Steve Ramirez knows how devastating painful memories can be. In 2015, when he was a doctoral candidate working on manipulating memories in mice, his close friend and lab partner, Xu Liu, died unexpectedly at the age of 37.

Haunted by memories of his friend, he found himself struggling to cope with his work and daily life.

Initially, Ramirez, who is now a professor of neuroscience at Boston University, wanted to numb the memories, which, on the face of it, may have seemed like a good idea. But the way he went about it had serious consequences. He turned to alcohol – and developed a severe drinking problem.

The realization that things had gone too far came years later, in February 2021, during a very average day in the office.

“I was thinking, ‘Okay, I’ve had three drinks and it’s 11am. How many more can I have throughout the day without being a mess at night?’” he says. As he pondered this, a rehab facility phoned to offer help – concerned friends had contacted them on his behalf.

“It clicked that in order for me to go to rehab, I’d have to take leave from work, and I wouldn’t really be around my friends or family,” he says. “The idea of losing everything suddenly became very real.”

A few days later, his friends and wife staged an intervention. The combination of events was eye-opening. “If there was ever a wake-up call, this was certainly it. And the next day, I went to my first group meeting.”

The meetings changed everything and Ramirez hasn’t had a drink since. “The group offered a connection with other people that I realised had been missing in my life since Xu died.”

Since then, Ramirez’s research has flourished and his memories of Liu’s death are helping to drive it forward, instead of being tied to feelings of helplessness.

Moreover, what he, and others like him, have found is that not only is it possible to manipulate painful memories like these, but you can do it without the harmful effects of alcohol.

Deleting brain cells
It’s still early days for research on memory editing, with most studies having only been carried out in animals. So far, the basic principles discovered suggest a sliding scale of possible interventions, from manipulation to complete deletion.

Memories are encoded in the brain as ‘memory traces’ – the physical changes that happen in the brain as you remember something. This can involve activity in a group of specific brain cells (neurons) or the wider connections between them.

The memory trace can get stronger or weaker over time. For example, it tends to strengthen during sleep. What’s more, every time we think of the memory, we may boost, weaken or even change it.

It’s during this process, called reconsolidation, that memories are modifiable – with a window of a few hours in humans.

This knowledge, along with advances in biotechnology, has given researchers the tools to manipulate memories.

One groundbreaking paper was published by a group of Canadian researchers in 2009. The team was able to pinpoint the brain cells in mice that lit up when the animals heard a sound that they’d come to associate with a mild electric shock, making them freeze with fear when they remembered it.

The researchers then managed to destroy those brain cells by targeting them with a chemical. Remarkably, after this intervention, the mice no longer froze when they heard the sound, suggesting the fearful memory had been erased.

“It was a transformative paper,” says Ramirez. “They took advantage of the idea that some cells in the brain are more active than others and those that are more active are the ones that are more likely to become part of a memory.”

But can you be sure the memory has really been deleted based on behavior alone? Ramirez admits that, although it certainly appears that way, we can’t be absolutely certain. “The absence of evidence isn’t evidence of absence,” he says. “It could still be there.”

He was, nevertheless, intrigued and followed up with more research in collaboration with Liu.

In 2012, Ramirez, Liu and other colleagues tried another approach. Rather than erasing cells, they tried to turn a memory on by activating particular brain cells using light from a fiber-optic cable inserted into the brains of genetically engineered mice (a method called optogenetics).

They started by identifying the brain cells in the hippocampus, the brain’s memory center, that activated when the mice received an electric shock – suggesting these cells were encoding the memory.

They then put the mice in a new environment (where there was nothing to remind them of the shock) and activated those same cells with light.

The result? The mice froze with fear in the absence of either a shock or a triggering cue. It appeared they had turned a memory on through cellular stimulation alone, strengthening the idea that memory can be manipulated through targeting brain cells.

From there, the findings kept rolling in. In a paper the following year, Liu and Ramirez used a similar method to implant a false memory into mice.

They also discovered that it’s possible to change emotions linked to specific memories. In this study, they activated a positive memory in mice while they were receiving a shock.

“When we did this, the memory of the shocks became less negative,” says Ramirez.

Fast forward a decade, and Ramirez and his colleagues are still uncovering fascinating insights.

In 2022, they showed that if you activate a positive memory in mice that are simultaneously recalling a negative memory, during the all-important reconsolidation phase, you can permanently suppress it – getting rid of fear responses. “This effect lasted for months,” says Ramirez.

What about humans?
It’s unlikely that these treatments will be offered to humans anytime soon. “What we do in rodents is very likely translatable to humans.

"But we have to be clever about it because we don’t want to go in and start engineering human brain cells or invasively modulating their activity without a scientifically and medically sound plan. Who knows what kind of unsafe side-effects could end up happening,” says Ramirez.

Instead, researchers, including Prof Emily Holmes, a clinical psychologist and neuroscientist from Uppsala University in Sweden, are trying to find non-invasive methods.

“I’m a big fan of some of the basic work that’s been done on mice. This research teaches us about the basic principles of memory, such as timings. We’ve drawn quite a lot on some of the original animal work.”

While there are promising methods to manipulate memories in humans, there’s no way to permanently delete them yet. But manipulation could well be the start.

It’s possible to weaken memories in humans and stop them flashing back, as they do in conditions like PTSD, with pharmacological or behavioral interventions.

Eye movement is thought to play a role in memory and recalling painful experiences, while tracking objects can, with professional help, reduce how frightening those memories feel - Image credit: Science Photo Library
One popular, evidence-based method is cognitive behavioral therapy (CBT), in which a therapist works with a client to help them reframe negative memories to make them feel less threatening.

This is supported by the work in mice showing that you can weaken a negative memory by activating a positive one.

Another approach, eye movement desensitization and reprocessing (EMDR), trains the brain to reprocess memories so they become less frightening.

For example, when recalling the memory during the reconsolidation window, people may be asked to simultaneously focus on something else, such as sounds, tapping or a moving object, disrupting the process.

In 2012, a team of researchers showed that when human participants viewed a neutral picture, accompanied by an electric shock, a memory trace appeared in a part of the brain that processes fear, called the amygdala.

The participants were then shown the images without the shock. Some were allowed to complete the reconsolidation process after seeing it, while others had theirs disrupted by being shown the image repeatedly.

The group that had been disrupted weren’t as fearful afterwards, while their brain scans showed that the memory trace appeared to have been erased. That said, it was a laboratory experiment and there are still no clinical trials to back it up.

It therefore remains to be seen if this would work in people with long-term, complex emotional memories.