Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Nov 1.
Published in final edited form as: Neuroscience. 2022 Sep 11;503:17–27. doi: 10.1016/j.neuroscience.2022.09.007

A possible mechanism for development of working memory impairment in male mice subjected to inflammatory pain

Alexander Papadogiannis 1, Eugene Dimitrov *
PMCID: PMC9588797  NIHMSID: NIHMS1835708  PMID: 36100034

Abstract

We studied the effects of inflammatory pain on working memory and correlated the pain effects with changes in dendritic spine density and glutamate signaling in the medial prefrontal cortex (mPFC) of male and female mice. Injection of Complete Freund’s Adjuvant (CFA) into the hind paw modeled inflammatory pain. The CFA equally decreased the mechanical thresholds in both sexes. The density of dendritic spines, as a marker for neuronal input, increased on the dendrites of both, pyramidal cells and interneurons in males but only on the dendrites of interneurons in CFA injected females. Next, we injected virus with glutamate sensor (pAAV5.hSyn.iGluSnFr) into the mPFC and used fiber photometry to record glutamate signaling during Y-maze spontaneous alternations test, which is a test for working memory in rodents. The detected fluorescent signal was higher during correct alternations when compared to incorrect alternations in both sexes. The CFA injection did not change the pattern of glutamate fluorescence during the test but the female mice made fewer incorrect alternations than their male counterparts. Furthermore, while the CFA injection decreased the expression of the glutamate transporter VGlut1 on the soma of mPFC neurons in both sexes, the decrease was sex dependent. We concluded that inflammatory pain, which increases sensory input into the mPFC neurons, may impair working memory by altering the glutamate signaling. The glutamate deficit that develops as a result of the pain is more pronounced in male mice in comparison to female mice.

Introduction

The mPFC regulates an array of cognitive processes, namely decision making, problem solving, emotional inhibition, attention, learning, long-term and working memory (Euston, Gruber and McNaughton, 2012). Numerous studies in humans and laboratory animals describe the dependence of working memory on other cortical and subcortical regions but indisputably the frontal cortex is the most essential component for both, memorization of information and execution of an action plan (Christophel et al., 2017; Euston, Gruber and McNaughton, 2012; van Asselen et al., 2006). Functional studies in human subjects describe the cortical processes such as short-term potentiation, glutamate release and interaction between the activity of different layers as fundamental for working memory (Woodcock et al., 2018; Bastos et al., 2018). Despite certain discrepancy in comparative nomenclature (Laubach et al., 2018), the rodent mPFC is accepted as anatomical and functional equivalent to the human and primate anterior cingulate and dorsolateral prefrontal cortex (Seamans, Lapish and Durstewitz, 2008). The rodent tests for working memory include various navigation and preference tasks. An elegant recent publication demonstrated that individual neurons as well as neuronal assembly in the mPFC are activated by both location on the maze and choice that rats make during the T-maze test (Yang and Mailman, 2018). Another study assigned the encoding of working memory to a subpopulation of excitatory neurons in the mPFC with stronger interconnectivity (Tian et al., 2018). The immense complexity of working memory brings into question the susceptibility of the memory processes to environmental perturbations that cause cognitive impairment such as stress and pain as major pathologies associated with cognitive decline.

The mPFC receives ample sensory information including nociceptive input (Tan et al., 2019) and it is the place where pain experience becomes conscious (Ong, Stohler and Herr, 2019). The mPFC regulates nociception by inhibiting the affective and sensory components of pain via long projections to the periaqueductal gray (PAG) and brain stem nuclei (Martin et al., 2013; Ong, Stohler and Herr, 2019; Jiang et al., 2014). However, persistent pain triggers an array of functional and morphological changes in the CNS defined as central sensitization (Latremoliere and Woolf, 2009). Diminished cognition, including working memory impairment, is an example of this process, one of many well recognized maladaptive consequences of chronic pain (Mazza, Frot and Rey, 2018). Neuroimaging studies in humans demonstrate that chronic pain leads to deactivation and thinning of the prefrontal cortex, decrease of neurotransmitters and alterations in neuronal activity (Kang et al., 2019). In rodents, where the mPFC is parceled into anterior cingulate (ACC), prelimbic (PrL), infralimbic (IL) and medial orbital (MO) cortex (Van De Werd et al., 2010), numerous studies describe similar changes in the activity and organization of the cortical neurons. The deleterious effects of chronic pain include changes in neuronal excitability and activity (Wei and Zhuo, 2001; Wu, Liang and Gao, 2016), connectivity (Cardoso-Cruz, Lima and Galhardo, 2013), release of neurotransmitters (Kang et al., 2021; Hung et al., 2014), receptor expression (Chung et al., 2017), the length and complexity of dendritic arbors (Kelly et al., 2016; Metz et al., 2009). All these changes define the process of central sensitization in the mPFC with subsequent disruption of the descending pain modulation circuit (Huang et al., 2019; Latremoliere and Woolf, 2009) and cognitive deterioration (Kummer et al., 2020).

Despite our knowledge about the negative impact of persistent pain, there are still major gaps in matching cognitive deficits to specific morphological and functional changes in the mPFC triggered by the pain conditions. Here, we used an inflammatory pain model to trace changes in the dendritic density of cortical neurons as a general marker for the level of neuronal input to the mPFC. In the following experiments, we compared the in vivo glutamate signaling during working memory task before and after seven days of inflammatory pain and finally, we evaluated the expression of VGlu1 as a presynaptic marker for glutamatergic input into the mPFC neurons.

Methods

Animals.

All animal work was done in accordance to the guidelines set by the IACUC at Rosalind Franklin University of Medicine and Science. Adult outbred male and female CD-1 mice, purchased from a vendor (Charles River, Worcester, MA), were used in these experiments. The mice were housed in conventional cages located in controlled AALAC-approved facility with maintained room temperature (20–22°), humidity (50–55%) and illumination (10:14 h light/dark cycle). Food and water were available ad libitum.

1. Behavior

1.1. Von Frey test:

Injection of CFA was used as a well-established rodent model for persistent inflammatory pain that does not inhibit locomotion, grooming and weight gain (Ren and Dubner, 1999). The control groups were injected with normal saline while the experimental groups were injected with 20 μl of CFA into the plantar side of the left hind paw.

Mice were habituated to the testing room and a box with mesh floor before nociceptive testing. Tactile sensitivity was measured using von Frey filaments (North Coast Medical, Morgan Hill, CA) applied to the plantar surface of the hind paw through the mesh floor. Filaments with different stiffness (range 0.008 g to 8 g weight) were used for each measurement. The starting filament was always 1.0 g weight. A quick withdraw, shaking or licking of the paw were considered a positive reaction.

