Lateral prefrontal cortex

Goals

• To discuss the effects of frontal lobe brain damage on human behavior and executive processes.
• To review the gross anatomy of the frontal lobes.
• To discuss the function of prefrontal cortex with respect to executive processes.
• To discuss the function of prefrontal cortex with respect to working memory.

Topic slide

Robert Knight (b. 1948) is a neurologist, presently at UC Berkeley, who has conducted extensive studies of the frontal lobes in humans. He has created a registry of patients with damage to the frontal lobes caused by strokes, and made these patients available to many collaborators.

Patricia Goldman-Rakic (1937-2003) was a professor of neurobiology here at Yale until her untimely accidental death in 2003. She studied the frontal lobes, primarily of monkeys, using a variety of anatomical, neurophysiological, and behavioral techniques. She championed the role of prefrontal cortex as an important substrate of working memory.

Reading

• Reading: PN6 Chapter 27

Overview

The frontal lobes play an outsized role in executive functions and working memory. However, it is important to recognize at the outset that other brain regions also participate in these functions.

It is also important to note that the frontal lobes occupy about 41% of the human cerebrum according to a 1998 study by Kennedy and colleagues. The frontal lobes also play important roles in other functions; for example, the caudal (posterior) part of the frontal lobes comprise premotor and primary motor cortex.

I will use the figure below to organize my lecture. This figure emphasizes three components of executive function:

While I like this summary, it should be evident that many of these concepts are difficult to cleanly separate. For example, in initiating a new rule, one must inhibit the old rule. Shifting rules means to express one rule while inhibiting the alternative rule. These rules, and the basis for shifting among them, may be held in working memory.

Clinical case studies

As we will see, patients with frontal lobe damage show deficits in planning, persevere with non-optimal responses, show altered emotional responses, and show memory disturbances (particularly for source of information, or recency of memory). I presented four illustrative case studies.

Dysexecutive syndrome

One of Robert Knight's patients, W.R., illustrates the range of deficits observed with frontal lobe lesions. Patient W.R. had a large bilateral glioblastoma of the frontal lobes. He provides a good, though tragic, example of dysexecutive syndrome. Test results show no decrease in W.R.'s intelligence. However, he showed deficits in the following domains:

• Planning – e.g., never took bar exam after completing law school.
• Working memory, temporal sequencing (as a tennis pro, he forget score, whose serve)
• Anhedonia – didn’t take pleasure from life – stopped dating, etc.
• Inappropriate affect
  • Responded with flat affect to his own grim prognosis, also responded similarly to death of his mother.

You can read a full account of patient W.R. here.

Concrete thinking

W.R. could also not generate counterfactuals. I described the case study of second patient with frontal lobe lesions who could not simulate alternative outcomes; i.e., the patient had no ability to generate a counterfactual. Most individuals, when shown an image of a tragic scene (e.g., a car crash) and asked to reappraise the situation to give it a more positive outcome, can do so easily (e.g., 'help is on the way', 'the accident is not as bad as it looks', 'at least nobody was injured'). This patient, Mrs. M., had extremely concrete thinking and could not reappraise. Rather, she got stuck on the details of the picture and demonstrated concrete thinking.

The following is an extended quote from the paper by Salas et al. which describes the patient. The emphasis is mine.

Concrete behavior, the inability to disengage from immediate experience in order to manipulate ideas and thoughts, has long been understood to be a common problem after frontal lobe lesions. However, there has been little consideration of the impact that concreteness may have on emotional functioning, specifically in the use of thinking to manipulate emotional responses. One widely studied emotion regulation strategy is reappraisal, which depends on several frontal lobe related cognitive control processes. While there have been numerous neuroimaging findings on reappraisal, no study has used brain injured patients to investigate this issue. The present case study is the first to describe the capacity to generate reappraisals in a patient (Mrs M), whose behaviour became concrete after a left prefrontal stroke. Using a picture-based reappraisal paradigm, her performance was compared to non-concrete brain-lesioned patients, and neurologically healthy controls. Although Mrs M showed relatively preserved overall cognitive function, she was completely unable to spontaneously generate reappraisals.

Again, quoting from Salas et al.

