Latest / Clinical Deep Dives / PSYCH 022: Animal Models in Psychiatry

Transcript

PSYCH 022: Animal Models in Psychiatry · Clinical Deep Dives

0:00
SPEAKER_01: How do you ask a mouse if it's hallucinating?
0:02
SPEAKER_01: You know, you really can't.
0:03
SPEAKER_01: A rodent cannot pull up a chair,
0:05
SPEAKER_01: sit across from a clinician,
0:07
SPEAKER_01: and just articulate some complex,
0:08
SPEAKER_01: systematized delusion about the CIA tracking its movements.
0:12
SPEAKER_00: No, obviously not.
0:13
SPEAKER_01: Right.
0:14
SPEAKER_01: But in trying to solve that completely impossible problem,
0:17
SPEAKER_01: we've actually accidentally mapped the biological engines of human fear,
0:22
SPEAKER_01: addiction,
0:22
SPEAKER_01: and despair.
0:24
SPEAKER_01: At its core,
0:25
SPEAKER_01: psychiatry is the study of how brain,
0:27
SPEAKER_01: mind,
0:27
SPEAKER_01: and experience interact,
0:29
SPEAKER_01: and crucially,
0:29
SPEAKER_01: how that incredibly delicate process can go wrong.
0:32
SPEAKER_00: Right.
0:32
SPEAKER_01: But when we try to approximate the human mind in an animal,
0:35
SPEAKER_01: it's essentially like trying to understand the vast,
0:37
SPEAKER_01: unpredictable open ocean by studying a controlled,
0:41
SPEAKER_01: shallow tide pool.
0:42
SPEAKER_00: I love that.
0:43
SPEAKER_00: And that is exactly the tension we navigate every single day in the lab.
0:46
I mean,
0:46
the tide pool operates on the exact same fundamental physical
0:49
principles,
0:50
right?
0:50
You've got water,
0:51
gravity,
0:52
salinity,
0:52
the basic biological drives for survival.
0:55
SPEAKER_00: So a tide pool isolates those specific mechanisms beautifully.
0:59
You can watch a crab scavenge or barnacle filter feed without a
1:02
hurricane getting in the way.
1:03
SPEAKER_01: Yeah,
1:04
SPEAKER_01: but the tide pool completely lacks the crushing pressures,
1:07
SPEAKER_01: the terrifying lightless depths,
1:09
SPEAKER_01: and the sweeping currents of the open sea.
1:12
SPEAKER_01: Animal models are our psychiatric tide pools.
1:14
SPEAKER_01: They isolate the mechanism,
1:16
but they can't possibly capture the existential entirety of the
1:20
human experience.
1:20
SPEAKER_00: And to be fair, we really shouldn't expect them to.
1:23
SPEAKER_00: Abstract thought,
1:24
SPEAKER_00: complex language,
1:25
SPEAKER_00: the uniquely human capacity for self-reflection,
1:28
SPEAKER_00: and the punishing rumination that often stems from it,
1:30
SPEAKER_00: those are features of the open ocean.
1:32
SPEAKER_00: But underneath all that, our subcortical structures are ancient.
1:36
SPEAKER_01: The evolutionary engine room.
1:37
SPEAKER_00: Exactly.
1:38
The amygdala,
1:38
the hippocampus,
1:39
the stratum,
1:40
those are highly conserved across species.
1:42
The hardware running a mouse's fear response is stunningly
1:46
similar to the hardware
1:47
SPEAKER_00: running yours.
1:48
SPEAKER_01: So, welcome to this deep dive.
1:49
We're going straight to the core of how we translate a mouse's
1:53
freezing behavior,
1:54
SPEAKER_01: or a dog's sudden sleep attack,
1:56
SPEAKER_01: into actionable neurobiological insights for human mental illness.
2:01
SPEAKER_01: And let's unpack the core problems straight away,
2:03
because a lot of people look at this field and think it's
2:04
inherently a fool's
2:06
SPEAKER_01: errand.
2:07
Looking for schizophrenia in a rat feels a bit like looking for
2:11
a Wi-Fi signal in a toaster.
2:12
Historically yes that was a massive stumbling block for a long
2:15
time the field was
2:16
completely obsessed with creating a schizophrenic rat or
2:19
depressed mouse right
2:21
researchers were trying to model entire psychiatric disorders
2:24
but there was a
2:24
paradigm shift a couple decades ago when we collectively just
2:28
well admitted defeat
2:29
on that front we realized you simply cannot model entire human
2:33
psychiatric
2:33
SPEAKER_01: disorders in animals because the human diseases are incredibly messy
2:37
SPEAKER_00: Highly heterogeneous, yes.
2:38
SPEAKER_00: The diagnostic criteria we use for humans,
2:41
SPEAKER_00: the DSM or ICD,
2:42
SPEAKER_00: they are based on subjective,
2:44
SPEAKER_00: self-reported symptoms.
2:45
SPEAKER_00: You know,
2:46
two humans can both be diagnosed with major depressive disorder
2:49
and share almost
2:50
SPEAKER_00: zero overlapping symptoms.
2:52
SPEAKER_01: Oh, absolutely.
2:53
SPEAKER_01: One can't sleep, won't eat, and is agitated.
2:56
SPEAKER_01: The other sleeps 16 hours a day, overeats, and can't get out of bed.
2:59
SPEAKER_00: Precisely.
3:00
To create a genetic mouse line that encapsulates all of that
3:03
contradiction is
3:04
SPEAKER_00: scientifically impossible.
3:07
so instead of looking for schizophrenia as a whole we break it
3:09
down into Lego
3:10
bricks we look for a specific symptom like an inability to
3:15
filter out background
3:16
noise we call those symptom domains things like negative
3:20
valence which covers fear
3:22
and anxiety or positive valence covering reward and addiction
3:25
or cognitive deficits
3:27
these domains cross diagnostic boundaries so a deficit in
3:31
working memory is a
3:31
hallmark of schizophrenia but it's also present in severe
3:34
depression and ADHD makes
3:35
SPEAKER_01: sense
3:36
By looking for well-defined phenotypes observable traits with
3:38
known biological
3:39
SPEAKER_00: bases and strong genetic homologies between humans and animals,
3:43
SPEAKER_00: we can do highly precise science.
3:45
Which brings us to the ultimate question of whether that
3:47
science is actually any good.
3:49
SPEAKER_01: How do we know if the model works?
3:52
SPEAKER_01: It all comes down to validity.
3:54
SPEAKER_01: There's this sort of holy trinity of validity in the literature.
3:57
SPEAKER_01: First we have face validity, which seems like the most intuitive one.
4:00
SPEAKER_00: It is.
4:01
SPEAKER_00: Face validity just asks on its face,
4:03
SPEAKER_00: does the animal's behavior physically resemble the human symptom?
4:08
SPEAKER_00: If we're modeling panic,
4:10
SPEAKER_00: and a rat exhibits frantic,
4:12
SPEAKER_00: darting flight behavior when a predator gets close,
4:15
SPEAKER_00: well that looks like human panic,
4:16
SPEAKER_00: it has strong face validity.
4:18
SPEAKER_01: But appearances are notoriously deceptive in neurobiology.
4:21
SPEAKER_01: I mean,
4:22
I might be crying because I'm depressed or I might be crying
4:24
because I'm chopping
4:24
SPEAKER_01: an onion.
4:25
SPEAKER_01: The output is exactly the same,
4:27
SPEAKER_01: but the mechanism is vastly different,
4:28
SPEAKER_01: which is why we need the second pillar,
4:30
SPEAKER_01: which is construct validity.
4:31
SPEAKER_00: Yes, construct validity is much more rigorous.
4:34
SPEAKER_00: It demands that the animal model reproduces the actual etiology,
4:38
SPEAKER_00: the underlying biological root cause of the human disorder.
4:42
SPEAKER_01: Can you give an example of that?
4:44
SPEAKER_01: Sure.
4:44
Let's say we find a specific highly penetrant genetic mutation
4:48
in human families
4:50
SPEAKER_00: that causes a neurodevelopmental disorder.
4:53
If we use gene editing to put that exact human mutation into a
4:57
mouse's DNA and the
4:58
SPEAKER_00: mouse develops the corresponding cognitive deficits.
5:00
SPEAKER_01: Then you've nailed it.
5:01
SPEAKER_00: Exactly.
5:02
SPEAKER_00: That model has exceptional construct validity.
5:05
SPEAKER_00: We aren't just mimicking the symptom.
5:07
SPEAKER_00: We've recreated the biological spark.
5:09
SPEAKER_01: And then there's predictive validity,
5:10
SPEAKER_01: which is essentially what the pharmaceutical industry lives and dies by.
5:14
SPEAKER_00: Right.
5:14
SPEAKER_00: Predictive validity evaluates whether a manipulation in the animal,
5:18
SPEAKER_00: usually giving it a drug,
5:19
SPEAKER_00: accurately predicts clinical effectiveness in humans.
5:22
SPEAKER_00: If a novel compound reduces freezing behavior in a mouse,
5:25
SPEAKER_00: does it actually cure anxiety in a human patient?
5:28
A model with high predictive validity has very low false
5:31
positive and false negative rates.
5:33
SPEAKER_01: And here's where it gets really interesting and,
5:35
SPEAKER_01: frankly,
5:36
SPEAKER_01: quite dangerous for clinical reasoning.
5:38
SPEAKER_01: Because you can have a test with fantastic face validity,
5:43
SPEAKER_01: but absolutely terrible predictive validity.
5:46
SPEAKER_01: Think of it like trying to measure wind speed with a thermometer.
5:48
SPEAKER_01: The thermometer is a great tool.
5:50
SPEAKER_01: It measures weather.
5:51
SPEAKER_01: But if you're trying to figure out if a hurricane is coming,
5:54
SPEAKER_01: the temperature reading is going to give you a massive false negative.
5:57
SPEAKER_00: And that happens all the time in preclinical psychopharmacology.