The mechanical allodynia was assessed at the 7th post injection day as described by Chaplan (Chaplan et al., 1994) and the mechanical thresholds were calculated by Dixon’s up and down method (Dixon, 1991). We used within-subject comparison for the photometry experiment, therefore the CFA injection was applied after obtaining base level thresholds and the mice were retested seven days after the CFA application.

1.2. Spatial memory test:

Y-maze spontaneous alternation test for rodents is used to asses spatial reference memory stored in the hippocampus and working memory, which depends on mPFC (Kraeuter, Guest and Sarnyai, 2019). The test relies on the rodent’s exploratory behavior and ability to navigate closed spaces. The mice normally prefer to enter a new arm instead of entering back into the just visited arm. The maze consists of three equal size arms (5 × 35 cm with 10 cm walls) positioned at 120° angle to each other. The mice, with the photometry headgear already attached, were placed in the center and left undisturbed to explore the maze. The video record was started 100 seconds after the mouse was placed on the maze and continued for another 300 seconds. The video records were used to estimate the total number of visits to the three arms and the number of correct entrances.

2. Golgi-Cox impregnation

2.1. Golgi staining:

The mice were killed with a lethal dose of pentobarbital, the brains removed, snap-frozen on dry ice and cut into 100 μm thick sections on a cryostat after marking the right hemisphere. The sections were processed following FD Rapid Golgi Stain protocol (FD NeuroTechnologies, Inc, Columbia, MD). Briefly, the sections were incubated in impregnation solution, in a dark container, at room temperature for 72 hours; the solution was replaced after 24 hours of incubation. After subsequent washes and a developing step, the sections were mounted on microscopic slides, dehydrated with ethanol, delipidated with chloroform substitute and coversliped.

2.2. Microscopy:

Bright-field images were obtained with Leica DM 5500 microscope equipped with a 100 X oil objective. The high-resolutions Z-stacks (75 to 95 single slides, 1392 × 1040 pixels, pixel size 0.01388 × 0.01388 μm, voxel depth 1 μm) were obtained from the entire tissue thickness of the entire right mPFC. The images were transferred to Image J for analysis. Third and fourth order neuronal dendrites with intact terminal tufts were selected for assessment of spine density. All visible spines were counted on 20 to 40 μm long dendritic segments and sorted into different categories by two independent experimentators blind to the animals’ condition. The spines on two to five dendrites from three to five neurons per animal were averaged separately for the two main types of cortical neurons, pyramidal cells and interneurons. The neurons were categorized as pyramidal based on their distinguished shape and prominent dendritic organization consisting of a single central dendrite on the opposite cell pole of a cluster of apical dendrites. The interneurons were sorted according to their smaller, round bodies with evenly radiating dendrites. The different types of spines, namely stubby, long, mushroom, branched and filamentous (Risher et al., 2014) were expressed as an average spine density, which was used for group comparison.

3. Stereotaxic surgeries.

3.1. Viral injections:

Mice were anesthetized with isoflurane at flow rate 400 ml per minute, 3% induction and 1–2% for maintenance. After confirming anesthesia by tail pinch, the mice were placed in the stereotaxic frame (Stoelting, Wood Dale, IL) and the skull was exposed via midline skin incision. Stainless-steel 32-gauge needle with a 10 μl Hamilton syringe was lowered into the right mPFC, through an opening drilled through the skull with coordinates +1.7 mm AP, − 0.4 mm right L and −2.0 mm DV to bregma. A volume of 120 nl of pAAV5.hSyn.iGluSnFr, viral particles titer 7×1012 per ml (cat. # 98929, Addgene, Watertown, MA) was injected over 2-minute period with an infusion pump and the needle remained in place for another 2 minutes.

3.2. Fiberoptic cannulae:

Similar stereotaxic technique with the same coordinates was used for placing the fiberoptic cannulae for photometry recordings. The cannulae (ferrule diameter 2.5 mm, fiber core 400 μm/NA0.39, fiber length 3mm, TeleFipho, Amuza Inc., Sen Diego, CA) were secured to the skull with bone screws and dental cement. The animals received analgesia with Flunixin in dose 2 mg/kg subcutaneously for three days and remained undisturbed for two to three weeks after the surgeries.

4. Photometry

4.1. Recording:

We used pAAV5.hSyn.iGluSnFr to deliver a high-sensitivity glutamate sensor into the mPFC. The glutamate sensor allows for in vivo recording of task-dependent activity from the transfected neuronal populations (Marvin et al., 2013). As described above, the mice were implanted with fiberoptic cannulae three weeks after the viral injections. The wireless photometry system (TeleFipho, Amuza Inc. San Diego, CA) allows for recording of fluorescent signal without an optical cable, therefore avoiding problems with cable artefacts and impeded animal movements. Each mouse underwent two training sessions to place the headgear and acclimate for wearing the headgear. The headgear consists of wireless transmitter emitting LED light at 495 nm with adjustable power from 10 to 300 μW and captures the fluorescent signal at speed 100 Hz by a photodetector with GFP wavelength filter. The device transmits the signal to a digital receiver with an antenna. The first two to three minutes of photometric recording were used to adjust the power of the emitted light and stabilize the fluorescent signal. The mice were placed on the Y-maze once the signal reached consistency and a video record of the mice ambulating on the maze was started after the first 100 seconds in order to avoid recording of cell activity caused by handling of the animal. The simultaneous photometric and video recordings continued for five minutes. The video records were used to timestamp the correct and incorrect alterations of the mice. The last second before entrance into a new arm, which coincides with the moment when the mouse head aligns with the chosen arm, was selected as timestamped event and categorized as correct or incorrect alternation depending on the animal choice.

4.2. Analysis:

The raw traces were processed by Amuza software and the maximum fluorescence was determined for the entire record in one second steps. Next, spreadsheets with time/fluorescent data were upload for further processing and analysis to pMAT, an open-source software (Bruno et al., 2021). The pMAT program creates a scaled control channel by fitting an exponential curve into the fluorescent signal, reduces noise and calculates ΔF/F for the entre trace. The fully processed ΔF/F plots than were presented as a normalized Z-score trace which included reference ticks for timestamped events. The Z-scores of the second when the animal head aligned with the chosen arm were compared between correct and incorrect alternations, before and seven days after CFA injection. The recorded Z-scores of correct and incorrect alternations were averaged per animal and plotted for statistical analysis. Each mouse underwent two photometry tests, one before and one seven days after CFA injection into the hind paw.

