Sustained attention performance deficits in the three-choice serial reaction time task in male and female rats after experimental brain trauma

Abstract

Impaired attention is central to the cognitive deficits associated with long-term sequelae for many traumatic brain injury (TBI) survivors. Assessing complex sustained attention post-TBI is clinically-relevant and may provide reliable avenues towards developing therapeutic and rehabilitation targets in both males and females. We hypothesized that rats subjected to a moderate TBI will exhibit attentional deficits seen as reduced accuracy and increased distractibility in an operant 3-choice serial reaction time task (3-CSRT), designed as an analogue of the clinical continuous performance test. Upon reaching baseline of 70% accuracy at the 300 ms cue, adult male and female Sprague-Dawley rats were subjected to a controlled cortical impact (2.8 mm deformation at 4 m/s) or sham injury over the right parietal cortex. After two weeks of recovery, they were retested on the 3-CSRT for ten days. Dependent measures include percent accuracy (overall and for each of the three cue ports), percent omissions, as well as latency to instrumental poke and retrieve reward. Results demonstrate that both males and females displayed reduced percent accuracy and increased omissions when re-tested post-TBI on 3-CSRT compared to Sham rats and to their own pre-insult baseline (p’s < 0.05). Performance accuracy was impaired consistently throughout the ten days of post-surgery re-testing, suggesting pronounced and long-lasting dysfunction in sustained attention processes. Deficits were specifically more pronounced when the cue was pseudorandomly presented in the left-side cue port (p < 0.05), mirroring clinical hemispatial neglect. These data demonstrate significant and persistent complex attention impairments in both sexes after TBI, rendering identifying efficient therapies for cognitive recovery as pivotal.

1. Introduction

Globally, traumatic brain injury (TBI) is a leading cause of death and disability (Hyder et al., 2007). Despite the precipitating event being vastly heterogeneous, a majority of moderate to severe TBI patients experience varying features of cognitive dysfunction, which impedes personal and economic success (Rabinowitz and Levin, 2014). Attentional dysregulation is frequent in TBI survivors and can persist chronically, lending in part to its debilitating nature as the capacity to sustain attention is essential for complex cognitive functions (Arcieniegas et al., 1999; Arcieniegas et al., 2002; Zoccolotti et al., 2000; Ben-David et al., 2011; Barman et al., 2016). Attentional and memory deficits are reported in nearly 30% of mild TBI patients (de Freitas Cardoso et al., 2019) and up to 65% of moderate or severe TBI survivors (Whiteneck et al., 2004). TBI-induced disability may also preclude rehabilitative strategies (Barman et al., 2016) and exacerbate attention-related comorbidities like attention-deficit hyperactivity disorder (ADHD) (Biederman et al., 2015; Asarnow et al., 2021). Sustained attention, often categorized as the basic attentional function that precipitates divided and selective attention refers to the ability to maintain goal-directed behavior, typically a relevant stimulus over an extended period (Robbins, 2002; Sarter et al., 2006). Clinical studies utilizing various neuropsychological assessments of attention, such as the continuous performance task (CPT) or the Wisconsin Card Sorting test (WCST), reveal impairments in sustained and divided, as well as flexible attention, respectively, across all levels of injury severity that can persist years following TBI (Ord et al., 2010a; Ord et al., 2010b; Zane et al., 2016; Zhao et al., 2018). Moreover, moderate-to-severe and severe TBI resulted in more pronounced deficits of executive function in the WCST (Ord et al., 2010a; Ord et al., 2010b) and sustained attention in the CPT (Riccio et al., 2002) compared to mild injury patients.

Despite the clinical pervasiveness of TBI-induced attentional impairments, experimental TBI has mainly focused on impairments to hippocampal-dependent learning and memory, utilizing tasks such as the Morris water maze (MWM). Empirical assessment of higher-order cognitive dysfunction remains understudied and is of utmost importance.