6:01
SPEAKER_00: An animal might look sad,
6:02
SPEAKER_00: it stops moving,
6:03
SPEAKER_00: it stops eating,
6:04
SPEAKER_00: so a researcher gives it a new drug,
6:06
SPEAKER_00: the animal starts moving again,
6:08
SPEAKER_00: and they rush to publish that they've cured depression.
6:10
SPEAKER_01: But then it fails in humans.
6:12
SPEAKER_00: Exactly.
6:13
If the drug consistently fails in human clinical trials,
6:16
the model lacked predictive validity.
6:18
We have to be incredibly disciplined about not taking animal
6:21
behaviors purely at face value.
6:23
SPEAKER_01: Let's apply that discipline right now, actually.
6:25
SPEAKER_01: I'm going to jump into our first major domain, which is negative balance.
6:29
SPEAKER_01: This is the dark side of the emotional spectrum.
6:32
SPEAKER_01: Fear, anxiety, hopelessness.
6:34
It's rooted heavily in those ancient subcortical survival
6:38
circuits you mentioned earlier.
6:39
SPEAKER_00: Specifically,
6:40
SPEAKER_00: we're talking about structures like the amygdala,
6:42
SPEAKER_00: the periaqueductal gray,
6:44
SPEAKER_00: and the bed nucleus of the citraea terminalis.
6:46
SPEAKER_00: These areas govern emotional states of aversiveness or discomfort.
6:50
SPEAKER_01: We colloquially use fear and anxiety interchangeably,
6:54
SPEAKER_01: but neurologically,
6:55
SPEAKER_01: they're entirely different beasts.
6:56
SPEAKER_01: It's all about the proximity of the threat.
6:58
SPEAKER_01: Let's use a metaphor.
7:00
Fear is when you're sitting in your office and suddenly thick
7:03
black smoke starts
7:04
SPEAKER_01: pouring under the door.
7:05
SPEAKER_01: You are in immediate acute danger.
7:08
SPEAKER_01: Anxiety is when you're sitting in your office,
7:09
SPEAKER_01: you hear a fire alarm three blocks away,
7:12
and you spend the next hour wondering if the wind is going to
7:14
shift in your
7:14
SPEAKER_01: direction.
7:15
That distinction maps perfectly onto the ethological findings
7:18
in rodents.
7:19
SPEAKER_00: The brain processes distal threats,
7:21
SPEAKER_00: the fire alarm three blocks away,
7:23
SPEAKER_00: completely differently than acute imminent threats.
7:26
SPEAKER_00: When a threat is distal,
7:27
SPEAKER_00: the animal engages in anxiety behaviors like freezing,
7:30
SPEAKER_00: concealment,
7:31
SPEAKER_00: and risk assessment.
7:32
SPEAKER_01: Keeping low profile.
7:33
SPEAKER_00: Yes.
7:34
SPEAKER_00: The animal lowers its posture to avoid detection.
7:37
SPEAKER_00: But when the threat is acute,
7:38
SPEAKER_00: the smoke under the door,
7:40
the animal transitions immediately into explosive flight or
7:43
defensive panic
7:44
SPEAKER_00: attacks.
7:45
So how do we trigger that acute panic in a lab setting the most
7:49
direct ethological
7:50
method is the defense test battery you place a rat in a long
7:54
circular runway then
7:55
an experimenter introduces a predator sometimes a live cat
7:59
sometimes a robotic
8:00
predator that begins closing in on the rat a robotic predator
8:03
yeah that's brilliant
8:05
but I assume that's primarily for construct validity so the
8:08
threat is exactly the
8:09
same every single time it ensures absolute standardization of
8:12
the stimulus yes
8:13
As the robotic predator approaches we quantify the rats flight
8:17
response if we
8:18
suddenly block the runway creating an unavoidable threat the
8:22
rats brain shifts
8:23
gears again it goes from flight to defensive threat and attack
8:26
it turns around and
8:27
fights exactly this model has excellent face validity and more
8:31
importantly drugs
8:32
like benzodiazepines which we use to treat human panic attacks
8:36
reliably reduce this
8:37
SPEAKER_00: runway flight behavior
8:38
SPEAKER_01: But dealing with live predators,
8:40
SPEAKER_01: or complex circular runways,
8:42
SPEAKER_01: introduces a lot of mechanical noise.
8:44
There's a much more elegant,
8:45
purely neurological way to induce panic,
8:47
isn't there?
8:48
Yes,
8:49
bypassing the external environment entirely and going straight
8:52
to the brain stem.
8:53
We can electrically stimulate the dorsal periaqueductal gray or
8:57
the D-Pag.
8:58
SPEAKER_01: What exactly is happening in the D-Pag?
9:00
SPEAKER_00: The DPag harbors a highly sensitive hypoxia alarm system.
9:03
SPEAKER_00: It's basically the brain's internal sensor for suffocation.
9:07
When you stimulate it,
9:08
the rat exhibits profound,
9:10
explosive,
9:11
panic-like responses.
9:13
SPEAKER_00: It's not just running.
9:14
SPEAKER_00: It's a frantic, desperate escape behavior.
9:17
SPEAKER_01: And this is where the construct's validity shines.
9:19
SPEAKER_01: Because human panic disorder isn't just one monolithic thing.
9:22
SPEAKER_01: Some people get heart palpitations, some get dizzy.
9:25
But there is a specific respiratory subtype of panic disorder
9:28
where patients feel
9:29
SPEAKER_01: this overwhelming,
9:30
SPEAKER_01: terrifying sense of smothering or air hunger.
9:33
SPEAKER_00: Exactly.
9:33
The DPAC stimulation in rats perfectly models that specific
9:37
respiratory subtype in humans.
9:39
SPEAKER_00: It shows us that this type of panic isn't a cognitive failing.
9:42
SPEAKER_00: It's a false alarm in the brain's deepest suffocation circuitry.
9:46
SPEAKER_01: Okay,
9:46
SPEAKER_01: let's shift from that unconditioned raw panic to conditioned fear,
9:50
SPEAKER_01: because this is the gateway into learning theory.
9:52
SPEAKER_01: And it's how we start to unpick the neurobiology of trauma and PTSD.
9:56
SPEAKER_00: Conditioned fear relies on standard Pavlovian principles.
10:00
SPEAKER_00: You take a neutral condition stimulus, let's say a specific audio tone.
10:04
SPEAKER_00: You play the tone and immediately follow it with an aversive,
10:08
SPEAKER_00: unconditioned stimulus,
10:09
SPEAKER_00: which is usually a very mild electrical foot shock.
10:12
It only take a few pairings for the amygdala to wire those two
10:15
things together.
10:16
SPEAKER_01: Tone equals shock.
10:17
SPEAKER_00: It's extremely rapid, robust learning.
10:20
SPEAKER_00: Once the association is formed, we can measure the learned fear.
10:23
SPEAKER_00: A common way is fear-potentiated startle.
10:26
SPEAKER_00: If you play a loud noise to a rat, it'll flinch.
10:29
SPEAKER_00: That's the baseline acoustic startle reflex.
10:31
But if you play that same loud noise while the fear conditioned
10:34
tone is playing in
10:35
SPEAKER_00: the background,
10:36
SPEAKER_00: the rat will flinch significantly more violently.
10:39
SPEAKER_00: The underlying state of fear potentiates the acoustic startle.
10:42
SPEAKER_01: We see that in humans with PTSD.
10:44
A car backfiring might make me jump,
10:46
but it makes a combat veteran hit the deck.
10:48
SPEAKER_01: The baseline threat arousal just magnifies the startle.
10:52
SPEAKER_00: Absolutely.
10:53
But the truly vital clinical application here isn't just
10:56
understanding how fear is learned.
10:57
SPEAKER_00: It's understanding how it's extinguished.
10:59
SPEAKER_01: Extinction.
11:00
SPEAKER_01: This is the biological bedrock of exposure therapy.
11:03
If you take that fear conditioned rat and you play the tone
11:07
over and over again,
11:07
SPEAKER_00: 50 times maybe,
11:09
SPEAKER_00: but you never deliver the foot shock,
11:11
SPEAKER_00: the brain starts to learn a new rule.
11:13
SPEAKER_00: The rat learns that the tone is now safe.
11:16
SPEAKER_00: The conditioned freezing behavior gradually diminishes.
11:19
SPEAKER_00: This is extinction.
11:20
SPEAKER_01: And crucially, extinction is not forgetting.
11:23
SPEAKER_01: The rat hasn't forgotten the shock.
11:25
It's actively laying down a new inhibitory memory trace right
11:30
over the old fear memory.
11:32
SPEAKER_00: That's a critical distinction.
11:33
SPEAKER_00: And because it's an active learning process,
11:35
SPEAKER_00: it relies heavily on neuroplasticity,
11:38
SPEAKER_00: specifically the NMDA glutamate receptors in the amygdala.
11:41
SPEAKER_00: This gave us a massive pharmacological target.
11:44
SPEAKER_00: Researchers reasoned that if we can boost NMDA receptor function,
11:48
SPEAKER_00: maybe we can speed up this extinction learning.
11:50
SPEAKER_01: Which led to the trials with desicloserine.
11:52
SPEAKER_00: Yes.
11:52
SPEAKER_00: Desicloserine is a partial agonist at the NMDA receptor.
11:56
SPEAKER_00: In the animal models,
11:57
giving desicloserine before extinction training significantly
12:00
accelerated the
12:01
SPEAKER_00: reduction in fear.
12:02
SPEAKER_00: It basically made the exposure therapy work faster and sick better.
12:05
And that translated directly to the clinic for human-specific
12:08
phobias and social anxiety.
12:11
Giving a patient decycloserine just before an exposure therapy
12:14
session like putting
12:16
them in a glass elevator if they have a fear of heights
12:19
actually improves the
12:20
SPEAKER_01: clinical outcome.
12:21
Though I know the literature is a bit messier when it comes to
12:24
complex PTSD.
12:25
SPEAKER_00: Well,
12:26
SPEAKER_00: yes,
12:26
the clinical reality of PTSD often involves multiple compounded
12:30
traumas over years,
12:31
which is vastly more complex than a single tone shock pairing
12:34
in a pristine lab
12:36
SPEAKER_00: environment.