5. Immunocytochemistry

5.1. Immunostaining for VGlut1 and GFP:

Mice of both sexes were stereotaxically injected into the right mPFC with pAAV5.hSyn.iGluSnFr as described in the previous section. The animals were killed one week after receiving 20 μl of saline or CFA into the plantar side of the left hind paw. The brains were fixed in 4% paraformaldehyde and 40 μm thick sections were cut with Leica vibrotome. The sections were incubated in blocking solution made of 2% normal donkey serum and 1 X PBS for 2 hours followed by incubation in cocktail of primary antibodies that contained rabbit anti-GFP antibody (# 75-131, PhosphoSolutions, Aurora, CO) in 1:2000 dilution and Guinea pig anti-VGlut1 (# AB5905, Millipore, Burlington, MA) in 1:10 000 dilution for 48 hours at 4° C. The staining was developed by 5-hour incubation in mixture of secondary antibodies, FITC anti-rabbit (# 711-225-152) and Alexa 594 anti-Guinea pig (# 706-005-148, Jackson ImmunoResearch, Inc., West Grove, PA), diluted 1:400. The sections were mounted on microscopic slides and coversliped with aqueous medium.

5.2. Microscopy:

Confocal Zeiss 510 microscope was used to obtain Z-stack images that included the entire thickness of labeled mPFC neurons. Only neurons with clear nuclei and intact cell shape with borders completely visible on both sides of the tissue sections were chosen for analysis. The collected Z-stacks consisted of 20 to 25 single slides, frame size 1024 × 1024 pixels, pixel size 0.06975 × 0.06975 μm and voxel depth of 0.42389 μm. An average of 5 to 6 Z-stacks per animal were collected and processed further in Image J. The Z-stacks were converted into 8-bit maximum projections, the red and the green channels were split into single channel, the neuronal bodies were outlined in the green channel and the outlines were transferred to the thresholded red channel. Next, we measured the integrated density (sum of all pixels above threshold expressed in arbitrary units) from the outlined area which included the entire surface of the cells. The results were averaged per animal and compared between the different groups.

Statistics.

For Golgi-Cox and VGlut1 staining experiments, the mice were randomly assigned to a control (10 males and 10 females) or CFA (12 males and 15 females) treated group. For photometry experiments, we followed a within-subject experimental design and the recordings were obtained before and after each mouse (9 males and 8 females) was injected with CFA. Data are presented as mean ± SEM and analyzed with Graph Prism 9 software. Two-way ANOVA followed by Tukey’s post-hoc analyses was used to evaluate the development of mechanical allodynia, total dendritic spine densities, working memory performance, photometric recordings and expression of VGlut1. The spine density of each type of dendritic spine was taken as independent variable and compared among the groups by One-way ANOVA and Tukey’s post-hoc. The accepted level of significance was P < 0.05 in all tests.

Results

Subcutaneous inoculation of CFA suspension into the plantar surface of the hind paw triggers inflammatory response which in addition to the development of swelling and redness of the paw also causes mechanical allodynia (hypersensitivity). The mechanical allodynia, defined as an increased response to non-nociceptive stimuli, is an expression of the phenomenon of central sensitization, which reflects the changes in CNS caused by pain (Latremoliere and Woolf, 2009). We applied von Frey filaments to the plantar surface of the hind paw to detect difference in the mechanical thresholds between vehicle injected controls and CFA treated groups (Figure 1A). In addition, the mechanical thresholds of mice assigned to the photometry experiment were evaluated before, and seven days after, CFA injection into the hind paw (Figure 1B). The mechanical thresholds were significantly decreased 7 days post-inoculation when compared to control mice (Figure 1A) or when compared to the pre-injection values (Figure 1B). The decrease in the mechanical thresholds was similar between the control groups and CFA groups of both sexes, from an average of 4.0 ± 1.4 g in control males and 4.4 ± 0.9 g in control females down to an average of 0.3 ± 0.4 in males and 0.2 ± 0.3 in females (Two-way ANOVA, significant for CFA factor F(1,42) = 246, P < 0.0001 but not significant for Sex factor F(1,42) = 0.9, P > 0.05 and Interaction F(1,42) = 0.4, P > 0.05). Comparable results were obtained from the second group of mice included in the photometry experiment, where the mechanical thresholds decreased after the CFA injection in both sexes, from an average of 4.2 ± 2.0 g in control males and 3.9 ± 0.8 g in control females down to an average of 0.2 ± 0.2 in males and 0.3 ± 0.3 in females (Two-way ANOVA, significant for CFA factor F(1,30) = 92, P < 0.0001 but not significant for Sex factor F(1,30) = 0.2, P > 0.05 and Interaction F(1,30) = 0.01, P > 0.05). Therefore, the inoculation of CFA into the hind paw led to the development of hypersensitivity in both sexes.

Figure 1: The CFA injection into the hind paw decreased the mechanical thresholds in male and female mice.

Figure 1:

Open in a new tab

Panel A shows the substantial decrease of the mechanical thresholds in the CFA groups when compared to their respective vehicle treated controls of both sexes seven days after the CFA application. The decrease of mechanical thresholds indicates the development of allodynia. Panel B shows similar difference in the mechanical thresholds that developed seven days after CFA injection in groups of mice, which were included in photometry experiment. **** - Main effect significant for CFA factor, - P < 0.0001. Data = Mean ± SEM, n = 9 to 12 for males and 10 to 12 for females in A and n = 9 for males and 8 for females in B.

Morphological changes in neuronal dendritic trees and dendritic spines as a part of the process of central sensitization by various pain models have been described in the relevant literature (Stratton and Khanna, 2020). Pain causes dendritic spine dysgenesis and changes of dendritic length in the spinal cord, hippocampus and cortex (Benson et al., 2020; Tyrtyshnaia and Manzhulo, 2020; Kelly et al., 2016). Yet, how pain affects the two main types of cells, pyramidal and interneurons, and if the potential pain effects are different in males and females is still not known. Here, we used Golgi-Cox staining method to compare the effects of CFA on dendritic spine density in the mPFC neurons. The Golgi impregnated neurons were divided into pyramidal and interneurons according to their shape, size and dendritic arbors, while the dendritic spines were grouped according to their general classification (Risher et al., 2014), (Figure 2).

Figure 2: The Golgi-Cox impregnation of neurons in the mPFC revealed the detail morphology of the stained cells.

Figure 2:

Open in a new tab

Panel A shows a group of pyramidal neurons with distinguished shape and an apical dendrite. Panel B shows an interneuron with round body and radiating, evenly spaced dendrites of a similar size. The dendritic spines, shown in C, were classified as “stubby” (1), “long” (2), “mushroom” (3), “branched” (4) and “filamentous” (5). Scale bar = 20 μm for A and B and 5 μm for C.