Reappraisal involves changing the way that we think about a situation, usually reframing the meaning of an aversive event in less negative or more positive terms … To achieve this goal, several cognitive control processes are thought to be required, including: (1) keeping the automatic appraisal of a negative situation to mind (working memory), (2) decreasing the salience of that appraisal (inhibition), (3) generating alternative interpretations (working memory manipulation and verbal ability), (4) identifying and engaging the interpretation that is most appropriate (set shifting), (5) keeping the new appraisal in mind (working memory maintenance), and (6) keeping track of the success of regulation (monitoring)..

Disinhibition syndrome

The term disinhibition (inhibition of inhibition) suggests the release of behaviors that are ordinarily inhibited. The French neurologist Lhermitte described very curious behavior of frontal patients. Lhermitte’s patients had disinhibition syndrome (sometimes known as environmental dependence syndrome). These patients cannot help but to act on objects in their environments – even when it is highly inappropriate.

Lhermitte created elaborate scenarios and brought patients into those scenarios to see how they would behave.

• Examples included a woman who picked up a syringe and injected Lhermitte. When put in a room with knitting needles and wool, she immediately started knitting.
• Another example was of a man who, upon seeing a bed in the doctor’s apartment, disrobed and got into the bed. When brought into a room in which a painting was leaning against the wall, he picked up a hammer and nail and hung the picture.
• Interesting side note: in a ‘cocktail party’ scenario, the woman immediately started serving others, the man immediately acted like a guest.

Lhermitte's patients can also be said to engage in utilization behavior which he described in a 1983 paper in the journal Brain, entitled 'Utilization behaviour' and its relation to lesions of the frontal lobes'. Quoting from his paper:

"The tactile, visuotactile and visual presentation of objects compels the patients to grasp and use them. This behaviour was obtained with miscellaneous utilitarian objects. For the patients, the presentation of objects implies the order to grasp and use them. It is proposed that the balance between the subject's dependence on and independence from the outside world is disturbed. With frontal lesions, the inhibitory function of the frontal lobes on the parietal lobes is suppressed. The result is a release of the activities of the parietal lobes so that the subject becomes dependent on visual and tactile stimulation from the outside world. Five cases are reported as examples: one anatomoclinical case with bilateral lesions of the frontal lobe. The role of lesions affecting different parts of the frontal lobes is discussed."

Acquired psychopathy

The findings for patients with damage restricted to dorsolateral PFC are typically different from patients with damage to orbitofrontal and ventral PFC. A case study of patient with an orbital frontal lesion following a car accident as an infant reveals this distinction. The patient showed an apparent quick recovery, but demonstrated the following:

• At 3 years, she was observed not to respond to verbal or physical punishment.
• She engaged in chronic lying, stealing, impulsive, aggressive towards others, risky and promiscuous sex leading to a teenage pregnancy.
• She could not keep job, and ignored the infant.
• She never expressed guilt or remorse.
• When imaged at age 20, bilateral damage of orbital and ventromedial prefrontal cortex was discovered.

This was a case of acquired psychopathy, a topic for later discussion this semester.

Anatomy of frontal lobe

As I stated above, the frontal lobes constitute about 41% of the cerebral cortex. I showed several slides illustrating the extent of the frontal lobes on the dorsal and ventral surfaces, and on the medial surface.

Major gyri of the frontal lobe

• Superior frontal gyrus
• Middle frontal gyrus
• Inferior frontal gyrus
  • Pars Triangularis
  • Pars Opercularis
  • Pars Orbitalis
• Orbital frontal gyrus
• Gyrus rectus
• Precentral gyrus (primary motor, M1)

Prefrontal cortex

Prefrontal cortex is that part of the frontal lobe anterior to premotor cortex. Prefrontal cortex can be defined by its thalamic connections, which are through dorsomedial thalamus (also called the medial dorsal thalamus).

The exact borders between different areas of prefrontal cortex are difficult to draw. Nevertheless, researchers frequently distinguish among the following areas.

• Lateral prefrontal cortex
  • Dorsolateral Prefrontal Cortex (dlPFC)
  • Ventrolateral Prefrontal Cortex (vlPFC)
• Dorsomedial Prefrontal Cortex (dmPFC)
• Ventral prefrontal cortex
  • Ventromedial Prefrontal Cortex (vmPFC)
  • Orbital Frontal Cortex

As we learned with language disorders, the strokes or causes of brain damage can be quite extensive and encroach upon several different regions. For example, a car accident that damages ventral prefrontal cortex will likely involve both vmPFC and orbital frontal cortex.

Thus, there is a degree of imprecision related to the function of the frontal lobes and to the anatomy of the frontal lobes.