12:36
Let's look at more generalized anxiety,
12:38
where there isn't a specific tone or a specific trauma.
12:41
SPEAKER_01: The classic models here revolve around approach avoidance conflicts.
12:44
SPEAKER_00: Rodents are natural foragers,
12:46
SPEAKER_00: so they possess an innate drive to explore novel environments for food,
12:50
SPEAKER_00: but they are also prey animals.
12:52
Out in the wild,
12:52
an open,
12:53
brightly lit,
12:54
elevated space is exactly where a hawk will snatch them.
12:57
So their neurobiology is constantly weighing the drive to
13:00
explore against the drive
13:01
SPEAKER_00: to hide,
13:02
SPEAKER_00: and we exploit that conflict.
13:04
SPEAKER_01: The elevated plus maze is the poster child for this.
13:06
SPEAKER_01: It's exactly what it sounds like.
13:08
SPEAKER_01: A raised platform shaped like a plus sign.
13:11
SPEAKER_01: Two arms have high, dark walls so they feel safe and enclosed.
13:15
The other two arms are completely open,
13:16
essentially just diving boards into the abyss.
13:19
SPEAKER_01: You put a mouse in the center.
13:21
SPEAKER_01: A calm mouse will eventually creep out onto the open arms to explore.
13:25
SPEAKER_01: An anxious mouse will stay huddled in the dark, walled arms.
13:28
We use similar paradigms like the light dark box where the
13:31
animal chooses between a
13:32
brightly illuminated chamber and a pitch black one or the open
13:35
field test where we
13:36
measure how much time the animal spends clinging to the
13:39
perimeter walls versus
13:40
SPEAKER_00: venturing into the exposed center.
13:42
SPEAKER_01: Okay,
13:43
let's unpack this for a second because I want to push back on
13:45
these approach
13:45
SPEAKER_01: avoidance models.
13:47
The literature explicitly notes that these specific mazes were
13:50
primarily developed
13:51
SPEAKER_01: to validate the effects of benzodiazepines,
13:53
SPEAKER_01: like Valium or Xanax.
13:56
If our primary tool for measuring anxiety was built entirely
13:58
around depicting one
14:00
SPEAKER_01: specific class of drugs,
14:02
SPEAKER_01: aren't we just trapped in a massive confirmation bias loop?
14:05
SPEAKER_00: It's one of the most glaring translational failures in our field.
14:08
SPEAKER_00: You've hit the nail on the head.
14:10
These tests are exquisitely sensitive to the gabergic
14:13
mechanisms of benzodiazepines.
14:15
If you give a rat a benzo,
14:16
it will happily stroll out onto the open arms of the plus maze.
14:20
But consider SSRI selective serotonin reuptake inhibitors like
14:24
fluoxetine or sertraline.
14:26
which are the absolute gold standard first line pharmacological
14:29
treatments for
14:30
human generalized anxiety disorder right but if you give an
14:33
SSRI to a rat and put
14:35
it in the elevated plus maze it often fails the SSRI does not
14:38
consistently increase
14:39
SPEAKER_00: time spent in the open arms
14:41
If an alien scientist landed on Earth and only used the
14:44
elevated plus maze to
14:45
SPEAKER_00: screen drugs,
14:45
SPEAKER_00: they would literally throw away every SSRI,
14:48
completely missing the most effective anti-anxiety medications
14:51
we possess.
14:52
SPEAKER_01: Because the thermometer doesn't measure wind speed.
14:54
SPEAKER_01: We're using the wrong biological ruler.
14:56
SPEAKER_01: Yeah.
14:56
SPEAKER_01: So how do we bypass this?
14:58
SPEAKER_01: How do we test anxiety without relying on a test biased toward benzos?
15:02
SPEAKER_00: We look to different behavioral outputs.
15:04
SPEAKER_00: The marble burying test is one.
15:07
You place a mouse in a cage filled with deep bedding and two
15:09
dozen glass marbles.
15:11
Mice will naturally engage in defensive burying of novel,
15:15
unfamiliar objects.
15:16
SPEAKER_01: So an anxious mouse will frantically bury every single marble,
15:20
SPEAKER_01: whereas a calm mouse will just ignore them.
15:22
SPEAKER_00: That's the theory.
15:23
SPEAKER_00: But the construct validity there is hotly debated.
15:26
SPEAKER_00: Are we measuring anxiety toward the object?
15:29
SPEAKER_00: Or are we measuring a repetitive compulsive phenotype?
15:33
Some argue that drugs reducing marble bearing are actually just
15:37
increasing passive
15:38
SPEAKER_00: freezing behavior,
15:39
SPEAKER_00: which is really just a different flavor of anxiety.
15:42
SPEAKER_01: A better alternative seems to be the shock probe bearing test.
15:45
SPEAKER_00: Much better.
15:46
SPEAKER_00: You introduce a small electrified probe into the cage.
15:50
SPEAKER_00: The rat approaches, sniffs it, receives a mild shock, and retreats.
15:54
Then the rat will actively use its forepaws to kick bedding
15:58
over the probe to bury it.
15:59
This is a highly specific,
16:00
active defensive response to a confirmed localized threat.
16:03
SPEAKER_00: It's much cleaner than marble burying.
16:05
SPEAKER_01: But there's another confound we haven't mentioned, which is sedation.
16:09
SPEAKER_01: Benzodiazepines make you sleepy.
16:18
SPEAKER_00: Locomotor confounds ruin a lot of data.
16:21
SPEAKER_00: To bypass behavior entirely, we use stress-induced hyperthermia.
16:26
SPEAKER_01: Taking their temperature.
16:27
SPEAKER_00: Exactly.
16:28
SPEAKER_00: When a mammal experiences acute stress,
16:30
the sympathetic nervous system triggers a slight but measurable
16:33
increase in core
16:34
SPEAKER_00: body temperature.
16:35
Often just the minor stress of the researcher picking the rat
16:38
up to insert a rectal
16:39
SPEAKER_00: thermometer is enough to trigger the response.
16:42
We can measure this physiological spike independently of
16:45
whether the rat is running
16:46
SPEAKER_00: around or sedated in a corner.
16:48
SPEAKER_00: It's a pure autonomic readout.
16:49
SPEAKER_01: I want to scale up from rodents for a moment.
16:52
SPEAKER_01: Social anxiety is profoundly complex.
16:54
SPEAKER_01: It's the fear of judgment, of observation.
16:57
SPEAKER_01: How do you model that?
16:58
We have to move up the evolutionary ladder to non-human
17:01
primates like macaques or
17:03
SPEAKER_00: marmosets because their social hierarchies are much closer to ours.
17:07
SPEAKER_00: The classic model is the human intruder test.
17:09
SPEAKER_01: It sounds like a home invasion.
17:10
SPEAKER_00: It's remarkably simple, actually.
17:12
SPEAKER_00: You have a primate in its enclosure.
17:15
A human researcher who the primate has never seen before steps
17:18
into the room and
17:19
SPEAKER_00: stands perfectly still,
17:21
SPEAKER_00: maintaining a specific profile,
17:23
SPEAKER_00: purposefully avoiding direct eye contact.
17:26
just looming in their peripheral vision and for a primate that
17:29
ambiguous presence
17:30
triggers profound behavioral inhibition they freeze but more
17:34
tellingly they begin
17:36
exhibiting displacement behaviors they'll start scratching
17:39
themselves obsessively
17:40
or engaging in frantic auto grooming it's the primate
17:44
equivalent of a socially
17:45
anxious human nervously biting their nails or checking their
17:48
phone at a party it
17:50
SPEAKER_00: responds beautifully to clinical anxiolytics
17:53
SPEAKER_01: But the human intruder introduces noise.
17:55
SPEAKER_01: I mean, I might be taller than you or wear a different color lab coat.
17:58
To remove that experimenter variable,
18:00
there's a newer touchscreen paradigm.
18:02
SPEAKER_00: Yes.
18:03
SPEAKER_00: This leverages their ethology much better.
18:05
The monkey is trained to touch specific shapes on a computer
18:08
screen to run a juice reward.
18:09
SPEAKER_00: It's a calm, routine task.
18:11
SPEAKER_00: But randomly,
18:11
the screen will flash a high-definition image of an unknown
18:14
dominant male monkey of
18:16
SPEAKER_00: the same species staring directly at them.
18:18
SPEAKER_01: Which is a severe social threat challenge in primate language.
18:22
SPEAKER_00: The reaction is instantaneous.
18:24
SPEAKER_00: You measure their behavioral inhibition.
18:25
SPEAKER_00: Do they physically recoil from the screen?
18:28
How long does it take them to overcome the fear and touch the
18:30
screen again to get the juice?
18:32
It directly taps into the neurobiology of social dominance and
18:36
submission without
18:37
SPEAKER_00: the noise of a human experimenter in the room.
18:39
SPEAKER_01: It's incredible how precise we can get.
18:42
SPEAKER_01: Let's pivot slightly within negative valence,
18:44
moving from the hyperarousal of anxiety to the hyperarousal of
18:47
depression.
18:48
SPEAKER_01: We're talking about hopelessness, despair, reduced effort.
18:52
SPEAKER_01: How do you quantify despair in a mouse?
18:55
The most ubiquitous acute screens are the forced swim test for
18:59
rats and the tail
19:00
SPEAKER_00: suspension test for mice.
19:01
The forced swim test involves placing a rodent in a sheer
19:04
walled cylinder of water.
19:06
SPEAKER_00: They can't touch the bottom and they can't climb out.
19:08
SPEAKER_01: So it's an inescapable stressor.
19:10
SPEAKER_00: Right.
19:11
Initially,
19:11
the animal swims frantically,
19:13
desperately trying to find an escape route.
19:15
SPEAKER_00: But after a few minutes, it stops struggling.
19:17
SPEAKER_00: It adopts a posture of immobility,
19:19
making only the absolute minimal movements required to keep its
19:23
nose above the
19:23
SPEAKER_00: waterline.