The CFA injection led to increases in the total spinal density of pyramidal neurons but only in male mice. While the average density of spines in the Male control group was 1.1 ± 0.2 spines per μm, the Male CFA group showed an average density of 1.6 ± 0.1 spines per μm. Treatment with CFA did not change spine density of pyramidal neurons in female mice (Female control 1.9 ± 0.3 spines/μm versus Female CFA with 1.8 ± 0.3 spines/μm. ANOVA comparison of the total spine counts was significant of Sex factor (F1,21 = 22.2, P < 0.001) and Interaction Sex X CFA (F1,21 = 8.2, P < 0.01) but not for CFA factor). Further analysis with Tukey’s post-hoc test revealed significant difference between Male control and Female control, P < 0.0001; Male control and Male CFA, P < 0.05; Male control and Female CFA, P < 0.01 (Figure 3A).

Figure 3. Inflammatory pain increased the dendritic spine density on pyramidal neurons in the mPFC of male mice but not in female mice.

Figure 3.

Open in a new tab

The graph in panel A compares the changes in the total spine density after CFA injection, where the total spine density increased in male mice but remained the same in females; Note that the dendritic spine density in control females is higher than the spine density in control males. Two-way ANOVA, post-hoc analysis: * P > 0.05, *** - P < 0.001 and **** - P < 0.0001, Data = Mean ± SEM. The graph in panel B shows that the CFA injection increased the total spine density in males without affecting significantly any particular type of spine. The density of dendritic spines on the pyramidal neurons of control female mice was higher than the males for stubby, long and mushroom spines but neither of the spine type densities was changed by the CFA injection in female mice. One-way ANOVA, post-hoc: * P > 0.05, ** -P < 0.01 and *** - P < 0.001. Data is expressed as “box-and whiskers plot”, Box = median and Whiskers = 5th to 95th percentile, n = 5 to 6 for males and 6 to 8 for females.

The effects of CFA injection were unevenly distributed among the different spine types. ANOVA analysis found significant difference among the groups in the densities of stubby, F3,21 = 3.6, P < 0.05; long, F3,21 = 5.4, P < 0.01 and mushroom spines, F3,21 = 11.4, P < 0.001 but did not find significant changes in the spine density of branched and filamentous spines (Figure 2 B). However, while the total spine density increased, the CFA injection did not change the density of any specific type of dendritic spine in males or females (Figure 3B).

In contrast, the CFA injection increased the total spine density on the dendrites of interneurons in both sexes. In male mice, the CFA injection caused increase from an average of 0.8 ± 0.2 spines/μm to 1.2 ± 0.1 spines/μm and in female mice from an average of 1.1 ± 0.1 spines/μm to 1.4 ± 0.1 spines/μm. ANOVA analysis of the total spine density was significant for Sex factor (F1,23 = 12.4, P < 0.01) and CFA factor (F1,23 = 19.4, P < 0.001) but not for Interaction. Tukey’s post-hoc test showed difference between Male control group and Female control group, P < 0.05; Male control and Male CFA, P < 0.05; Control female and CFA female, P < 0.05 and Control male and CFA female, P < 0.001 (Figure 4 A).

Figure 4. Inflammatory pain increased the dendritic spine density on interneurons in the mPFC of male and female mice.

Figure 4.

Open in a new tab

The panel A shows the total spine density of interneurons in control and CFA injected mice. Despite the higher spine density found in the control females when compared to control males, the CFA injection still increased the total spine density in females. Two-way ANOVA post-hoc analysis: * P > 0.05 and **** - P < 0.0001, Data = Mean ± SEM. Remarkably, the graph in panel B shows that only the density of mushroom spines in male mice was significantly increased by CFA injections. One-way ANOVA post-hoc: * P > 0.05, ** - P < 0.01 and *** - P < 0.001. Data is expressed as “box-and-whiskers plot”, Box = median and Whiskers = 5th to 95th percentile, n = 5 to 6 for males and 6 to 9 for females.

Significant differences among the densities of various spine types were found for long, F3,23 = 5.7, P < 0.01, mushroom F3,23 = 39.3, P < 0.0001, branched F3,23 = 4.3, P < 0.05 and filamentous spines F3,23 = 4.9, P < 0.01 (Figure 3 B). Interestingly, as in pyramidal neurons, the CFA injection did not increase the density of any specific type of dendritic spines in interneurons in female mice but did in male mice, the change between Control male and CFA male group was significant for mushroom spines (Figure 4 B).

The increased spine density in the cortical neurons triggered by inflammatory pain led us to suspect changes in glutamatergic signaling very likely driven by the increased nociceptive input to mPFC. Therefore, we tested the hypothesis that the CFA injection alters glutamate signaling, which may affect some of the cognitive functions executed by mPFC. To test our hypothesis, we used in vivo fiber photometry of fluorescent signal emitted by the glutamate sensor iGluSnFr during Y-maze spontaneous alternations test. The fluorescent signal showed multiple glutamate transients throughout the recording session but subsided to base line during periods of relative inactivity such as grooming (Figure 5 A and B). The numbers of correct alternations (the animal chooses to enter a new maze arm) were strongly aligned with intense glutamate transients but incorrect alternations (the animal chooses to enter the recently visited maze arm) were not and that reduced the overall ΔF/F Z-score of the wrong choices in both sexes; Two-way ANOVA main effect for sex and choice as variables showed significance only for choice (F1,30 = 5.9, P < 0.05) but not for sex or choice X sex interaction, P > 0.05 (Figure 5 D). After a week of recuperation, the mice were injected with CFA into the hind paw and the Y-maze test was repeated seven days after the CFA injection. The pattern of glutamate signaling during exploration of the maze remained the same, higher Z-scores accompanied correct alternations than incorrect alternations, which again led to statistical significance for choice (Two-way ANOVA, F1,30 = 25, P < 0.001, Figure 5 E). Furthermore, we did not detect significant difference between fluorescent signals corresponding to correct (Two-way ANOVA main effect for pain, F1,30 = 0.1, P > 0.05, sex, F1,30 = 1.4, P > 0.05 and interaction, F1,30 = 0.2, P > 0.05) or incorrect alternations (Two-way ANOVA main effect for pain, F1,30 = 2.7, P > 0.05, sex, F1,30 = 0.1, P > 0.05 and interaction, F1,30 = 1.6, P > 0.05) when compared before and after CFA treatment. However, while the CFA injection did not affect the total number of Y-maze entries (average number of entrances for Pre-CFA males 17 ± 3; pre-CFA females 16 ± 4; CFA males 14 ± 3 and CFA females 17 ± 5, Two-way ANOVA, main effect, P > 0.05 for sex and CFA), the male mice with CFA injection made fewer correct alternations when compared to the control males, (average number of correct entrances for pre-CFA males 14 ± 4; pre-CFA females 12 ± 3; CFA males 9 ± 3 and CFA females 13 ± 2; Two-way ANOVA, Tukey’s post-hoc Control males versus CFA males, q9 = 4.1, P < 0.5), which produced interaction between sex and CFA variable, (Two-way ANOVA, main effect, significant for interaction sex X CFA, F1,30 = 4.4, P < 0.5, Figure 5 F).