The application of complex behavioral paradigms that assess clinically relevant symptomatology may elucidate the pathogenic mechanisms that underlie, in part, impairments to attention following TBI. Our laboratory previously demonstrated that a controlled cortical impact (CCI) produced injury severity-dependent impairments to cognitive flexibility, a form of executive function assessed by the attentional set shifting test (AST; Bondi et al., 2014; Njoku et al., 2019; Minchew et al., 2021), which is analogous to the clinical WCST (Stuss et al., 2000). Such studies support the pre-clinical utility and sensitivity of complex behavioral assessments mediated largely by the prefrontal cortex to examine clinically relevant cognitive sequelae in models of TBI. Furthermore, biological sex has been implicated in hippocampal-dependent tasks (Yagi and Galea, 2019), attentional processes, and etiology of neuropsychological disorders like ADHD and schizophrenia (Wickens et al., 2018). It is well established that female TBI patients report increased recovery time and symptom severity suggesting the possible influence of sex on recovery (Gupte et al., 2019). In most observational and clinical studies of TBI, women represent approximately 30% of subjects (Biegon, 2021), a finding that warrants both sexes should undergo stringent cognitive assessments after TBI (Bruns and Hauser, 2003).

To evaluate whether parietal cortex injury induces deficits to visuospatial sustained attentional functioning in adult male and female rats, we employed the well-validated 3-Choice Serial Reaction Time (3-CSRT; Fig. 1) Task, a modified version of the 5-Choice Serial Reaction Time (5-CSRT) Task and rodent analog of the clinical CPT (Carli et al., 1983; Robbins, 2002; McGaughy et al., 2002; Dalley et al., 2004; Totah et al., 2009). We hypothesized that both male and female rats subjected to a moderate TBI would exhibit chronic reductions in sustained attention and increased distractibility relative to their baseline (i.e., pre-injury) performance. Understanding the extent of impairment to complex cognitive functioning after TBI in both sexes may provide insight into the neural correlates of visuospatial attention and contribute to the development of targeted therapeutic interventions.

Fig. 1.

The three-choice serial reaction time (3-CSRT) task. During each 30 min session, one of three cue port locations was pseudorandomly illuminated for 300 ms after an 8 s ITI. The rat had 5 s to respond into one of the cue holes (Totah et al., 2009). A correct response (A) resulted in sugar pellet reward delivery into the food magazine on the wall opposite from the cue holes. An incorrect response (B) or omission of response altogether (C) led to house light being turned off for 5 s. After an error, a new trial was initiated when the rat nose poked into the empty food magazine.

2. Results

2.1. Health and wellness

There were no health issues with the rats as they all gained weight over time. No rats were eliminated after surgery, hence the final statistical analyses were performed on 19 rats.

2.2. 3-CRST training

Thirteen adult male Sprague Dawley rats reached criterion performance at the testing level (i.e., 300 ms cue duration) after 107.85 ± 15.5 training sessions. At the 15 s, 5 s, 2 s, and 1 s cue durations, it took 5.61 ± 1.23, 4.15 ± 0.37, 10.07 ± 2.79, and 34.15 ± 6.85 training sessions, respectively, for the male rats to meet criterion performance (i.e., >80% accuracy and < 20% omissions). At the 500 ms cue duration, male rats reached criterion performance (i.e., > 75% accuracy, < 20% omissions) after 30.84 ± 8.90 training sessions, whereas at the 300 ms cue duration, the male rats met criterion performance (i.e., > 70% accuracy and < 20% omissions) after 23 ± 7.93 training sessions. Fig. 2 A, C, E display percent accuracy, percent omissions, and total trials per session, respectively, for male rats, suggesting the criterion performance was met and the task was learned appropriately prior to surgery. Individual data points are presented on the graph accordingly. No statistical analyses were necessary prior to group designations.

Fig. 2.

Training on 3-CSRT for adult male (left panels: A, C, E) and female (right panels: B, D, F) Sprague Dawley rats. Rats (n = 19) were trained in the task using initial cue duration of 15 s. Cue duration decreased at each training level until rats met performance criteria using a 300 ms cue duration (see Materials and Methods). Individual data points are presented on each column accordingly. A, B. Mean (±S.E.M.) percentage accuracy, the primary measure of sustained attention, at each training level. Accuracy decreased slightly when stages became increasingly difficult with shorter cues, however the patterns were nearly identical in both sexes and baseline performance remained above the 70% cutoff. C, D. Mean (±S.E.M.) percentage omissions at each training level, which remained well below the 20% criterion requirement. E, F. Mean (±S.E.M.) number of total trials per session at each training level.