19:24
SPEAKER_01: And historically,
19:25
we've interpreted that transition from active struggling to
19:28
passive floating as
19:30
SPEAKER_01: behavioral despair,
19:32
SPEAKER_01: learned helplessness.
19:34
The animal essentially decides,
19:35
you know,
19:36
nothing I do matters,
19:37
so I give up.
19:38
SPEAKER_00: And the predictive validity is why this test is so famous.
19:42
SPEAKER_00: If you administer a clinically effective antidepressant like an SSRI or a
19:46
SPEAKER_00: tricyclic,
19:47
the animal will spend significantly more time actively swimming
19:51
and trying to climb
19:51
SPEAKER_00: before it finally gives up.
19:53
But let's look at the construct validity because there is a
19:56
huge debate here.
19:57
SPEAKER_01: Is the rat really despairing?
19:59
SPEAKER_01: Or is it just exhibiting high-level cognitive adaptation?
20:02
SPEAKER_01: I mean,
20:03
SPEAKER_01: if I'm thrown into the middle of the ocean,
20:04
SPEAKER_01: thrashing around is going to drain my energy and drown me faster.
20:07
SPEAKER_01: Floating conserves energy.
20:08
SPEAKER_01: Maybe the rat isn't depressed.
20:10
SPEAKER_01: Maybe it's just smart.
20:11
SPEAKER_00: That critique is heavily cited.
20:13
Are we measuring despair or are we measuring adaptive energy
20:16
conservation?
20:17
SPEAKER_00: But regardless of the philosophical interpretation of the behavior,
20:20
SPEAKER_00: the pharmacological reality remains.
20:23
SPEAKER_00: The test reliably detects drugs that augment monoamine transmission.
20:28
SPEAKER_00: What's fascinating is how we use it today.
20:30
SPEAKER_00: Standard SSRIs require weeks of chronic dosing to work in humans,
20:34
SPEAKER_00: but they work acutely in the forced swim test.
20:37
SPEAKER_00: Now,
20:37
researchers are using this exact same test to map the
20:39
mechanisms of rapid acting
20:41
SPEAKER_00: agents like ketamine and psilocybin.
20:43
SPEAKER_01: Proving that a single dose of ketamine,
20:45
SPEAKER_01: acting on glutamate rather than serotonin,
20:47
SPEAKER_01: can induce an immediate,
20:48
SPEAKER_01: resilient,
20:49
SPEAKER_01: active coping strategy in the animal that lasts for days.
20:51
SPEAKER_00: Exactly.
20:52
SPEAKER_00: But despair is only one half of the depressive phenotype.
20:56
SPEAKER_00: We also have to measure anhedonia,
20:57
SPEAKER_00: the profound loss of the capacity to feel joy or seek reward.
21:01
SPEAKER_01: We measure that by looking at what the animal usually loves.
21:04
SPEAKER_01: The sucrose preference test is the gold standard.
21:07
SPEAKER_01: Rodents have a massive sweet tooth.
21:09
SPEAKER_01: If you put two bottles in their cage,
21:11
SPEAKER_01: one with plain water,
21:12
SPEAKER_01: one with sugar water,
21:14
SPEAKER_01: they will almost exclusively drink the sugar water.
21:16
But when they are in a depressive-like state,
21:19
that preference completely collapses.
21:21
SPEAKER_01: They'll drink from both bottles equally or barely drink at all.
21:24
SPEAKER_01: They just don't care anymore.
21:26
The progressive ratio test takes that a step further by
21:29
measuring not just
21:30
SPEAKER_00: preference,
21:30
SPEAKER_00: but motivation.
21:32
SPEAKER_00: How hard will you work for the reward?
21:34
SPEAKER_00: The rat presses a lever to get a sugar pellet.
21:36
SPEAKER_00: The first pellet costs one press, the next costs two.
21:40
SPEAKER_00: Then five, ten, twenty, fifty.
21:42
SPEAKER_01: So we're looking for the breakpoint.
21:43
The exact moment the neurobiological cost benefit analysis
21:47
shifts and the animal
21:48
SPEAKER_00: decides the physical effort is no longer worth the dopamine hit.
21:52
In anteedonic animals that breakpoint plummets,
21:54
they give up much earlier.
21:56
SPEAKER_01: But how do we get the animal to that anteedonic state in the first place?
22:00
Acute stressors like the forced swim test don't really capture
22:03
the chronic grinding
22:04
SPEAKER_01: daily friction that usually precipitates clinical depression in humans.
22:08
SPEAKER_00: For that, we rely on chronic stress models.
22:11
SPEAKER_00: The most robust is the chronic social defeat model.
22:14
SPEAKER_00: This leverages the intense territoriality of male mice.
22:18
You take a subject mouse and you place them into the home cage
22:21
of a much larger,
22:22
SPEAKER_00: highly aggressive,
22:23
SPEAKER_00: dominant male,
22:24
SPEAKER_00: usually a specifically bred aggressive strain.
22:27
So the subject mouse is immediately attacked and physically
22:30
subordinated?
22:31
SPEAKER_00: Yes.
22:32
SPEAKER_00: The attack is stopped before serious injury occurs,
22:34
but then the subject mouse is forced to live in that same cage
22:37
for 24 hours,
22:38
SPEAKER_00: separated only by a perforated plexiglass divider.
22:42
He can't be bitten,
22:43
but he can smell,
22:44
see,
22:45
and hear the aggressor constantly.
22:47
SPEAKER_00: It is an environment of inescapable chronic psychosocial stress.
22:51
SPEAKER_00: You repeat this with a new aggressor every day for 10 days.
22:54
SPEAKER_01: And the behavioral changes in the subject mouse are just catastrophic.
22:57
SPEAKER_00: They exhibit profound social avoidance.
23:00
SPEAKER_00: If you put a completely novel,
23:02
SPEAKER_00: non-aggressive mouse in their arena,
23:04
SPEAKER_00: the defeated mouse will just cower in the far corner.
23:07
SPEAKER_00: Their sucrose preference drops.
23:08
SPEAKER_00: Their progressive ratio breakpoints plummet.
23:11
SPEAKER_00: They are neurobiologically depressed.
23:13
here's where it gets really interesting and honestly represents
23:16
one of the most
23:16
frustrating historical blind spots in neuroscience for decades
23:20
this incredibly
23:21
powerful robust model for depression was fundamentally
23:25
mathematically flawed
23:27
because it only worked on males the territorial aggression
23:30
required to make the
23:31
model work is a male specific ethological trait female mice
23:36
generally speaking do
23:37
SPEAKER_00: not display or elicit that type of fierce
23:41
SPEAKER_00: Which is mind-boggling when you look at the epidemiology.
23:52
SPEAKER_01: Women experience clinical depression at roughly twice the rate of men.
23:55
We were studying the neurobiology of a disease predominantly
23:58
affecting women by
23:59
SPEAKER_01: exclusively looking at male brains.
24:01
And we know now that the molecular reality of depression is
24:04
sexually dimorphic.
24:05
SPEAKER_00: The transcriptomic profiles,
24:06
the actual genes being turned on and off in the brain during
24:09
chronic stress,
24:10
SPEAKER_00: are vastly different between male and female mice.
24:13
By excluding females,
24:14
we were literally missing half the biological picture.
24:17
SPEAKER_01: Thankfully,
24:18
the literature outlines some brilliant ethological workarounds
24:21
that have recently
24:22
SPEAKER_01: solved this.
24:23
SPEAKER_00: It requires tricking the social circuitry?
24:26
SPEAKER_00: One method involves applying male odorants,
24:28
urine or specific pheromones to the female subject before
24:32
placing her in the
24:32
SPEAKER_00: aggressor's cage.
24:34
SPEAKER_00: The male's olfactory system registers a male rival and he attacks.
24:38
But there's a more organic adaptation involving maternal
24:41
aggression,
24:41
right?
24:41
SPEAKER_00: Yes.
24:42
SPEAKER_00: You house a female aggressor with a castrated male.
24:45
Because the male is castrated,
24:46
they don't breed,
24:47
but they form a tight pair bond.
24:49
The female establishes that cage as her territory and views the
24:53
castrated male as her mate.
24:55
SPEAKER_00: If you then drop a female intruder into that cage,
24:58
SPEAKER_00: the resident female's maternal territory defense circuits fire rapidly.
25:03
SPEAKER_00: She will fiercely attack the female intruder.
25:05
SPEAKER_01: Finally,
25:06
allowing us to subject female mice to chronic social defeat and
25:09
opening up an
25:10
entirely new frontier of studying sex-specific neurobiology and
25:13
depression.
25:14
It's a perfect example of how the design of the model dictates
25:17
the absolute limits
25:18
SPEAKER_01: of our knowledge.
25:19
Let's cross the spectrum now,
25:20
moving from negative valence to positive valence.
25:23
Positive valence covers emotional states that are intrinsically
25:26
attractive or rewarding.
25:28
SPEAKER_00: In a healthy brain, this system drives us to eat, mate, and survive.
25:33
SPEAKER_00: But in psychiatric disorders,
25:34
we are usually looking at what happens when this
25:37
dopamine-driven system becomes
25:39
SPEAKER_00: completely unmoored,
25:40
SPEAKER_00: leading to destructive hyperarousal or addiction.
25:43
SPEAKER_01: The ultimate expression of hyperarousal is mania in bipolar disorder.
25:48
SPEAKER_01: And again, this seems impossible to model.
25:50
I can't imagine a rat experiencing grandiosity or maxing out a
25:54
credit card or
25:55
SPEAKER_01: talking a mile a minute.
25:57
SPEAKER_00: We can't model the cognitive grandiosity,
25:59
SPEAKER_00: but we can model the biological triggers and the behavioral outcomes.
26:03
SPEAKER_00: If you look at human bipolar patients,
26:05
severe sleep disturbance is one of the most reliable triggers
26:07
for a manic episode.
26:09
SPEAKER_00: So researchers leverage that using a sleep deprivation paradigm.
26:12
SPEAKER_01: This is the flowerpot technique.
26:14
SPEAKER_00: It's conceptually simple.