Figure 5: Effect of CFA injection on glutamate signaling in the mPFC during exploration of Y-maze.

Figure 5:

Open in a new tab

Panel A is an example of a raw fiber-photometric trace recorded from a female mouse during Y-maze test. Panel B shows the processed trace of the record in A with calculated Z-sore of normalized ΔF/F fluorescence. In order to avoid handling artefacts, the fluorescent signal and mouse entrances were scored after the first 100 seconds on the maze. The red and black marks above the trace indicate the 1 second time-stamped events of correct (red) and incorrect (black) entrances of the mouse. The marks were made visible on the panel by increasing their size approximately 10 times to font of 12. The Z-scores that correspond to the time-stamped events (the mouse head aligns with the chosen maze arm, correct or incorrect) were plotted for analysis. The high level of glutamate signaling persists throughout the test except during grooming, an example for which is found between 350 to 375 seconds on the trace record. Panel C shows the viral expression in the mPFC and a fiber-optic track, where the tip of the fiber is positioned above the PrL subdivision of the mPFC. Panel D is a Two-way ANOVA comparison of glutamate fluorescence of correct versus incorrect alternations made by male and female mice before they were injected with CFA; * - main effect significant for choice. The ANOVA graph in panel E analyses the glutamate fluorescence of correct and incorrect alternations on the seventh day post CFA injection; **** - main effect significant for choice. The graph in panel F compares the number of correct alternations made by mice of both sexes before and after the CFA injection. The male mice treated with CFA made fewer correct alternations than the control males, post-hoc analysis, *P < 0.05; * - main effect, significant for interaction Sex X CFA, Data = Mean ± SEM, n = 9 males and 8 females per group. Abbreviations: Cg – cingulate cortex, PrL – prelimbic cortex and IL – infralimbic cortex.

Next, we asked the question if there would be any quantifiable differences in the expression of the presynaptic VGlut1 transporter by the mPFC neurons transfected with pAAV5.hSyn.iGluSnFr. Mice of both sexes were injected into the right mPFC with 120 nl of the virus, followed three weeks later by CFA inoculation into the left hind paw. The mice were killed on the seventh post inoculation day and their brains processed for immunostaining. We used mice that never participated in photometry experiments because the implantation of fiberoptic cannulae produces debris and fibrous reaction from the surrounding area that may interfere with antibody penetration into the tissue and reduce the quality of the immunostaining. The pAAV5.hSyn.iGluSnFr transduced a number of neurons mostly in layers II and III of PrL subdivision of mPFC with a few scattered cells visible in layer V (Figure 6). The nature of the neurons, pyramidal or interneurons, could not be determined from the viral labeling (Figure 6 A to D and F). The presynaptic VGlut1 was extensively expressed throughout the cortex as small puncta. The density of VGlut1 puncta was estimated from maximum projections of Z-stacks incorporating the entire surface of the labeled cell bodies. The integrated density of VGlut1was similar between control males and females but was significantly lower in mice injected with CFA (Two-way ANOVA main effect significant for CFA, F1,16 = 45, P < 0.001 and sex factor, F1,16 = 6.6, P < 0.05 but not for interaction F1,16 = 0.4, P = 0.5). Overall, the inflammatory pain decreased VGlut1 expression in both sexes as shown by the post hoc analysis (Control males vs CFA males, q10 = 5.2, P < 0.001; Control Females vs CFA females, q10 = 4.2, P < 0.01 and Control females vs CFA males, q10 = 6.6, P < 0.0001, Figure 6 E).

Figure 6: CFA induced changes in the expression of the presynaptic marker VGlut1 by the mPFC neurons labeled by AAV2.hSyn.iGluSnFr.

Figure 6:

Open in a new tab

Panels A to D show the expression of VGlut1(red puncta) and iGluSnFr labeled neurons (green) in (A) control, (B) CFA treated male mouse, (C) female control and (D) CFA treated female mouse. The graph in panel E compares the expression of VGlut1, quantified as integrated density, by the mPFC neurons of control and CFA injected groups. Post-hoc analysis revealed that seven days of inflammatory pain caused difference in VGlut1 expression between Control males and CFA males, *** P < 0.001, Control females and CFA females, ** P < 0.01 and Control females and CFA males, **** P < 0.0001. Panel F shows the distribution of iGluSnFr-labeled neurons confined mostly to layers II and III of the PrL cortex. Roman numerals in panel F indicate cortical layers, layer VI is not included in the image. n = 4 controls and 6 CFA for both sexes. Data = Mean ± SEM. Scale bar = 10 μm for A to D and 200 μm for F.

Discussion

The study investigated the effects of inflammatory pain on the density of dendritic spines and glutamate signaling in the mPFC and correlated the pain-induced changes in spine density and expression of VGlut1 to deficit in working memory of mice. The main findings of the study include the detection of higher spine density on the mPFC neurons of female mice when compared males and the relative resistance of dendritic spines in female mice to changes under pain conditions. The pain-induced reduction of VGlut1expression by the mPFC neurons was also more pronounced in males. Furthermore, we found that only the correct but not the incorrect alternations on Y-maze are associated with glutamate release in the mPFC and that inflammatory pain impairs working memory only in male but not in female mice. Therefore, male mice experience more often bouts of glutamate deficiency related to the mPFC functions than female mice subjected to pain.