Six adult female Sprague Dawley rats reached criterion performance at the testing level (i.e., 300 ms cue duration) after 186.66 ± 13.4 training sessions. The female rat training group met criterion performance at the 15 s, 5 s, 2 s, and 1 s cue duration after 6.5 ± 1.17, 4.33 ± 0.95, 20 ± 11.89, and 129.66 ± 6.23 training sessions, respectively. Female rats reached criterion performance at the 500 ms cue duration after 8.66 ± 1.08 training sessions. At the 300 ms cue duration level, female rats met criterion after 17.5 ± 4.55 training sessions. Fig. 2 B, D, F display percent accuracy, percent omissions, and total trials per session, respectively, for female rats, suggesting the criterion performance was met and the task was learned appropriately prior to surgery. Individual data points are presented on the graph accordingly.

2.3. Acute neurological assessments

Statistically significant differences between TBI and Sham rats in hindlimb reflex paw withdrawal latency were observed for left (TBI male range = 165.86 ± 5.69 s, Sham male range = 21.83 ± 1.78 s, TBI female range = 177 ± 15.88 s, Sham female range = 24.33 ± 1.2 s; F_3,15 = 171.32, p < 0.001), and right (TBI male range = 160.43 ± 5.52 s, Sham male range = 16 ± 1.57 s, TBI female range = 171 ± 16.86 s, Sham female range = 19.67 ± 1.33 s; F_3,15 = 165.71, p < 0.001) sides after cessation of anesthesia. Newman Keuls post hoc tests rendered significant effects between both male and female TBI versus their respective Sham groups (p’s < 0.05).

Significant differences in righting reflex were also observed between TBI and Sham rats for return to righting ability (TBI male range = 379.71 ± 34.15 s, Sham male range = 128.82 ± 7.93 s, TBI female range = 466.33 ± 5.81 s, Sham female range = 135 ± 6.66 s; F_3,15 = 36.39, p < 0.001). Newman Keuls post hoc tests rendered significant effects between both male and female TBI versus their respective Sham groups (p’s < 0.05).

2.4. Post-injury 3-CSRT testing

2.4.1. Percent accuracy

Fig. 3A shows the effects of injury, sex, and test session on percent accuracy in the 3-CSRT at days 14–24 post-surgery (Sham males: n = 6, Sham females: n = 3; TBI males: n = 7; TBI females: n = 3). Repeated measures ANOVA revealed that parietal TBI induced significant and persistent performance deficits seen as reduced percent accuracy (i.e., impaired sustained attention) compared to Sham groups and baseline levels (Group effect: F_3,15 = 17.39, p < 0.001; Day effect: F_10,150 = 18.33, p < 0.001; Group × Day interaction: F_30,150 = 2.63, p < 0.001). Moreover, Newman Keuls post hoc analyses showed that both male and female TBI rats performed significantly worse than their Sham counterparts (p’s < 0.05), but there were no differences between males and females in either TBI or Sham conditions (p’s > 0.05). Impairments in sustained attention occurred without detrimental effects on motivation to perform reflected in total trials. Both Sham groups maintained total trials at > 150 per session, similar to pre-surgery baselines (Fig. 2 E,F), while TBI groups maintained total trials throughout testing at > 100 per session, well above the 10% requirement of Sham levels for this task (data not shown).

Fig. 3.

Mean (±S.E.M.) percent accuracy (A) and percent omissions (B) in the 3-CSRT during post-surgery testing for both males and females. Parietal TBI induced significant and persistent performance deficits seen as reduced percent accuracy (i.e., impaired sustained attention) (A) and increased omissions (i.e., increased distractibility) (B) in both male and female rats compared to their respective Sham groups and baseline levels (*p’s < 0.05). However, there were no differences between males and females in either TBI or Sham conditions (p’s > 0.05), n = 6–7/group (males); 3/group (females).

