a consolidated memory can actually alter the memory. Reconsolidation is the process that stabilizes the memory once again, and it also evolves with the age of the memory; older ones are less sensitive to disruption. But of course this implies that since an older memory likely went through a labile stage, its overall integrity is still compromised. Mostly though the use of protein inhibitors in chick, rats, mice and gerbils and through PET scans in humans, researchers have discovered that although the hippocampus, amygdala and auditory cortex are involved in memory consolidation after initial training, these structures are not always needed for reconsolidation. There are exceptions, however. In taste avoidance and fear conditioning both processes need protein synthesis in the same brain areas.
Thanks to recent research efforts, our understanding of memory has become less pixelated. But the overall picture is still fuzzy. Neuroscientists are an even longer way from explaining what happens in the brain when a concept is understood. There are surely consolidation-reconsolidation mechanisms involved, and they'll probably prove to be even more intricate than those operating in memory formation. Such a framework undoubtedly facilitates the assimilation,consolidation and distortion of additional memories. And we can only imagine what has to happen in the brain to modify or dislodge a deeply implanted, erroneous idea.
Sources:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3056265/
Alberini, C.M. Mechanisms of memory stabilization: are consolidation and reconsolidation similar or distinct processes? Trends in Neuroscience, 28, 51-56. 2005
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2680680/?tool=pubmed