Dendritic spines are platforms for synaptic connections and their mailability by neuronal input is well known (Runge, Cardoso and de Chevigny, 2020). Release of both, glutamate and GABA in the vicinity of dendritic shaft triggers quick development of dendritic spines equipped with the corresponding synaptic scaffolds (Kwon and Sabatini, 2011; Oh et al., 2016). However, the remarkable plasticity of dendritic spines under physiological conditions makes them susceptible to changes under unfavorable conditions such as stress and pain. Animal studies show that chronic pain may increase or decrease dendritic length and spine density in cortical neurons and may lead to overactivation or deactivation of the mPFC (Kelly et al., 2016; Metz et al., 2009; Wu, Liang and Gao, 2016). The variability in the results depends on the pain model and duration of pain. In humans, the presumed mechanism responsible for cognitive impairment in patient with chronic pain is the loss of gray matter and thinning of the cortex (Kang et al., 2019). In our study we found an increase in the total dendritic spine density on the mPFC neurons, which is consistent with the report of increased excitability of these neurons in mice subjected to CFA injection (Wu, Liang and Gao, 2016). The mPFC receives dense afferents from the sensory cortex, thalamus and amygdala that convey nociceptive information (Jefferson, Kelly and Martina, 2021). While a longer lasting pain may inhibit synaptogenesis and spine formation, the increase of spine density in our experiment is very likely a result of increased nociceptive input to the mPFC driven by inflammatory pain with relatively short duration. Our results are supported by the published literature that show an increase of excitatory transmitter glutamate in the mPFC after pain conditions with short duration (Giordano et al., 2012).

The dendritic spine density was affected differently in male mice and female mice. Our explanation for the sex difference is that the total spine density was higher in the females even under control conditions which may dampen the effect of increased nociceptive input on synaptogenesis and development of new dendritic spines. The observed high spine density in female cortex recaptures data from published literature which shows dependence of the spine density on age and sex, with higher expression of dendritic spines in the mPFC neurons of female mice (Delevich et al., 2020). Furthermore, the density of mushroom spines, which are considered to be the most stable and fully functional spines (Runge, Cardoso and de Chevigny, 2020), was not affected at all in female mice. However, the pain increased the mushroom spine density on the dendrites of interneurons in males. Therefore, we suggest that one factor responsible for intact working memory in females with pain is the unaltered density and stability of mushroom spines on the dendrites of mPFC neurons. Interestingly, our findings parallel the recently described sex differences in the expression and activity of inhibitory mPFC neurons after nerve injury (Jones and Sheets, 2020; Shiers et al., 2018). Maladaptive sex differences affecting the mPFC interneurons are reported not only in pain models but also after chronic stress (Girgenti et al., 2019).

Working memory is distributive in nature or many brain areas are involved with various aspects of the memory processes such as sensory and spatial information, attention and outcome strategy (Christophel et al., 2017). The role of mPFC in working memory is to encode information about the influx of stimuli and to generate adequate action in response to the most salient stimuli. In other words, the mPFC formulates the decisions that guide behavior. In both human subjects and rodents, while the spatial information is encoded in the hippocampus, the completion of spatial tasks is guided by the mPFC (van Asselen et al., 2006; Euston, Gruber and McNaughton, 2012). These tasks are executed by neuronal assemblies in which single neurons are active during various aspects of the task. For example, electrophysiological recordings of the mPFC neurons in rats demonstrate that some neurons are active during different maze locations and some neurons fire around the time when the animal enters the maze arms (Yang and Mailman, 2018). The activity of single neuronal units was recaptured by field potential from neuronal groups that increased their activity around the choice that the animal made on the maze (Yang and Mailman, 2018). Here, we used a biosensor to detect glutamate signaling during similar working memory task in mice. While our photometry recordings lack the single cell resolution of the quoted above study, they demonstrate similar increased neuronal activity at the second before the mouse enters into a new arm. This observation is consistent with the reported increased firing rate at the moment of the animal choice (Yang and Mailman, 2018). However, the authors also report neuronal activity during wrong choices, while our recordings detected a robust glutamate release only before the correct entrances. The glutamate triggered fluorescence was much lower before incorrect entrances. A possible explanation for this discrepancy is that the photometry recordings demonstrate the dynamics of glutamate signaling in the neuronal populations under the fiberoptic cannula but not the actions potentials of the entire population of active neurons. It is possible that neuronal excitation that underlines wrong choices is generated by some of the numerous other neurotransmitters that regulate the neuronal activity in the mPFC. For example, optogenetic activation of parvalbumin and somatostatin neurons in the mPFC of rats worsens working memory performance (Kamigaki and Dan, 2017). The possible impairment of working memory by overactive interneurons in the mPFC very well matches our data that shows increased density of the functionally active mushroom spines on the interneurons only in male mice subjected to pain, the same group that underperformed on Y-maze test.

The likelihood of working memory dependence on sufficient glutamate release in the mPFC finds support in the changes of VGlut1 expression by the mPFC neurons in mice with inflammatory pain. The function of glutamate transporters is recycling of the glutamate spillover at the synaptic clefts of glutamatergic synapses and their expression is tightly linked to the level of glutamate signaling (Wojcik et al., 2004). While the photometric traces lack the necessary single cell resolution and did not detect a significant change of glutamate signaling after CFA injection, the expression of VGlut1 by the mPFC neurons was substantially reduced by the pain. The detected VGlut1 decrease was sex-dependent, a fact that supports our notion for glutamatergic deficiency as an underlying mechanism for memory impairment in male mice with inflammatory pain. The small but significant decrease of VGlut1 expression by the mPFC neurons is likely a harbinger for upcoming further decline of glutamate signaling in the cortex as the pain progresses, a process described in rodents (Guida et al., 2015) and in humans (Naylor et al., 2019; Kang et al., 2021).

In summary, inflammatory pain negatively affected working memory in male but not in female mice. The very likely mechanism for the memory impairment is the increased density of mushroom spines on the dendrites of interneurons in male mice, which may suppress glutamate signaling during working memory task. The ability of female cortical neurons to buffer nociceptive input by maintaining their spine density relatively constant prevents any drastic changes in the neuronal output of the mPFC as illustrated by normal execution of working memory task by female mice subjected to pain.