2.4.2. Percent omissions

Fig. 3B displays the effects of injury, sex, and test session on percent omissions in the 3-CSRT at days 14–24 post-surgery. Repeated measures ANOVA determined that parietal TBI induced significant and persistent performance deficits seen as increased percent omissions (i.e., increased distractibility) compared to Sham groups and baseline levels. There was no significant Group effect (F_3,15 = 17.39, p = 0.17), but there was a significant Day effect (F_10,150 = 8.29, p < 0.001) and significant Group × Day interaction (F_30,150 = 1.61, p < 0.05). Subsequently, Newman Keuls post hoc analyses showed that both male and female TBI rats performed significantly worse than their Sham counterparts (p’s < 0.05), but there were no differences between males and females in either TBI or Sham conditions (p’s > 0.05).

2.4.3. Percent accuracy - individual cue port breakdown

In order to appreciate injury-induced deficits in goal-directed behavior and sustained attention within specific sections of the visual field in the operant chambers, percent accuracy was calculated for each of the three cue ports where lights were pseudorandomly presented: left (L), center (C), and right (R). Fig. 4 A–C shows the effect of injury, sex, and test session on percent accuracy in each of the three cue ports. Specifically, for performance related cue presentation in the L nose port, repeated measures ANOVA determined that parietal TBI induced significant and persistent reductions in percent accuracy compared to Sham groups and baseline levels (Fig. 4A). There were significant Group (F_3,15 = 5.77, p < 0.01) and Day (F_10,150 = 3.11, p < 0.01) effects, but no significant Group × Day interaction (F_30,150 = 1.37, p = 0.11). Subsequently, Newman Keuls post hoc analyses showed that both male and female TBI rats performed significantly worse than their Sham counterparts in the L cue port (p’s < 0.05), but there were no differences between males and females in either TBI or Sham conditions (p’s > 0.05). Moreover, performance deficits were specifically most pronounced when the cue was presented in the left-side cue port compared to the others (i.e., percent accuracy dropping to nearly 30% on first day of post-injury testing). For performance related to cue presentation in the C port (Fig. 4B), ANOVA rendered significant effects of Group (F_3,15 = 8.91, p < 0.01) and Day (F_10,150 = 3.06, p < 0.01), as well as a significant Group × Day interaction (F_30,150 = 1.65, p < 0.05). Newman Keuls post hoc analyses revealed TBI groups performed significantly worse than their Sham counterparts (p’s < 0.05), while male rats displayed significantly lower percent accuracy than females in both TBI and Sham conditions (p’s < 0.05). For performance related to cue presentation in the R port (Fig. 4C), repeated measures ANOVA rendered significant effects of Group (F_3,15 = 5.03, p < 0.05) and Day (F_10,150 = 2.67, p < 0.01), albeit no significant Group × Day interaction (F_30,150 = 1.33, p = 0.14). Newman Keuls post hoc analyses again revealed TBI groups performed significantly worse than their Sham counterparts (p’s < 0.05), while male rats displayed significantly lower percent accuracy in the R cue port than females in both TBI and Sham conditions (p’s < 0.05).

Fig. 4.

Mean (±S.E.M.) percent accuracy for individual cue-ports: left (A), center (B), and right (C). TBI rendered significant and lasting deficits in sustained attention in both males and females within each cue port, with performance impairments being most pronounced when the cue was presented in the left-side hole (i.e., accuracy dropping to nearly 30% on first day of post-injury testing), mirroring clinical hemispatial neglect. Newman Keuls post hoc analyses revealed TBI groups performed significantly worse than their Sham counterparts in all cue-ports (*p’s < 0.05), while male rats displayed significantly lower percent accuracy than females in both TBI and Sham conditions for the center (B) and right (C) cue holes (^#p’s < 0.05), which could be attributed to higher female sham performance levels than expected.

2.4.4. Latency to choice

Motivation to perform in the 3-CSRT was further assessed by the latency to choice (i.e., instrumental poke) in both male and female rats. Fig. 5A shows the effects of injury, sex, and test session on latency to choice in the 3-CSRT at days 14–24 post-surgery. Repeated measures ANOVA revealed that parietal TBI induced significant elevations in latency to choice (Group effect: F_3,15 = 3.86, p < 0.05; Day effect: F_9,135 = 13.52, p < 0.001; Group × Day interaction: F_27,135 = 1.69, p < 0.05). Newman Keuls post hoc analyses showed that both male and female TBI rats took longer to engage in poking than their Sham counterparts (p’s < 0.05), and there were no differences between males and females in either TBI or Sham conditions (p’s > 0.05). Nevertheless, these elevations did not reflect overall motivation level alterations, as TBI rats engaged in poking with an approximate 0.5–1 s delay versus Sham rats, whose typical latency to choice is 1 s or less, and total trials remained at > 100 per session after injury as well.