26:15
You place a rodent on a very small platform like an inverted
26:18
flowerpot surrounded
26:19
SPEAKER_00: by a pool of water.
26:21
SPEAKER_00: The platform is just big enough for the rat to sit, eat and groom.
26:24
But when a mammal enters REM sleep,
26:26
they experience atonia,
26:27
complete muscle relaxation.
26:29
So the moment the rat hits REM sleep,
26:31
it goes limp,
26:32
its nose hits the water,
26:33
and it wakes up.
26:34
SPEAKER_00: It completely deprives them of REM sleep.
26:37
SPEAKER_00: If you enforce this for 72 hours, the animal's neurobiology snaps.
26:41
SPEAKER_00: It enters a state that remarkably mimics an acute manic episode.
26:45
SPEAKER_00: The rat becomes hyperactive, exhibiting stereotypic repetitive pacing.
26:49
SPEAKER_00: Even when you take them off the platform and allow them to sleep,
26:52
SPEAKER_00: they exhibit severe insomnia,
26:54
SPEAKER_00: they become highly aggressive,
26:55
SPEAKER_00: and they exhibit profound hypersexuality.
26:58
All classic hallmarks of human mania and the predictive
27:01
validity is the clincher
27:02
here it is if you treat these sleep deprived rats chronically
27:06
with mood stabilizers
27:07
specifically lithium these manic behaviors are almost entirely
27:11
blocked the model
27:12
proves that the biological link between sleep architecture and
27:14
mood regulation is
27:16
SPEAKER_00: conserved across species
27:18
SPEAKER_01: Let's drill down into impulsivity.
27:19
SPEAKER_01: The literature divides this into impulsive action and impulsive choice.
27:23
SPEAKER_01: And clinically, it's vital to separate the two.
27:26
SPEAKER_01: Impulsive action is essentially a failure of the brakes.
27:29
SPEAKER_01: You know you shouldn't do something,
27:31
but the prepotent motor response fires before the prefrontal
27:33
cortex can veto it.
27:34
We measure impulsive action using the stop signal reaction time
27:38
task,
27:38
the SSRT.
27:40
SPEAKER_00: It's a beautifully precise test of inhibitory control.
27:43
The rat is trained to press a lever quickly when a go signal
27:46
flashes to get a reward.
27:48
But on a small fraction of the trials,
27:50
a no-go signal flashes milliseconds after the go signal.
27:54
SPEAKER_00: The rat has to abruptly cancel its initiated motor response.
27:58
It's like swinging a baseball bat and trying to stop mid-swing
28:01
when you realize
28:01
SPEAKER_01: it's a curveball.
28:02
SPEAKER_00: Exactly.
28:03
We calculate how long of a delay between the two signals the
28:06
rat can tolerate and
28:07
SPEAKER_00: still successfully hit the brakes.
28:09
We also use the five choice serial reaction time task,
28:12
which tests sustained spatial attention.
28:15
SPEAKER_00: The rat is facing an array of five holes.
28:17
It has to wait for a light to flash in one of the holes,
28:20
then poke its nose in to get a treat.
28:22
SPEAKER_00: If it gets impatient and pokes its nose in before light flashes,
28:25
SPEAKER_00: that's a premature response,
28:27
SPEAKER_00: a failure of impulsive action.
28:29
SPEAKER_01: And drugs like adamoxetine or methylphenidate,
28:31
SPEAKER_01: which we use for ADHD in humans,
28:34
SPEAKER_01: significantly reduce those premature responses in rats.
28:37
SPEAKER_00: Then we have impulsive choice.
28:39
SPEAKER_00: This is cognitive impulsivity.
28:41
SPEAKER_00: It's the classic delayed discounting paradigm.
28:44
SPEAKER_00: Would you rather have a small reward right now or a massive reward later?
28:49
SPEAKER_00: An impulsive brain heavily discounts the value of future rewards.
28:53
SPEAKER_00: It can't wait.
28:54
But the textbook raises a fascinating translational roadblock
28:57
here In human
28:59
SPEAKER_01: studies,
28:59
I can offer you 10 pounds today or 100 pounds next year I can
29:02
just verbally explain
29:03
SPEAKER_01: the future You can't explain the concept of next week to a macaque monkey
29:07
SPEAKER_00: That's a huge point.
29:08
SPEAKER_00: In animal models, the subject must experientially learn the delay.
29:12
SPEAKER_00: They have to press the large reward lever hundreds of times and sit there
29:16
experiencing the 30 second delay over and over until the brain
29:19
maps the temporal
29:20
SPEAKER_00: cost.
29:21
SPEAKER_00: So when we compare rodent data to human data, we have to be careful.
29:25
We can only accurately compare it to human tasks where the
29:28
human subjects also had
29:29
to sit in a room and actually experience the delay in real time
29:32
rather than just
29:33
SPEAKER_00: answering a hypothetical questionnaire about the future.
29:35
Let's link impulsivity to compulsion because on the ward they
29:39
can look identical.
29:40
SPEAKER_01: A patient keeps gambling or a patient keeps washing their hands.
29:43
SPEAKER_01: But the mechanism is totally different.
29:45
SPEAKER_01: Impulsivity is acting without thinking.
29:47
Compulsion is feeling physically forced to act despite knowing
29:51
the action is senseless.
29:52
SPEAKER_01: How do we isolate compulsion in a lab?
29:55
We look for perseverative responding,
29:57
the inability to break a cognitive loop.
30:00
SPEAKER_00: The reversal learning paradigm isolates this perfectly.
30:03
SPEAKER_00: You train an animal on a simple rule.
30:05
SPEAKER_00: The left lever gives food.
30:07
SPEAKER_00: The right lever does nothing.
30:08
SPEAKER_00: They learn it perfectly.
30:09
SPEAKER_00: Then, without any warning, you reverse the contingencies.
30:13
SPEAKER_00: Now the right lever gives food.
30:15
An impulsive animal might just start mashing both levers
30:17
randomly in frustration.
30:19
SPEAKER_00: But the compulsive animal will exhibit perseveration.
30:21
SPEAKER_00: It will continue to press the left lever 50,
30:24
SPEAKER_00: 100 times,
30:24
completely ignoring the overwhelming empirical evidence that
30:27
it's no longer
30:28
SPEAKER_00: working.
30:29
The behavioral flexibility circuit in the prefrontal cortex is
30:32
completely jammed.
30:33
SPEAKER_01: We also see this rigid compulsion in eating disorders.
30:37
SPEAKER_01: The textbook details activity-based anorexia,
30:40
which has to be one of the most deeply counterintuitive models
30:43
in all of science.
30:45
SPEAKER_00: It defies basic survival logic.
30:47
SPEAKER_00: If you restrict a mammal's food intake,
30:49
SPEAKER_00: its evolutionary programming should force it to conserve energy,
30:52
SPEAKER_00: to sleep,
30:53
SPEAKER_00: to lower its metabolic rate,
30:56
SPEAKER_00: In the activity-based anorexia model,
30:58
SPEAKER_00: you restrict a rat's access to food to only a short window each day,
31:01
SPEAKER_00: usually just an hour or two.
31:03
SPEAKER_00: But you also give them free access to a running wheel.
31:05
SPEAKER_01: And paradoxically, instead of resting, they start running.
31:08
SPEAKER_00: They run excessively.
31:09
SPEAKER_00: And the more they run, the less they eat during their designated window.
31:13
SPEAKER_00: It creates a lethal neurobiological feedback loop.
31:16
The starvation triggers hyperactivity,
31:18
which suppresses appetite,
31:19
causing more starvation.
31:20
If researchers don't intervene,
31:22
the rats will literally run themselves to death.
31:24
it's a harrowing mirror of the lethal drive we see in severe
31:28
human anorexia nervosa
31:30
where hyperactivity and severe caloric restriction lock
31:33
together but because we
31:35
don't have any genuinely effective targeted pharmacotherapies
31:38
for anorexia yet we
31:40
can't fully validate the predictive validity of the model on
31:44
the other side of the
31:44
eating disorder spectrum we model binge eating using the
31:47
cafeteria diet if you give
31:49
SPEAKER_00: rodents access to
31:51
SPEAKER_00: Highly palatable,
31:52
SPEAKER_00: high fat,
31:53
SPEAKER_00: high carbohydrate human foods,
31:55
SPEAKER_00: essentially junk food,
31:57
SPEAKER_00: and restrict their access to it,
31:58
SPEAKER_00: they will gorge.
32:00
The vital finding here is that chronic exposure to this
32:02
cafeteria diet causes a
32:04
SPEAKER_00: severe downregulation of the dopamine receptors in the striatum.
32:07
SPEAKER_01: Which is the exact same neurological signature we see in drug addiction.
32:11
The brain gets flooded with so much dopamine from the junk food
32:14
that it removes
32:15
SPEAKER_01: receptors to protect itself,
32:16
SPEAKER_01: meaning the animal has to eat even more next time just to feel normal.
32:20
Which brings us right to the door of substance use disorders,
32:22
the hijacked brain.
32:24
The modern clinical framework views substance use disorder
32:27
entirely as an acquired
32:28
SPEAKER_00: disease of neuroplasticity.
32:30
SPEAKER_00: The addictive substance fundamentally rewires the reward circuitry,
32:35
causing the brain to radically overvalue the drug at the total
32:38
expense of natural
32:40
SPEAKER_00: rewards like food,
32:41
SPEAKER_00: water or social connection.
32:42
The easiest way to measure drug reward is condition place
32:45
preference,
32:46
or CPP.
32:47
SPEAKER_01: You put a mouse in a box with two distinct chambers.
32:50
SPEAKER_01: One has striped walls, one has spotted walls.
32:53
SPEAKER_01: Day one, you lock them in the striped room and inject them with cocaine.
32:56
SPEAKER_01: Day two, you lock them in the spotted room and inject them with saline.
32:59
After a few days of this,
33:00
you open the partition and let them choose where to hang out.
33:03
SPEAKER_00: And overwhelmingly, they will choose the striped room.
