Decoding Memory: How Brain Waves, Drugs, and a Tiny Receptor Subunit Connect

Hey there! Let’s dive into something super interesting about how our brains work, specifically that handy mental scratchpad we call working memory. You know, the ability to hold onto information just long enough to use it? It’s crucial for pretty much everything we do, from remembering a phone number for a few seconds to following instructions.

Now, imagine that scratchpad isn’t working right. That’s a big challenge for folks dealing with conditions like schizophrenia. While current treatments help with some symptoms, they often don’t touch these cognitive issues, especially working memory problems. And honestly, these cognitive struggles are often the biggest hurdle to living independently and finding work. So, understanding *why* working memory goes awry is a huge deal.

Scientists like me are always trying to figure out the brain’s secrets. We know that working memory relies on the brain’s electrical activity, specifically these synchronized pulses or “brain waves” called oscillations. Gamma oscillations, buzzing away between 30 and 100 Hz, seem particularly important for higher-level thinking and memory tasks. Interestingly, in people with schizophrenia, we often see weird patterns in these gamma waves – maybe too much baseline activity but not enough when they’re actually trying to *do* a memory task.

This is where things get really intriguing. There’s a type of brain receptor called the NMDA receptor (NMDAR). These receptors are like gatekeepers for signals between brain cells, and they’re known to be involved in learning and memory. When these receptors aren’t working correctly (a concept called NMDAR hypofunction), it’s been linked to schizophrenia symptoms. And guess what? Drugs that block NMDARs – antagonists like ketamine, PCP, and MK-801 – can actually *mimic* some of the symptoms of schizophrenia, including those changes in gamma oscillations!

So, I and my colleagues thought, “Okay, if these drugs mess with both working memory and gamma waves, and they model schizophrenia symptoms, maybe there’s a direct link between the two?” We wanted to explore this connection and also figure out if a specific piece of the NMDA receptor, a subunit called GluN2D, plays a special role.

Putting Working Memory to the Test

To do this, we turned to our amazing mouse friends. We used a clever task called the Trial-Unique-Non-match to Location (TUNL) on a touchscreen. Think of it like a mini-game for mice that tests their working memory. They see a light in one spot, remember it during a short delay, and then have to touch a *new* light that appears alongside the old one to get a treat. It’s a neat way to test memory that’s similar to how we test cognition in humans, making the findings more relevant.

We trained both regular “wildtype” mice and special “knockout” mice that were genetically engineered to lack the GluN2D subunit. Once they were pros at the TUNL task, we gave them different NMDAR antagonist drugs (PCP, MK-801, S-ketamine, R-ketamine) or just saline (salty water, our control) and watched how they performed. Crucially, while they were doing the task, we were also recording their brain activity using tiny electrodes implanted in key memory areas: the prefrontal cortex (PFC) and the hippocampus. This allowed us to see the brain waves *as* they were trying to remember things.

What We Found: Memory and the GluN2D Link

The first big takeaway was about working memory accuracy. As expected, most of the drugs messed up the mice’s performance. They made more mistakes on the TUNL task. But here’s the kicker: PCP, one of the drugs, only disrupted working memory in the *regular* mice. The mice missing the GluN2D subunit? They were surprisingly protected from PCP’s negative effect on memory! This strongly suggests that the GluN2D subunit is really important for PCP to have its working memory-disrupting effect.

MK-801, another drug, impaired working memory in *both* types of mice, meaning its effects likely aren’t specifically tied to the GluN2D subunit. The ketamines (S and R) also impaired memory, but their effects seemed shorter-lived, mainly showing up in the first part of the testing session. This fits with how quickly ketamine is cleared from the body compared to PCP or MK-801.

Interestingly, the GluN2D-knockout mice performed just as well as the regular mice on the TUNL task when they *weren’t* on drugs. This tells us that just missing the GluN2D subunit on its own doesn’t necessarily cause working memory problems, but it makes the brain less vulnerable to the effects of certain drugs like PCP.

Brain Waves: Baseline Buzz and Task-Related Rhythms

Next, we looked at the brain waves, specifically gamma oscillations. We checked the “baseline” gamma power – the brain’s general buzz when the mice weren’t actively doing the task (during the waiting period between trials).

In the prefrontal cortex (PFC), most of the drugs increased this baseline gamma power. This is similar to what’s seen in people with schizophrenia. The GluN2D-knockout mice actually had higher baseline gamma in the PFC overall, and they showed a heightened response to PCP in the PFC compared to regular mice.

In the hippocampus, a brain area critical for memory, things were a bit different. Only PCP significantly increased baseline gamma power, and just like with working memory, this effect was only seen in the *regular* mice, not the GluN2D-knockouts. This is another piece of evidence pointing to GluN2D’s specific role in how PCP affects the hippocampus and, potentially, memory. The GluN2D-KO mice also had lower baseline low gamma in the hippocampus compared to regular mice.

But the really exciting part came when we looked at brain activity *during* the task, especially during the “maintenance phase” – that short delay where the mouse has to hold the location information in mind. We saw a clear increase in low gamma power (30-40 Hz) in the hippocampus during this maintenance phase. And here’s the key: this increase was significantly *higher* when the mouse was going to make a *correct* choice compared to an incorrect one! It’s like this specific low gamma rhythm in the hippocampus is an electrical signature of successful working memory maintenance.

Then, we gave the mice the drugs again and looked at this task-induced low gamma. In the saline-treated mice, we still saw that nice bump in low gamma during correct trials. But in mice treated with *any* of the NMDAR antagonists (PCP, MK-801, S-ketamine, R-ketamine), that difference disappeared. The task-induced low gamma didn’t increase significantly during correct trials anymore. This strongly suggests that these drugs disrupt working memory by messing up this specific, task-related low gamma activity during the maintenance phase.

Connecting the Dots

So, what does all this tell us? Well, for starters, it confirms that NMDAR antagonists are powerful tools for studying working memory problems and related brain activity changes that we see in conditions like schizophrenia.

More specifically, our findings highlight the GluN2D subunit as a key player in the effects of PCP on both working memory and hippocampal baseline gamma. This subunit is particularly found on certain types of brain cells called PV interneurons, which are crucial for generating gamma oscillations and regulating brain circuits. It seems PCP might be disrupting working memory partly by acting on GluN2D receptors on these interneurons, throwing off the delicate balance of brain activity.

While MK-801 also disrupts working memory, it doesn’t seem to rely on GluN2D, suggesting different NMDAR subtypes or even other brain systems might be involved in its effects.

Perhaps most importantly, we’ve identified a specific brain wave pattern – low gamma in the hippocampus during memory maintenance – that seems directly tied to successful task performance and is disrupted by *all* the NMDAR antagonists we tested. This gives us a concrete electrophysiological target to study further when thinking about how these drugs, and potentially conditions like schizophrenia, impair working memory.

It’s a complex puzzle, but each piece helps us get closer to understanding the intricate ways our brains handle memory and how things can go wrong. By linking specific receptor subunits, brain regions, and oscillation patterns to working memory performance, we’re hopefully paving the way for better treatments down the line.

Source: Springer