Fig. 5.

Mean (±S.E.M.) latency to choice (s) (A) and latency to pellet consumption (s) (B) for adult male and female rats following TBI. A. Injury rendered subtle but significant elevations in latency to operant poking in both sexes (*p’s < 0.05), albeit the approximate 0.5–1 s elevation during the first several days of post-surgery testing did not influence overall goal-directed behavior and motivation to perform on the task. There were no differences between males and females in either TBI or Sham conditions (p’s > 0.05). B. There were no injury- or sex- related alterations in latency to sugar pellet consumption following correct responses (p’s > 0.05).

2.4.5. Latency to pellet consumption

Motivation to perform in the 3-CSRT was also assessed by the latency to pellet consumption in both male and female rats following correct responses. Fig. 5B displayed the effects of injury, sex, and test session on latency to pellet consumption at days 14–24 post-surgery (i.e., testing days 1–10). Repeated measures ANOVA revealed that there were no injury-related alterations in latency to consume the sugar pellet (Group: F_3,15 = 2.53, p = 0.10; Day: F_9,135 = 1.29, p = 0.25; Group × Day interaction: F_27,135 = 0.72, p = 0.84) for either male or female rats.

3. Discussion

The goal of the study was to characterize the 3-CSRT as a sensitive and reliable assessment tool of higher-order attentional functioning, specifically, measures of sustained attention and distractibility, after moderate TBI in adult male and female rats. The 3-CSRT, a modified version of the five-choice serial reaction time task, which was designed as an analogue of the human continuous performance test, requires subjects to divide attention between three nose-poke holes in an operant chamber, in which incrementally decreasing cues are randomly presented, to obtain a sucrose pellet reward. Notably, our paradigm utilized a pre-injury training period, establishing a reliable baseline comparison to post-injury attentional performance. The data showed that the 3-CSRT was sensitive to measures of attention in both male and female rats after TBI, as both sexes displayed decreased percent accuracy and increased percent omissions compared to their baseline and Sham controls, suggesting impaired sustained attention and distractibility, respectively. Furthermore, sustained attention deficits were specifically more pronounced when the cue was pseudorandomly presented in the left-side cue port (Fig. 4A), mirroring clinical hemispatial neglect following right hemisphere injury. Motivation to perform the task remained high over the ten-day post-injury testing period, as evidenced by no alterations in latency to consume the sugar pellet reward, engaging in significant numbers of total trials (>100), and despite a small but significant injury-induced elevation in latency to poke (approx. 0.5–1 s per trial).

In our study, we report that a moderate CCI leads to significant and lasting deficits in goal-directed behavior, seen as reduced accuracy throughout the ten days of testing post-injury, and increased omissions, the latter being more pronounced during the first days of testing and returning towards baseline in the second half of the testing period, while accuracy remained dramatically reduced in both male and female adult rats. Additional testing time points beyond one month post-injury are thus of interest and warrant further exploration. Future studies should also investigate effects of severe TBI on sustained attention. Based on clinical studies (Ord et al., 2010a; Ord et al., 2010b; Riccio et al., 2002), severe injury will likely lead to more pronounced deficits, albeit we also demonstrated that, in the attentional set-shifting test, moderate TBI (2.8 mm cortical deformation depth) resulted in slightly more robust impairments than severe TBI (3.0 mm depth) in male rats, possibly due to cross-hemispheric compensatory recruitment by colossal interactions causing interference effects through partially functional brain regions ipsilateral to injury, such as the hippocampus (Bondi et al., 2014).