33:06
The brain has associated those specific spatial cues with the
33:10
massive dopamine
33:11
SPEAKER_00: spike of the drug.
33:12
But the text is quite blunt about CPP having limited construct
33:16
validity for human addiction.
33:17
SPEAKER_01: Why?
33:18
SPEAKER_00: Because it's a totally passive model.
33:20
SPEAKER_00: The experimenter is injecting the drug.
33:22
SPEAKER_00: The animal is just a recipient.
33:25
But human addiction is defined by active,
33:27
desperate,
33:28
operant drug-seeking behavior.
33:30
SPEAKER_00: The patient is doing the work to acquire the drug.
33:32
SPEAKER_01: So we move to active self-administration.
33:34
SPEAKER_00: Yes.
33:36
SPEAKER_00: We surgically implant intravenous catheters into rodents or primates.
33:40
We place them in operant chambers and train them that pressing
33:42
a lever delivers a
33:43
SPEAKER_00: micro infusion of a drug directly into their bloodstream.
33:46
SPEAKER_01: And the translation here is stark.
33:48
Animals will readily,
33:50
eagerly self administer almost every single drug that humans
33:53
abuse.
33:54
SPEAKER_01: Cocaine, heroin, methamphetamine, nicotine, alcohol.
33:58
SPEAKER_00: Interestingly,
33:59
they will not self-administer classic hallucinogens like LSD or
34:02
psilocybin,
34:03
which maps perfectly onto the fact that those drugs do not
34:06
drive compulsive
34:07
SPEAKER_00: addiction in humans.
34:08
SPEAKER_01: But pressing a lever for a hit of cocaine doesn't equal addiction.
34:11
SPEAKER_01: Lots of humans try cocaine and don't become addicted.
34:14
We have to model the clinical reality of the disease,
34:16
which is the loss of control.
34:18
SPEAKER_01: The progressive ratio test comes back into play here.
34:21
SPEAKER_01: We measure the breakpoint.
34:22
SPEAKER_01: Will the rat press the lever 100 times for a single infusion?
34:25
SPEAKER_01: 500 times.
34:25
SPEAKER_01: 1000 times.
34:28
SPEAKER_00: It measures the escalating motivational value of the drug.
34:31
SPEAKER_00: But the most clinically relevant model,
34:34
the one with the highest construct validity for the sheer
34:36
destruction of addiction,
34:38
SPEAKER_00: is modeling compulsive use despite severe negative consequences.
34:41
SPEAKER_01: This is the foot shock paradigm.
34:43
SPEAKER_00: Yes.
34:44
Once an animal has established a strong self-administration
34:47
habit,
34:47
you change the rules.
34:49
SPEAKER_00: Now,
34:49
SPEAKER_00: pressing the lever delivers the hit of cocaine,
34:51
SPEAKER_00: but it simultaneously delivers a painful electrical shock to their feet.
34:55
A healthy animal or one that has only had brief exposure to the
34:58
drug will stop
34:59
immediately the neurobiological logic says the pain isn't worth
35:03
the reward exactly
35:05
SPEAKER_00: But an addicted animal,
35:07
one whose prefrontal cortex has been downregulated by chronic
35:10
extended drug access,
35:11
SPEAKER_00: cannot stop.
35:13
SPEAKER_00: The compulsive drive overrides the pain signals.
35:15
SPEAKER_00: They will continue to press the lever,
35:17
SPEAKER_00: enduring severe repeated shocks,
35:19
SPEAKER_00: just to maintain their drug levels.
35:21
It perfectly models the human patient who continues to inject
35:24
heroin despite losing
35:26
SPEAKER_00: their marriage,
35:26
SPEAKER_00: their career,
35:27
SPEAKER_00: their health,
35:28
SPEAKER_00: and their freedom.
35:29
SPEAKER_01: And the tragedy of addiction doesn't stop when the drug stops.
35:32
SPEAKER_01: We have to model relapse, extinction, and reinstatement.
35:36
SPEAKER_00: You force the animal into abstinence by removing the drug.
35:39
SPEAKER_00: The lever stops working.
35:40
SPEAKER_00: Eventually, the lever pressing extinguishes.
35:43
SPEAKER_00: Then we study what triggers reinstatement,
35:45
what causes them to relapse and start frantically pressing the
35:48
lever again.
35:48
SPEAKER_01: And the triggers are identical to human triggers.
35:51
SPEAKER_01: A stressor, like a random foot shock, will trigger relapse.
35:55
SPEAKER_01: A small priming dose of the drug given by the researcher.
35:58
SPEAKER_01: Or, most powerfully, just the drug paired cue.
36:02
Just flashing the light that used to turn on when the drug was
36:04
delivered is enough
36:05
SPEAKER_01: to send the animal into a frenzy of drug seeking.
36:08
And this leads to perhaps the most terrifying finding in
36:11
addiction neuroscience,
36:12
SPEAKER_00: the incubation of craving.
36:14
You would logically think that the longer someone is abstinent,
36:17
the weaker their craving gets.
36:19
SPEAKER_01: Time heals all wounds.
36:20
SPEAKER_00: But the neuroplasticity of addiction works in reverse.
36:24
If you measure how hard an animal will work for the drug after
36:27
one day of abstinence,
36:28
it's high.
36:29
But if you force them into abstinence for 30 days or 60 days
36:33
and then expose them
36:34
SPEAKER_00: to a trigger,
36:35
SPEAKER_00: the magnitude of their relapse,
36:36
SPEAKER_00: the sheer violence of their lever pressing,
36:39
SPEAKER_00: is exponentially higher.
36:41
SPEAKER_00: The craving actually intubates and amplifies over time.
36:44
SPEAKER_01: Which explains why a patient could be clean for a year,
36:47
SPEAKER_01: walk past the pub they used to drink at,
36:49
and suddenly experience a craving so intense it feels like a
36:52
physical blow.
36:54
SPEAKER_01: The brain was quietly sharpening that knife in the background.
36:57
SPEAKER_01: Let's shift gears now into the third broad domain.
37:00
SPEAKER_01: Cognition, social functioning, and sleep.
37:03
Cognitive deficits are often the most debilitating
37:05
treatment-resistant symptoms in
37:06
SPEAKER_00: psychiatry,
37:07
SPEAKER_00: especially in schizophrenia.
37:09
To understand cognition,
37:10
we have to start at the very bottom before a thought even
37:13
forms.
37:13
SPEAKER_00: Sensory gating.
37:15
SPEAKER_00: Pre-attentive processing.
37:16
SPEAKER_01: Your brain is constantly bombarded with sensory data.
37:19
SPEAKER_01: The hum of the fridge, the feeling of your socks.
37:21
SPEAKER_01: If your brain processed all of it at full volume, you'd go mad.
37:24
SPEAKER_01: So it filters it out.
37:25
SPEAKER_00: We measure this filtering using pre-pulse inhibition, or PPI.
37:29
SPEAKER_00: It relies on the acoustic startle reflex.
37:31
SPEAKER_00: If I sneak up behind you and detonate a firecracker,
37:34
SPEAKER_00: you will exhibit a massive full-body startle response.
37:38
But if I tap you firmly on the shoulder a fraction of a second
37:42
before I detonate
37:43
SPEAKER_00: it,
37:43
SPEAKER_00: your brain registers the tap,
37:45
SPEAKER_00: the pre-pulse.
37:46
It uses that minor stimulus to warn the system and actively
37:49
dampen the startle
37:50
SPEAKER_00: response to the subsequent loud noise.
37:52
SPEAKER_00: You'll still jump, but it will be heavily inhibited.
37:55
SPEAKER_01: In a healthy brain, that pre-pulse inhibition is intact.
37:58
SPEAKER_01: But in patients with schizophrenia,
38:00
SPEAKER_01: particularly those actively experiencing psychosis,
38:03
SPEAKER_01: this early sensory filtering is fundamentally broken.
38:07
SPEAKER_00: The prepulse does nothing.
38:08
SPEAKER_00: The loud noise is processed at maximum intensity.
38:11
SPEAKER_00: The brain is utterly overwhelmed by incoming stimuli,
38:14
SPEAKER_00: which many theorists believe is the foundational mechanism that generates
38:17
SPEAKER_00: delusions.
38:18
The brain is desperately trying to make sense of a tsunami of
38:21
unfiltered sensory data.
38:23
SPEAKER_01: And PPI is completely conserved across species.
38:26
SPEAKER_01: We can put a rat in a soundproof box,
38:28
SPEAKER_01: play a quiet tone,
38:29
SPEAKER_01: then a loud tone,
38:30
SPEAKER_01: and measure the flinch.
38:32
SPEAKER_01: If we give them an antipsychotic,
38:34
SPEAKER_01: like haloperidol or clozapine,
38:36
SPEAKER_01: we can watch it restore their pre-pulse inhibition in real time.
38:40
SPEAKER_00: Moving up from sensory gating,
38:41
SPEAKER_00: we look at working memory,
38:43
SPEAKER_00: the ability to hold and manipulate information transiently.
38:46
SPEAKER_00: This is heavily reliant on the prefrontal cortex.
38:49
SPEAKER_00: In rodents, we use the radial arm maze.
38:52
SPEAKER_00: Imagine a central platform with eight distinct arms radiating outward.
38:56
SPEAKER_00: Food is placed at the end of each arm.
38:58
SPEAKER_00: An efficient rat will visit each arm exactly once.
39:01
It has to hold a real-time map in its head like I've cleared
39:03
arms one three and
39:04
four don't go back there if it reenters an arm it's already
39:07
emptied that is
39:08
quantified as a working memory error but translating working
39:11
memory from rodents to
39:12
humans is fraught rodents are brilliant for early screening but
39:16
evolutionarily our
39:17
prefrontal cortices diverged tens of millions of years ago the
39:20
rodent prefrontal
39:21
SPEAKER_00: architecture lacks the granular complexity of the primate brain
39:25
SPEAKER_01: Which is why primate models are irreplaceable here.
39:28
SPEAKER_01: Monkeys can perform delayed match to sample tasks.