Executive functions

Most of my discussion of Executive Functions and Working memory will focus on Dorsolateral and Ventrolateral Prefrontal cortex.

Shifting and initiating behavior

One task that is often used to demonstrate shifts in behavior is the Oddball task. I presented an example of this task drawn from a study conducted by Goldman-Rakic and me. Here are the salient points from that discussion:

• Subjects view a train of stimuli that include ‘standards’ and ‘targets’. The standards require one type of response (which can be the withholding of a response – a ‘no go’). The targets require another type of response. Usually there are many more standards than targets.
  • A variant of the task includes infrequent ‘novel’ stimuli that have the same frequency of occurrence as ‘targets’, but which do not require a response.
• Target events (whether the response is button pressing, or simply keeping a mental count) activate dorsolateral prefrontal cortex (specifically, the middle frontal gyrus).
• Novel events do not activate these areas when they are task-irrelevant and do not require a specific response.
• However, if you switch the task-relevance of targets and novels by requiring subjects to respond to what were previously novels and ignore the previous targets, then the novels activate the dorsolateral PFC instead.

What cognitive functions do oddball tasks engage?

• Oddball tasks require subjects to switch responses (i.e., respond with one response to a target and a different response to a standard).
• But they also require a shift in the task rules – after a long string of stimuli requiring one kind of response, this new stimulus requires a different response – i.e., the oddball task also requires response selection.
• But in switching to a new response, the subject must now withhold the prepotent response – i.e., the response to the prior stimulus.
• So oddball tasks involve task switching, rules, response selection, and response inhibition – it is difficult to disentangle these task components. But the oddball task does show that the requirement to stop doing the thing you’ve been doing and now do something different engages dorsolateral PFC (dlPFC).

Planning

A standard task used to study planning is the Tower of London puzzle in which a subject must move colored balls on a spindle to match a pattern, and complete this in the fewest moves.

Patients with left anterior frontal lesions were far worse on this task than control groups that had brain lesions in other regions.

Task complexity

Some studies have revealed a posterior-to-anterior gradient of frontal lobe activation as a function of task complexity. Tasks with greater complexity engage progressively more anterior regions of frontal cortex.

Stimulus and Response Selection

I used the Filtering Task as an example of stimulus and response selection. In this task, subjects are asked to generate actions (verbs) you could do with an object. Low filtering objects had few possible actions while High filtering objects hand many possible actions. These are the typical results:

• The High Filtering objects activated inferior frontal cortex in healthy individuals.
• Patients with inferior dorsal lateral frontal cortex lesions have difficulty performing the high filtering task. This corroborates the imaging results in healthy individuals.

Rule Following, Capture, and Perseveration

A favorite task of neuropsychologists is the Wisconsin Card Sort task (WSC). Here is a summary of the points I made in lecture:

• Subjects are presented with cards with simple shapes in different colors. There can be one or more identical objects on the cards. Thus, the cards can be matched on the basis of color, symbol shape, or number. The subject matches a card and the experimenter gives feedback as to whether the response is correct or not. The experimenter then changes the matching rule without telling the subject, who must then learn the new rule.
• Patients with frontal lobe lesions do poorly on this task. They tend to perseverate with the old rule and do not shift to a new rule despite the negative feedback from the experimenter.
• Healthy subjects perform well. fMRI studies show strong activation of dorsolateral PFC during a version of this task designed for scanning. In this version, subjects were given a cue to shift the matching rule. This then becomes very similar to the Oddball task.

Inhibition

One of the most interesting aspects of prefrontal function is disinhibition. This has been studied in patients (as described above) with frontal lobe damage in behavioral tasks. It has also been studied using physiological measures, as summarized below:

• Lesions of prefrontal cortex enhance auditory, somatosensory, and visual sensory evoked potentials. This suggests that sensory input in healthy individuals is inhibited by PFC, and thus disinhibited by PFC lesions.
• I provided an example of the auditory P30 response – an evoked potential occurring about 30 msec after the onset of an auditory tone. Frontal lesions enhance the amplitude of P30, suggesting that is it normally inhibited by the frontal lobe.

Are PFC patients better at some tasks than healthy controls?

The MatchStick Task tests the hypothesis that the dlPFC is crucial for defining a set of responses suitable for solving a particular task, and the biasing the selection (‘sculpting the response space’, Frith, 2000). Patients with dlPFC lesions were better at solving atypical (i.e., ‘outside the box’) problems than healthy controls.