In the current study, the female group experiment rendered less females completing training than males, due to the fact that females required an average 186.66 ± 13.4 training sessions, whereas males required 107.85 ± 15.5 training sessions to reach surgery. The rate-limiting step for females seemed to be level 4 (i.e., 1 s cue), where females required 129.66 ± 6.23 training sessions to reach criterion versus 34.15 ± 6.85 for males. Whether the delay was seasonal or pertaining to adjusting to the recent laboratory set-up, behavioral parameters (i.e., accuracy, omissions, total trials) did not differ between males and females once criterion was achieved (see Fig. 2). Previous studies employing 3-CSRT have successfully determined that n = 3/group in various experiments suffices for statistical power given the high performance stability and low variability (Pehrson et al., 2013). Meanwhile, we determined that TBI-induced deficits occur similarly when performing surgery on rats trained on an abbreviated version of the 3-CSRT in which rats are injured after reaching criterion at level 3 (i.e., 2 s cue), thus ongoing and future experiments will employ the abbreviated version of the task for efficiency purposes (Rennerfeldt et al., 2022).

Herein, we utilized 3 light-cue apertures to identify the attentional functioning of the left, center, and right visual fields following a unilateral, right hemisphere TBI. The findings suggest that both male and female rats exhibited decreased cue-port accuracy in all three spatial presentations, but it was more pronounced on the left visual field, indicative of unilateral spatial neglect of the visual field contralateral to the site of injury. This injury outcome is often observed clinically in TBI and stroke patients and strongly impedes a positive functional recovery trajectory (Posner et al., 1984; Chen et al., 2016; Di Gregorio et al., 2021). To the best of our knowledge, this is the first report of hemispatial neglect impeding complex cognitive function in a rodent unilateral CCI model and represents a captivating finding that warrants further investigation. Sensory neglect per se following bilateral CCI of the mPFC in rats has been previously characterized utilizing the “adhesive dot test” (Shear et al. 2002; Cutler et al. 2006). In both studies, sensory neglect was measured by latency to remove an adhesive dot from the radial wrist of the forepaw, such that TBI rats were significantly impaired vs sham, and persisted up to 20 days post-injury, suggesting potential long-term impairments. It is thus critical to consider that hemispatial neglect combined with a fast-paced, high-demand cognitive task, may lead to combined deficits in higher-order attentional processing after TBI.

When assessing percent accuracy in the center and right cue ports, not only were injured males and females impaired in comparison to their respective Sham counterparts, but TBI and Sham males both performed at lower accuracy levels than TBI and Sham females, respectively. This effect does not appear to be due to males performing poorly compared to females, but rather to females performing at higher levels than the overall average (i.e., >80% for Sham females in center versus typical 70–72% in males and females prior to surgery). Moreover, previous studies utilizing motor function and spatial learning also reported no sex differences following TBI (Bondi et al., 2015; Free et al., 2017).

Albeit limited, pre-clinical assessments of post-injury attentional impairments have been previously reported utilizing the 5-CSRT, a task analogous to the 3-CSRT, which uses 5 light-cue apertures in addition to alternative injury and behavioral paradigms (e.g., post-injury training). In overall support of our findings, Vonder Haar and colleagues (2016) reported that subgroups of adult male rats deemed to be “vulnerable” or “chronically impaired” when subjected to moderate-severe bilateral frontal CCI injury displayed early (i.e., 2–5 weeks post-injury) and late (i.e., 5–14 weeks post-injury) deficits in measures of sustained attention (i.e., percent accuracy) and distractibility (i.e., percent omissions) as assessed by the 5-CSRT at the 0.5 s cue duration (Vonder Haar et al., 2016). It remained unknown whether biological sex contributed to differences in disruptions of attentional functioning following adulthood TBI. However, using an alternative 5-CSRT paradigm where training occurred after mild adolescent TBI, Kaukas and colleagues (2021) revealed sex-specific impairments such that female rats exhibited deficits to sustained attention and distractibility during challenging cue durations (i.e., 1 s) compared to male rats during mid-adolescence. However, the authors tempered their findings as the overall power to detect sex differences was low in the trials with decreased stimulus duration. Previous studies utilizing the 5-CSRT demonstrate impaired task acquisition in non-injured adult female rats as task difficulty increases, for example, when the stimulus duration is decreased or a variable ITI is presented compared to male rats (Bayless et al., 2012). Our study found that during 3-CSRT training, the female rats exhibited a higher number of training sessions at the 2 s cue level to meet criterion performance compared to the male rats, even though once criterion was reached, performance parameters were nearly identical. Due to the prolonged 3-CSRT training period for the female rats, we were limited to a small sample size, which was still appropriately powered to detect similar deficits as in males post-injury.