39:31
SPEAKER_01: A screen drops,
39:32
SPEAKER_01: they see an object,
39:33
SPEAKER_01: the screen goes opaque for 20 seconds,
39:35
SPEAKER_01: and then they have to pick the matching object from a complex lineup.
39:38
And because primate prefrontal homology is so tight to humans,
39:41
the data translates flawlessly.
39:44
SPEAKER_00: In fact,
39:44
SPEAKER_00: the Cantab cognitive assessment system,
39:47
SPEAKER_00: which clinicians use in hospitals every day to assess human dementia and
39:50
SPEAKER_00: schizophrenia,
39:51
was directly reverse engineered from these specific primate
39:55
touchscreen tasks.
39:56
SPEAKER_01: Let's talk about declarative memory, spatial memory.
39:59
The textbook brings up the absolute titan of behavioral
40:02
neuroscience,
40:03
the Morris Water Maze.
40:05
SPEAKER_00: It is the most famous spatial memory test in the world.
40:08
SPEAKER_00: You have a large circular pool filled with opaque water.
40:12
SPEAKER_00: Just beneath the surface,
40:13
SPEAKER_00: completely invisible to the swimming animal,
40:15
SPEAKER_00: is a small resting platform.
40:17
SPEAKER_00: You drop a rodent in.
40:18
It swims frantically until it accidentally bumps into the
40:20
hidden platform and climbs up.
40:22
SPEAKER_00: Over multiple days,
40:23
the rodent uses visual cues on the walls of the laboratory to
40:26
triangulate the exact
40:27
SPEAKER_00: spatial location of that hidden platform.
40:29
You test their memory by taking the platform out and watching
40:33
to see if they swim
40:34
SPEAKER_01: in tight focus circles exactly where the platform used to be.
40:37
SPEAKER_01: It's incredibly elegant.
40:39
SPEAKER_01: But I want to push back on this.
40:40
The water maze is universally used,
40:43
but it has a colossal biological confound.
40:46
You are dropping a rat into a pool of water where it cannot
40:50
touch the bottom.
40:50
SPEAKER_01: It thinks it is going to drown.
40:52
SPEAKER_01: It is swimming for its life.
40:53
SPEAKER_01: Doesn't that acute survival panic completely spike their cortisol levels?
40:57
Are we actually measuring spatial memory or are we measuring
41:01
how severe stress
41:02
SPEAKER_01: degrades hippocampal function?
41:04
SPEAKER_00: That is a phenomenal critique,
41:06
and it is the exact reason many researchers are moving away
41:08
from the water maze.
41:09
SPEAKER_00: The physical exhaustion of swimming,
41:11
SPEAKER_00: the thermal stress of the water temperature,
41:14
SPEAKER_00: the massive sympathetic nervous system activation,
41:16
SPEAKER_00: these are all massive confounding variables.
41:19
The stress of impending drowning can alter the results entirely
41:23
independently of
41:23
SPEAKER_00: the actual memory circuit you are trying to study.
41:25
SPEAKER_01: So what's the alternative?
41:26
SPEAKER_01: How do you test spatial memory without the drowning confound?
41:29
SPEAKER_00: We use the barn's maze.
41:31
It's a large elevated circular table with 20 circular holes
41:35
drilled around the perimeter.
41:36
SPEAKER_00: Underneath 19 of the holes, there's just a drop to the floor.
41:40
SPEAKER_00: But underneath one specific hole is a dark enclosed safe escape box.
41:45
SPEAKER_01: You put the mouse in the brightly lit center of the table.
41:48
SPEAKER_00: Right.
41:49
SPEAKER_00: And think about the ethology of a mouse.
41:51
SPEAKER_00: It is a prey animal.
41:53
Its entire evolutionary biology screams that standing in a
41:56
bright open space means
41:58
SPEAKER_00: a hawk is going to eat it.
41:59
SPEAKER_00: It desperately wants to find a dark enclosed space.
42:04
SPEAKER_00: So it explores the edges until it finds the escape box.
42:07
SPEAKER_01: But crucially, it isn't drowning.
42:09
SPEAKER_01: It's uncomfortable.
42:10
SPEAKER_01: It's motivated by a natural ethological aversion to open spaces.
42:14
SPEAKER_01: But it isn't an acute survival panic.
42:16
SPEAKER_00: Precisely.
42:17
SPEAKER_00: Over several trials using the exact same spatial cues around the room,
42:20
SPEAKER_00: the mouse learns to run in a straight line directly to the correct hole.
42:24
It measures the exact same hippocampal-dependent spatial
42:27
declarative memory as the
42:28
SPEAKER_00: water maze,
42:29
but it completely removes the confounding variable of acute
42:32
survival stress.
42:33
SPEAKER_00: It's a much cleaner, far superior model.
42:36
SPEAKER_01: Let's touch briefly on social functioning.
42:38
SPEAKER_01: Attachment.
42:39
SPEAKER_01: We mentioned primate social anxiety, but what about the positive side?
42:43
SPEAKER_01: Affiliation.
42:44
SPEAKER_01: Love.
42:45
SPEAKER_00: For that, we owe an enormous debt to the prairie vole.
42:49
SPEAKER_00: The vast majority of rodent species are highly periscuous.
42:51
SPEAKER_00: They mate and move on.
42:53
SPEAKER_00: But the prairie vole is uniquely monogamous.
42:56
SPEAKER_00: They form profound, lifelong pair bonds.
42:59
SPEAKER_00: They share parenting duties.
43:00
SPEAKER_00: If a male prairie vole's partner dies,
43:03
SPEAKER_00: he will often refuse to take a new mate,
43:05
SPEAKER_00: exhibiting behaviors that strikingly resemble human grief.
43:08
SPEAKER_01: It's heartbreaking, but scientifically, it's a goldmine.
43:11
SPEAKER_00: It allowed us to decode the neurobiology of attachment.
43:14
SPEAKER_00: By studying prairie voles,
43:15
we discovered the absolute criticality of oxytocin and
43:18
vasopressin in cementing
43:19
SPEAKER_00: social bonds.
43:21
SPEAKER_00: If you block oxytocin receptors in a prairie vole's brain,
43:24
SPEAKER_00: they become just as promiscuous as a standard mouse.
43:27
SPEAKER_00: It proved that attachment isn't just a psychological construct,
43:29
SPEAKER_00: it's a hardwired neurochemical circuit.
43:32
Finally in this domain we have sleep and arousal and this is
43:35
where we have to tell
43:35
the story of narcolepsy because it is arguably the single
43:38
greatest triumph of
43:39
SPEAKER_01: construct validity in the history of our field.
43:42
SPEAKER_00: Human narcolepsy is devastating.
43:45
It's characterized by excessive daytime sleepiness and sudden
43:48
uncontrollable sleep attacks.
43:50
SPEAKER_00: Most notably,
43:51
SPEAKER_00: patients experience cataplexy,
43:53
SPEAKER_00: a sudden,
43:54
SPEAKER_00: complete loss of voluntary muscle tone while they are fully awake,
43:59
SPEAKER_00: usually triggered by a strong,
44:00
SPEAKER_00: positive emotion like laughing.
44:02
SPEAKER_01: For a long time, we had no idea what caused it.
44:05
SPEAKER_01: But in the late 20th century,
44:06
SPEAKER_01: researchers identified a specific colony of dogs,
44:10
SPEAKER_01: primarily Doberman,
44:10
SPEAKER_01: Pinschers,
44:11
SPEAKER_01: and Labradors,
44:12
SPEAKER_01: that exhibited naturally occurring narcolepsy.
44:14
SPEAKER_00: It was startling to watch.
44:15
SPEAKER_00: The dogs would be playing,
44:17
SPEAKER_00: or about to eat a treat,
44:18
SPEAKER_00: experiencing high positive emotion,
44:20
SPEAKER_00: and they would suddenly collapse to the floor in a cataplectic fit,
44:23
SPEAKER_00: fully conscious but paralyzed.
44:26
SPEAKER_00: Researchers developed the food-elicited cataplexy test to quantify it.
44:29
SPEAKER_01: But the real breakthrough was the genetics?
44:31
SPEAKER_00: Yes.
44:32
SPEAKER_00: Because these dogs naturally possessed the disease,
44:35
SPEAKER_00: researchers systematically sequenced their genome to find the flaw.
44:39
SPEAKER_00: And in doing so,
44:40
SPEAKER_00: they discovered an entirely unknown neurobiological system,
44:44
SPEAKER_00: the orexin receptor system,
44:46
SPEAKER_00: also known as Hippocrutin.
44:47
They proved that a specific genetic mutation causing a loss of
44:51
function in the
44:51
SPEAKER_00: orexin system was the root cause of the canine narcolepsy.
44:55
SPEAKER_01: And when they took that canine data back to humans, they struck gold.
44:59
We now know definitively that human narcolepsy type 1 is caused
45:02
by the autoimmune
45:03
SPEAKER_01: destruction of rexin-producing neurons in the hypothalamus.
45:07
SPEAKER_00: It is the ultimate translational triumph,
45:09
discovering the exact biological mechanism in an animal that
45:12
maps flawlessly onto
45:13
SPEAKER_00: the human etiology.
45:14
SPEAKER_00: It led to genetic mouse models,
45:16
and it is directly responsible for the entirely new classes of
45:19
targeted orexin
45:20
SPEAKER_00: antagonist sleep medications we use in the clinic today.
45:23
Which brings us to a critical juncture,
45:25
the integration,
45:27
the clinical reasoning,
45:28
SPEAKER_01: Because the narcolepsy story is beautiful, but it's the exception.
45:31
SPEAKER_01: It is not the rule.
45:33
SPEAKER_01: This leads us to the preclinical bottleneck.
45:35
SPEAKER_01: If our models are as sophisticated as we've just described,
45:38
if we can map fear extinction and working memory and
45:41
impulsivity with such surgical
45:43
SPEAKER_01: precision,
45:44
Why do the vast majority of psychiatric drugs that cure mice
45:47
fail completely when
45:48
SPEAKER_01: they reach human clinical trials?