Frontal lobes and memory

Recency memory

Patients with frontal lobe lesions do poorly in remembering which stimulus was more recently seen than a competing stimulus. That is, they have poor memory for when a stimulus was presented.

Source memory

Patients with frontal lobe lesions do poorly in remembering where they learned a particular fact. They remember the fact, but may falsely believe they had already known the fact, or falsely believe they learned the fact earlier in the experiment.

While source memory difficulties are widely reported with frontal lesions, a more recent study by Thaiss and Petrides has found no such deficits. Rather, these investigators found source memory deficits with temporal lobe lesions, and suggest earlier studies included patients with very large lesions that included both frontal and temporal lobe. Thus, there is some controversy about whether source memory is a frontal lobe function.

Working memory

Working memory in monkeys

Monkeys with dlPFC lesions do poorly on delayed memory tasks that depend upon working memory. An example of such a task would have a monkey observe an experimenter bait one of two covered wells with food. A partition would then be placed between the monkey and the food wells so that the monkey can no longer see the wells. When the partition is removed after the delay, the monkey with dlPFC lesions cannot remember which well was baited with food.

However, the dlPFC monkeys can perform a slightly modified version of the task that associates a visual cue with the baited well. This task depends upon associative (long term memory) and not on working memory.

Oculomotor delayed response task

Patricia Goldman-Rakic devised a working memory task in which a monkey was required to fixate on a central cross in a visual display. While fixating, a visual target stimulus was briefly illuminated in the periphery of the display and then turned off. While maintaining fixation on the central cross, the monkey had to keep in working memory the location of the now invisible target. When the central fixation cross was extinguished a few seconds later, the monkey had to make a saccade to the remembered location. Correct saccades were rewarded with juice. Here are the important results:

• Neurons in the monkey’s dlPFC fired during the delay period when the target location was not visible, but was held in working memory. Different cells encoded different spatial locations – i.e., cells had ‘memory fields’.
• Cells in the monkey’s intraparietal sulcus also fired during the delay period.
  • We will later encounter a parietal-frontal reaching system (Lecture 16) that has similar properties to this working memory system.
• These cells were interpreted as components of a system for maintaining locations in working memory.

Working memory studies in humans

Collaborative studies by Goldman-Rakic and McCarthy showed that the right middle frontal gyrus with dlPFC was strong activated during spatial working memory tasks in humans. Activation in this region increased when more items were held in working memory. But activation plateaued after 3 items.

Some studies have suggested that activity in category-specific regions of cortex (e.g., fusiform gyrus face area, and parahippocampal place area) also increases in the delay period of a working memory task. This may indicate that dlPFC activation is keeping ‘online’ stimulus content that is held in other brain regions.

Working memory and attention

Many researchers (including Marvin Chun from Yale, and Anna Nobre from Oxford) have argued that working memory is best conceptualized as a form of attention that is directed to internal representations. Thus, the frontal-parietal control systems we discussed for selective attention are also the same control systems used for working memory.

Involuntary attention (emotional distraction) and Working Memory

I used a 2006 study from my lab as an example of the effects of distraction upon working memory. This study illustrates the manner in which novel, distracting information commandeers attention and disrupts working memory processes in dlPFC. Here is the gist:

• Subjects were briefly shown 3 faces that disappeared at the start of a delay period. After the delay period, a single face appeared and subjects had to report whether that single face was one of the three faces previously seen.
• During the delay period, task-irrelevant distracting visual stimuli were presented. These stimuli were either patterns, neutral pictures, or emotional pictures.
• dlPFC showed strong activation during the delay period (as we’ve already discussed in other experiments). However, dlPFC activation was reduced when the neutral and emotional pictures were presented – particularly when the emotional pictures were presented.
• Task performance also went down with distraction – particularly emotional pictures.
• vlPFC showed strong activation by the emotional distracters, as did the amygdala. In other words, there was a reciprocal relationship between dlPFC and vlPFC.
• In this study it is not clear whether the vlPFC activity represents a ‘circuit-breaker’ process.
  • Other studies suggest that the vlPFC activity represents protection against distraction.

Videos

Prerecorded lecture for Fall 2020

This prerecorded video transitions from lecture 16 into 17

Previously recorded live lectures

The video embedded below was recorded in Fall, 2019.