In the current study, we have identified the 3-CRST as a novel method of characterizing attentional deficits, particularly sustained attention and distractibility, in male and female adult rats following experimental TBI. We also identified that sub-analysis of 3-CSRT results on a cue-port-specific basis may be beneficial for studying hemispatial neglect, another common post-TBI sequela. The 3-CSRT brings major advantages, including precise timing and standardization of behavior, and allows for the study of motor planning, decision-making, and multisensory integration with a level of temporal precision more familiar to primate studies (Whitlock, 2014). The task utilized the most attentionally taxing stimulus (i.e., 0.3 msec) to evaluate post-injury performance. Furthermore, impairment to higher degrees of attentional processing may also be transferrable to other aspects of attention, such as those involved in cognitive flexibility, where shifting attention to alternating stimuli requires increased attentional effort than attending to a single salient stimulus. Specifically, our laboratory has demonstrated severity-dependent alterations in cognitive flexibility as assessed by the attentional set shifting task (AST) (Bondi et al., 2014), as well as attenuation of injury-related behavioral flexibility alterations following chronic treatment with cholinergic (Njoku et al., 2019) and serotonergic (Minchew et al., 2021) compounds. Therefore, future studies will apply 3-CSRT in conjunction with AST to examine both bottom-up processes in which salient stimuli evoke attentional effort and top-down processes that require intentional allocation of attention, respectively. A multimodal behavioral approach will elucidate the facets of attentional processing that are impaired following injury and ultimately establish pre-clinical assessments that parallel the complexity of cognitive dysfunction following human TBI, ultimately aiding in the development of reliable and translatable therapeutic interventions.

4. Materials and Methods

4.1. Subjects and housing

Thirteen age-matched (3 months old) adult male (n = 13; 300–325 g) and six normal-cycling females (n = 6; 260–290 g) Sprague-Dawley rats (Envigo RMS, Inc., Indianapolis, IN) were utilized in this study. Rats were pair-housed in standard laboratory ventilated polycarbonate cages (37×25×18 cm) in a temperature- (21 ± 1C°) and light- (from 7:00 a.m. to 7:00p.m.) controlled vivarium with ad libitum food and water. During operant task training and testing, the rats were placed on mild food restriction to maintain 80% of their body weight (13 g/day/per rat) but had free access to water. The rats were randomly assigned to the following groups: Male, TBI (n = 7) and Sham controls (n = 6); Female, TBI (n = 3) and Sham controls (n = 3). All procedures were approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh. Every attempt was made to limit the number of animals and minimize suffering.

4.2. The 3-Choice serial reaction time task (3-CSRT)

The 3-CSRT is based on the well-validated 5-CSRT (Carli et al., 1983; Robbins, 2002; Totah et al., 2009). The 3-CSRT was performed in operant chambers (Harvard Apparatus, Holliston, MA) equipped with 3 LED illuminated light cue nose-poke apertures on one wall, an LED illuminated food magazine on the opposite wall to the cue holes, and a house light on the ceiling. Before training per se, rats first underwent a 20-min habituation session in which 10 sugar pellets were dispensed during the initial 10 s from the automated food magazine (Dustless Precision Pellets ^®, Bio-Serv, Flemington, NJ) into the food trough. The following day, rats underwent a 30-min pellet training session in which sugar pellets were dispensed every 30-s to accustom rats to pellet dispensing.

Rats were trained in six successive attentional levels in which the duration of pseudorandomly presented light cue gradually decreased (e.g., 15 s, 5 s, 2 s, 1 s, 500 ms, and 300 ms; Totah et al., 2009). Photosensors detected nose-pokes in the light-cue apertures and the food through. Correct responses, consisting of a nose-poke in the illuminated aperture, were rewarded by one sucrose pellet. Incorrect responses consisted of nose-poke responses to non-illuminated cue holes, whereas trial omissions occurred when rats neglected to respond within 5 s post-stimulus presentation. Incorrect responses and omissions were penalized by a “time-out” period in which the house light was turned off for 5 s, which could be shortened if the rat poked into the empty food magazine to start the next trial. Before the next cue presentation, an 8 s inter-trial interval (ITI) was presented (Fig. 1 presents a schematic of the task layout).