Highlights:

  1. Pain increases the cortical dendritic spine density in sex-dependent manner.

  2. Working memory depends on adequate glutamate release in the prefrontal cortex.

  3. Working memory is impaired only in male but not female mice subjected to pain.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Reference List:

  1. Bastos AM, Loonis R, Kornblith S, Lundqvist M and Miller EK (2018) ‘Laminar recordings in frontal cortex suggest distinct layers for maintenance and control of working memory’, Proc Natl Acad Sci U S A, 115(5), pp. 1117–1122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Benson CA, Fenrich KK, Olson KL, Patwa S, Bangalore L, Waxman SG and Tan AM (2020) ‘Dendritic Spine Dynamics after Peripheral Nerve Injury: An Intravital Structural Study’, J Neurosci, 40(22), pp. 4297–4308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bruno CA, O’Brien C, Bryant S, Mejaes JI, Estrin DJ, Pizzano C and Barker DJ (2021) ‘pMAT: An open-source software suite for the analysis of fiber photometry data’, Pharmacol Biochem Behav, 201, pp. 173093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cardoso-Cruz H, Lima D and Galhardo V (2013) ‘Impaired spatial memory performance in a rat model of neuropathic pain is associated with reduced hippocampus-prefrontal cortex connectivity’, J Neurosci, 33(6), pp. 2465–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chaplan SR, Bach FW, Pogrel JW, Chung JM and Yaksh TL (1994) ‘Quantitative assessment of tactile allodynia in the rat paw’, J Neurosci Methods, 53(1), pp. 55–63. [DOI] [PubMed] [Google Scholar]
  6. Christophel TB, Klink PC, Spitzer B, Roelfsema PR and Haynes JD (2017) ‘The Distributed Nature of Working Memory’, Trends Cogn Sci, 21(2), pp. 111–124. [DOI] [PubMed] [Google Scholar]
  7. Chung G, Kim CY, Yun YC, Yoon SH, Kim MH, Kim YK and Kim SJ (2017) ‘Upregulation of prefrontal metabotropic glutamate receptor 5 mediates neuropathic pain and negative mood symptoms after spinal nerve injury in rats’, Sci Rep, 7(1), pp. 9743. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Delevich K, Okada NJ, Rahane A, Zhang Z, Hall CD and Wilbrecht L (2020) ‘Sex and Pubertal Status Influence Dendritic Spine Density on Frontal Corticostriatal Projection Neurons in Mice’, Cereb Cortex, 30(6), pp. 3543–3557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dixon WJ (1991) ‘Staircase bioassay: the up-and-down method’, Neurosci Biobehav Rev, 15(1), pp. 47–50. [DOI] [PubMed] [Google Scholar]
  10. Euston DR, Gruber AJ and McNaughton BL (2012) ‘The role of medial prefrontal cortex in memory and decision making’, Neuron, 76(6), pp. 1057–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Giordano C, Cristino L, Luongo L, Siniscalco D, Petrosino S, Piscitelli F, Marabese I, Gatta L, Rossi F, Imperatore R, Palazzo E, de Novellis V, Di Marzo V and Maione S (2012) ‘TRPV1-dependent and -independent alterations in the limbic cortex of neuropathic mice: impact on glial caspases and pain perception’, Cereb Cortex, 22(11), pp. 2495–518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Girgenti MJ, Wohleb ES, Mehta S, Ghosal S, Fogaca MV and Duman RS (2019) ‘Prefrontal cortex interneurons display dynamic sex-specific stress-induced transcriptomes’, Transl Psychiatry, 9(1), pp. 292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Guida F, Luongo L, Marmo F, Romano R, Iannotta M, Napolitano F, Belardo C, Marabese I, D’Aniello A, De Gregorio D, Rossi F, Piscitelli F, Lattanzi R, de Bartolomeis A, Usiello A, Di Marzo V, de Novellis V and Maione S (2015) ‘Palmitoylethanolamide reduces pain-related behaviors and restores glutamatergic synapses homeostasis in the medial prefrontal cortex of neuropathic mice’, Mol Brain, 8, pp. 47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Huang J, Gadotti VM, Chen L, Souza IA, Huang S, Wang D, Ramakrishnan C, Deisseroth K, Zhang Z and Zamponi GW (2019) ‘A neuronal circuit for activating descending modulation of neuropathic pain’, Nat Neurosci, 22(10), pp. 1659–1668. [DOI] [PubMed] [Google Scholar]
  15. Hung KL, Wang SJ, Wang YC, Chiang TR and Wang CC (2014) ‘Upregulation of presynaptic proteins and protein kinases associated with enhanced glutamate release from axonal terminals (synaptosomes) of the medial prefrontal cortex in rats with neuropathic pain’, Pain, 155(2), pp. 377–387. [DOI] [PubMed] [Google Scholar]
  16. Jefferson T, Kelly CJ and Martina M (2021) ‘Differential Rearrangement of Excitatory Inputs to the Medial Prefrontal Cortex in Chronic Pain Models’, Front Neural Circuits, 15, pp. 791043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jiang ZC, Pan Q, Zheng C, Deng XF, Wang JY and Luo F (2014) ‘Inactivation of the prelimbic rather than infralimbic cortex impairs acquisition and expression of formalin-induced conditioned place avoidance’, Neurosci Lett, 569, pp. 89–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jones AF and Sheets PL (2020) ‘Sex-Specific Disruption of Distinct mPFC Inhibitory Neurons in Spared-Nerve Injury Model of Neuropathic Pain’, Cell Rep, 31(10), pp. 107729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kamigaki T and Dan Y (2017) ‘Delay activity of specific prefrontal interneuron subtypes modulates memory-guided behavior’, Nat Neurosci, 20(6), pp. 854–863. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kang D, Hesam-Shariati N, McAuley JH, Alam M, Trost Z, Rae CD and Gustin SM (2021) ‘Disruption to normal excitatory and inhibitory function within the medial prefrontal cortex in people with chronic pain’, Eur J Pain, 25(10), pp. 2242–2256. [DOI] [PubMed] [Google Scholar]
  21. Kang D, McAuley JH, Kassem MS, Gatt JM and Gustin SM (2019) ‘What does the grey matter decrease in the medial prefrontal cortex reflect in people with chronic pain?’, Eur J Pain, 23(2), pp. 203–219. [DOI] [PubMed] [Google Scholar]
  22. Kelly CJ, Huang M, Meltzer H and Martina M (2016) ‘Reduced Glutamatergic Currents and Dendritic Branching of Layer 5 Pyramidal Cells Contribute to Medial Prefrontal Cortex Deactivation in a Rat Model of Neuropathic Pain’, Front Cell Neurosci, 10, pp. 133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kraeuter AK, Guest PC and Sarnyai Z (2019) ‘The Y-Maze for Assessment of Spatial Working and Reference Memory in Mice’, Methods Mol Biol, 1916, pp. 105–111. [DOI] [PubMed] [Google Scholar]