45:50
SPEAKER_01: This is the crux of the entire field.
45:52
SPEAKER_01: The bottleneck exists because of how we define human disease.
45:56
We diagnose humans using syndromal behavioral clusters,
45:59
the DSM and the ICD.
46:01
We say a patient has major depressive disorder because they
46:03
check five out of nine
46:04
SPEAKER_01: symptom boxes on a clinician's clipboard.
46:06
But the biological reality beneath those checkboxes is vastly,
46:10
wildly heterogeneous.
46:12
SPEAKER_00: Exactly.
46:12
Let's say we have two patients in the waiting room,
46:14
both diagnosed with major depression.
46:17
Patient A's depression is driven by severe neuroinflammation
46:21
cytokine cascades.
46:22
SPEAKER_00: Patient B's depression is driven by chronic stress,
46:25
SPEAKER_00: altering their epigenetic transcription in the hippocampus.
46:28
SPEAKER_00: Clinically, they look identical.
46:29
SPEAKER_00: Now,
46:30
a pharmaceutical company develops a brilliant new drug
46:33
validated in the chronic
46:34
SPEAKER_00: social defeat mouse model.
46:36
the drug perfectly reverses the epigenetic stress adaptation so
46:39
they take that drug
46:40
they bypass the mice and they run a massive phase three
46:44
clinical trial on 1000
46:46
depressed human patients and the trial fails the drug is
46:50
shelved why because the
46:51
SPEAKER_00: drug only works on the underlying biology of patient B
46:54
It does absolutely nothing for patient A whose depression is
46:57
inflammatory.
46:58
But because our clinical trials group them all together under
47:00
the monolithic
47:01
SPEAKER_00: syndrome umbrella of major depression,
47:03
the overall statistical efficacy of the drug looks no better
47:06
than a placebo.
47:07
SPEAKER_01: The syndrome diagnosis is masking the biological precision of the drug.
47:11
SPEAKER_01: The drug didn't fail.
47:12
SPEAKER_01: Our diagnostic manual failed.
47:14
SPEAKER_01: So how do we fix it?
47:15
SPEAKER_00: The future relies on precision medicine and diversity.
47:19
SPEAKER_00: On the precision front,
47:20
tools like CRISPR-Cas9 gene editing allow us to create highly
47:24
specific genetic
47:25
SPEAKER_00: models that map onto actual human biology,
47:28
SPEAKER_00: not just human behavioral checklists.
47:31
RNA sequencing and proteomic screens are refining our construct
47:34
validity to an
47:35
SPEAKER_00: unprecedented degree.
47:37
SPEAKER_01: And on the diversity front, the textbook is adamant.
47:40
We cannot keep using exclusively male,
47:42
genetically identical,
47:44
highly inbred strains of mice.
47:45
SPEAKER_00: It guarantees clinical failure.
47:47
SPEAKER_00: Human populations are wildly heterogeneous.
47:51
SPEAKER_00: If we ignore sex as a fundamental biological variable,
47:54
SPEAKER_00: as we did with the social defeat model for decades,
47:56
we guarantee that our findings will only ever apply to half the
47:59
human population.
48:01
SPEAKER_00: True clinical reasoning requires us to understand individual,
48:04
SPEAKER_00: population-based,
48:05
and species-specific differences Let's pressure test all of
48:08
this I want to throw
48:09
some rapid-fire clinical scenarios at you based on everything
48:11
we've just covered to
48:12
see how these models actually apply to reasoning on the ward or
48:15
in the lab Let's do
48:16
SPEAKER_00: it Focus on the mechanisms Scenario 1
48:19
A researcher develops a new rat model testing a novel addictive
48:23
compound.
48:24
SPEAKER_01: When the drug is abruptly removed,
48:26
SPEAKER_01: the rat exhibits severe,
48:28
SPEAKER_01: quantifiable physical withdrawal symptoms shaking,
48:30
SPEAKER_01: temperature dysregulation,
48:32
SPEAKER_01: mirroring human physical dependence perfectly.
48:35
What specific type of validity is the researcher predominantly
48:38
claiming here?
48:38
SPEAKER_01: Pause the audio here if you want to think it through.
48:40
SPEAKER_00: That is a classic claim of face validity.
48:43
SPEAKER_00: On sheer observation,
48:44
SPEAKER_00: the behavior,
48:45
the physical manifestation of the withdrawal looks identical to
48:48
the human symptom
48:49
SPEAKER_00: it is intended to model.
48:50
SPEAKER_00: It doesn't guarantee the neurochemistry matches, but it looks right.
48:53
SPEAKER_01: Scenario two.
48:55
SPEAKER_01: You're consulting on a neurobiology paper.
48:58
The team is electrically stimulating the dorsal periaqueductal
49:01
gray,
49:02
the DPAY,
49:03
and RAT.
49:03
Which specific human psychiatric phenomenon are they most
49:06
accurately modeling?
49:08
SPEAKER_01: Pause the audio here if you want to think it through.
49:09
SPEAKER_00: They are modeling the respiratory subtype of panic disorder.
49:12
SPEAKER_00: Because the DPAY harbors that specific hypoxia-sensitive alarm system,
49:16
it perfectly mirrors panic attacks characterized by a sense of
49:19
smothering,
49:20
SPEAKER_00: asphyxiation,
49:21
SPEAKER_00: or profound air hunger.
49:22
SPEAKER_01: Scenario three.
49:24
SPEAKER_01: A massive pharmaceutical company synthesizes a novel compound.
49:29
It demonstrates incredibly potent anxiolytic effects in the
49:32
elevated plus maze.
49:33
SPEAKER_01: The rats are exploring the open arms fearlessly.
49:37
But it fails completely in phase two clinical trials for human
49:39
generalized anxiety disorder.
49:42
What known limitation of the plus maze likely explains this
49:45
multi-million pound failure.
49:47
SPEAKER_01: Pause the audio here if you want to think it through.
49:49
SPEAKER_00: The failure stems from the test's inherent historical bias.
49:52
The elevated plus maze was primarily developed and validated to
49:55
detect the
49:56
SPEAKER_00: GABAergic mechanisms of benzodiazepines.
49:59
SPEAKER_00: It is notoriously insensitive to other mechanisms.
50:02
SPEAKER_00: For instance, it frequently fails to detect the effects of SSRIs.
50:06
The novel compound likely worked via a non-benzodiazepine
50:09
mechanism that the maze
50:10
SPEAKER_00: was biologically blind to.
50:12
SPEAKER_01: Scenario 4.
50:13
SPEAKER_01: A lab is studying declarative spatial memory,
50:15
but they decide to explicitly abandon the Morris water maze in
50:19
favor of the Barnes
50:19
SPEAKER_01: maze.
50:20
SPEAKER_01: Why would they make that specific methodological choice?
50:23
SPEAKER_01: Pause the audio here if you want to think it through.
50:24
SPEAKER_00: Scenario 5
50:52
If you want to model the true clinical reality of human
50:55
substance use disorder,
50:56
SPEAKER_01: the absolute loss of control,
50:58
SPEAKER_01: which behavioral paradigm gives you the highest construct validity,
51:02
SPEAKER_01: pause the audio here if you want to think it through.
51:05
SPEAKER_00: Drug self-administration maintained despite aversive foot shock.
51:08
SPEAKER_00: That active model captures the core destructive compulsion of addiction,
51:12
the willingness of the subject to continue seeking the drug
51:15
even when it guarantees
51:16
SPEAKER_00: immediate,
51:16
SPEAKER_00: severe,
51:17
SPEAKER_00: and escalating physical punishment.
51:19
SPEAKER_01: Scenario 6.
51:20
SPEAKER_01: Last one.
51:20
SPEAKER_01: The study of naturally occurring narcolepsy in canine models,
51:23
SPEAKER_01: the Dobermans collapsing when excited,
51:25
led to the breakthrough discovery of which crucial
51:27
neurobiological system.
51:28
SPEAKER_01: Pause the audio here if you want to think it through.
51:30
It led directly to the discovery of the orexin or hypocretin
51:34
receptor system.
51:35
It remains the gold standard example of an animal model
51:38
providing perfect construct
51:39
validity that translated flawlessly into human pathophysiology
51:43
and subsequent drug
51:44
SPEAKER_00: development.
51:45
SPEAKER_01: So what does this all mean?
51:47
I want to leave you with a final provocative thought to mull
51:50
over as you head back
51:50
SPEAKER_01: to the clinic or step back into the lab.
51:53
We spend a massive amount of time wringing our hands over the
51:56
preclinical bottleneck.
51:58
We lament the fact that our animal models with their pristine
52:01
CRISPR knockouts and
52:02
SPEAKER_01: perfectly controlled environments are consistently highlighting discrete,
52:06
SPEAKER_01: reliable biological mechanisms.
52:09
SPEAKER_01: But then the drugs fail in human trials.
52:11
SPEAKER_00: We constantly blame the animals.
52:13
SPEAKER_00: We say the mice are failing to model human disease.
52:15
SPEAKER_01: Exactly.
52:16
SPEAKER_00: Yeah.
52:16
SPEAKER_01: But maybe the animals aren't failing at all.
52:18
Maybe our human diagnostic manuals are failing to map onto
52:22
biological reality.
52:23
Perhaps the mice are trying to tell us that major depressive
52:25
disorder isn't one disease.
52:27
Maybe it's a dozen different distinct biological lesions all
52:31
masquerading under the
52:32
SPEAKER_01: exact same syndromal trench coat.
52:34
SPEAKER_01: If we actually listen to the precision of the animal models,
52:37
it might just force us to completely burn down and rewrite how
52:40
we diagnose human
52:41
SPEAKER_01: beings.
52:41
And that translation from the tide pool to the open ocean is
52:44
the ultimate goal of
52:46
SPEAKER_00: psychiatric neuroscience.
52:47
SPEAKER_01: Here are the details of this episode.
52:49
SPEAKER_00: Animal models in psychiatry.
52:51
SPEAKER_00: Byline.
52:52
SPEAKER_00: Approximating the human mind, what models reveal, and what they cannot.