Rats advanced through training levels after meeting specific performance criteria. The first 4 training levels (e.g., 15 s, 5 s, 2 s, 1 s) required > 80% accuracy and < 20% omissions for 3 consecutive sessions. At the 500 ms level, rats were required to maintain > 75% accuracy for 6 consecutive sessions and < 20% omissions. The rats were considered ready for surgery once they met criterion performance of > 70% accuracy and < 20% omissions for 6 consecutive sessions at the 300 ms level, which was considered baseline performance utilized in post-injury behavior analysis. A criterion of minimal performance was set a priori consisting of completed choice trials (correct and incorrect) of at least 10 % of the total trials completed by the vehicle group in order to obtain meaningful accuracy and omissions measures in the task. No rats failed the minimum criterion performance. Dependent measures included percent accuracy for sustained attention [(# of correct responses / # total responses)*100]), percent omissions for distractibility [(# of incorrect responses / # total responses)*100]), as well as latencies to instrumental poking and pellet consumption to monitor motivation levels. Graphic State Notation 4^™ software was used to generate experiment control protocols and record responses.

4.3. Surgery

Upon meeting criterion performance at the testing cue duration (male n = 13; female n = 6), rats were randomly assigned to either a controlled cortical impact (CCI) or sham injury as previously described (Kline et al., 2002, 2004, 2010; Njoku et al., 2019; Minchew et al., 2021). Briefly, surgical anesthesia was induced and maintained with 4% and 2% isoflurane, respectively, in 2:1 N₂O:O₂. The rats were secured in a stereotaxic frame, and under aseptic conditions, a midline scalp incision was made. The skin and fascia were reflected to expose the skull. A craniectomy, encompassing bregma and lambda and the sagittal and coronal sutures, was made in the right hemisphere with a dental drill. The impacting rod was extended, and the impact tip (6 mm, flat) was centered and lowered through the craniectomy until it touched the dura mater. Then the rod was retracted, and the impact tip was advanced 2.8 mm farther to produce a moderate TBI (2.8 mm tissue deformation at 4 m/sec). Core body temperature was monitored and maintained at 37 ± 0.5 °C with a heating blanket. Immediately after the impact, anesthesia was discontinued and the incision was sutured. Except for the impact, sham rats underwent all surgical procedures.

4.4. Acute neurological assessments

After cessation of anesthesia, the rats were assessed for acute neurological outcome. Briefly, limb reflex ability was determined by gently squeezing the left and right hind paw every 5 s and recording the time to provoke a withdrawal response. Righting reflex was assessed by measuring the time required to turn from the supine to prone position three consecutive trials.

4.5. Post-injury 3-CSRT testing

Following CCI or sham injury, rats were allowed a total recovery period of 14 days, with mild food restriction reinstated on post-injury day 7. 3-CSRT testing at the 300 ms light cue duration was initiated as described above and continued for 10 days with one-day break at the mid point. Following the completion of behavior, rats were anesthetized with Fatal-Plus^® (0.3 mL, i.p.) and perfused transcardially with 200 mL of 0.1 M phosphate-buffered saline, followed by 500 mL of 4% paraformaldehyde in 0.1 M phosphate buffer solution (PBS).

4.6. Statistical analysis

Statistical analyses were conducted using StatView 5.0.1 software (Abacus Concepts, Inc., Berkley, CA) by a researcher blinded to group conditions. Measures were taken during the training period to establish a baseline to compare to post-injury performance and were analyzed by repeated measures analysis of variance (ANOVA). Post-injury measures included percent accuracy (overall and for individual cue ports), percent omissions, and latencies for choice and consumption. When significant ANOVA main effects or interactions were indicated, the Newman-Keuls post-hoc test was used to determine specific group differences. Results are expressed as the mean ± standard error of the mean (S.E.M.), and significance for all analyses was set at p ≤ 0.05.

Data availability

Data will be made available on request.