  24. Kummer KK, Mitric M, Kalpachidou T and Kress M (2020) ‘The Medial Prefrontal Cortex as a Central Hub for Mental Comorbidities Associated with Chronic Pain’, Int J Mol Sci, 21(10). [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kwon HB and Sabatini BL (2011) ‘Glutamate induces de novo growth of functional spines in developing cortex’, Nature, 474(7349), pp. 100–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Latremoliere A and Woolf CJ (2009) ‘Central sensitization: a generator of pain hypersensitivity by central neural plasticity’, J Pain, 10(9), pp. 895–926. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Laubach M, Amarante LM, Swanson K and White SR (2018) ‘What, If Anything, Is Rodent Prefrontal Cortex?’, eNeuro, 5(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Martin L, Borckardt JJ, Reeves ST, Frohman H, Beam W, Nahas Z, Johnson K, Younger J, Madan A, Patterson D and George M (2013) ‘A pilot functional MRI study of the effects of prefrontal rTMS on pain perception’, Pain Med, 14(7), pp. 999–1009. [DOI] [PubMed] [Google Scholar]
  29. Marvin JS, Borghuis BG, Tian L, Cichon J, Harnett MT, Akerboom J, Gordus A, Renninger SL, Chen TW, Bargmann CI, Orger MB, Schreiter ER, Demb JB, Gan WB, Hires SA and Looger LL (2013) ‘An optimized fluorescent probe for visualizing glutamate neurotransmission’, Nat Methods, 10(2), pp. 162–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mazza S, Frot M and Rey AE (2018) ‘A comprehensive literature review of chronic pain and memory’, Prog Neuropsychopharmacol Biol Psychiatry, 87(Pt B), pp. 183–192. [DOI] [PubMed] [Google Scholar]
  31. Metz AE, Yau HJ, Centeno MV, Apkarian AV and Martina M (2009) ‘Morphological and functional reorganization of rat medial prefrontal cortex in neuropathic pain’, Proc Natl Acad Sci U S A, 106(7), pp. 2423–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Naylor B, Hesam-Shariati N, McAuley JH, Boag S, Newton-John T, Rae CD and Gustin SM (2019) ‘Reduced Glutamate in the Medial Prefrontal Cortex Is Associated With Emotional and Cognitive Dysregulation in People With Chronic Pain’, Front Neurol, 10, pp. 1110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Oh WC, Lutzu S, Castillo PE and Kwon HB (2016) ‘De novo synaptogenesis induced by GABA in the developing mouse cortex’, Science, 353(6303), pp. 1037–1040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Ong WY, Stohler CS and Herr DR (2019) ‘Role of the Prefrontal Cortex in Pain Processing’, Mol Neurobiol, 56(2), pp. 1137–1166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ren K and Dubner R (1999) ‘Inflammatory Models of Pain and Hyperalgesia’, ILAR J, 40(3), pp. 111–118. [DOI] [PubMed] [Google Scholar]
  36. Risher WC, Ustunkaya T, Singh Alvarado J and Eroglu C (2014) ‘Rapid Golgi analysis method for efficient and unbiased classification of dendritic spines’, PLoS One, 9(9), pp. e107591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Runge K, Cardoso C and de Chevigny A (2020) ‘Dendritic Spine Plasticity: Function and Mechanisms’, Front Synaptic Neurosci, 12, pp. 36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Seamans JK, Lapish CC and Durstewitz D (2008) ‘Comparing the prefrontal cortex of rats and primates: insights from electrophysiology’, Neurotox Res, 14(2–3), pp. 249–62. [DOI] [PubMed] [Google Scholar]
  39. Shiers S, Pradhan G, Mwirigi J, Mejia G, Ahmad A, Kroener S and Price T (2018) ‘Neuropathic Pain Creates an Enduring Prefrontal Cortex Dysfunction Corrected by the Type II Diabetic Drug Metformin But Not by Gabapentin’, J Neurosci, 38(33), pp. 7337–7350. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Stratton HJ and Khanna R (2020) ‘Sculpting Dendritic Spines during Initiation and Maintenance of Neuropathic Pain’, J Neurosci, 40(40), pp. 7578–7589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Tan LL, Oswald MJ, Heinl C, Retana Romero OA, Kaushalya SK, Monyer H and Kuner R (2019) ‘Gamma oscillations in somatosensory cortex recruit prefrontal and descending serotonergic pathways in aversion and nociception’, Nat Commun, 10(1), pp. 983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tian Y, Yang C, Cui Y, Su F, Wang Y, Wang Y, Yuan P, Shang S, Li H, Zhao J, Zhu D, Tang S, Cao P, Liu Y, Wang X, Wang L, Zeng W, Jiang H, Zhao F, Luo M, Xiong W, Qiu Z, Li XY and Zhang C (2018) ‘An Excitatory Neural Assembly Encodes Short-Term Memory in the Prefrontal Cortex’, Cell Rep, 22(7), pp. 1734–1744. [DOI] [PubMed] [Google Scholar]
  43. Tyrtyshnaia A and Manzhulo I (2020) ‘Neuropathic Pain Causes Memory Deficits and Dendrite Tree Morphology Changes in Mouse Hippocampus’, J Pain Res, 13, pp. 345–354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. van Asselen M, Kessels RP, Neggers SF, Kappelle LJ, Frijns CJ and Postma A (2006) ‘Brain areas involved in spatial working memory’, Neuropsychologia, 44(7), pp. 1185–94. [DOI] [PubMed] [Google Scholar]
  45. Van De Werd HJ, Rajkowska G, Evers P and Uylings HB (2010) ‘Cytoarchitectonic and chemoarchitectonic characterization of the prefrontal cortical areas in the mouse’, Brain Struct Funct, 214(4), pp. 339–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Wei F and Zhuo M (2001) ‘Potentiation of sensory responses in the anterior cingulate cortex following digit amputation in the anaesthetised rat’, J Physiol, 532(Pt 3), pp. 823–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wojcik SM, Rhee JS, Herzog E, Sigler A, Jahn R, Takamori S, Brose N and Rosenmund C (2004) ‘An essential role for vesicular glutamate transporter 1 (VGLUT1) in postnatal development and control of quantal size’, Proc Natl Acad Sci U S A, 101(18), pp. 7158–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Woodcock EA, Anand C, Khatib D, Diwadkar VA and Stanley JA (2018) ‘Working Memory Modulates Glutamate Levels in the Dorsolateral Prefrontal Cortex during (1)H fMRS’, Front Psychiatry, 9, pp. 66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wu XB, Liang B and Gao YJ (2016) ‘The increase of intrinsic excitability of layer V pyramidal cells in the prelimbic medial prefrontal cortex of adult mice after peripheral inflammation’, Neurosci Lett, 611, pp. 40–5. [DOI] [PubMed] [Google Scholar]
  50. Yang Y and Mailman RB (2018) ‘Strategic neuronal encoding in medial prefrontal cortex of spatial working memory in the T-maze’, Behav Brain Res, 343, pp. 50–60. [DOI] [PubMed] [Google Scholar]

RESOURCES