Cisplatin treatment induces attention deficits and impairs synaptic integrity in the prefrontal cortex in mice

Patients treated for cancer frequently experience chemobrain, characterized by impaired memory and reduced attention. These deficits often persist after treatment, and no preventive or curative interventions exist. In mice, we assessed the effect of cisplatin chemotherapy on attention using the 5-choice serial reaction time task and on synaptic integrity. We also assessed the capacity of mesenchymal stem cells to normalize the characteristics of chemobrain. Mice were trained in the 5-choice serial reaction time task. After reaching advancement criteria at a 4-second stimulus time, they were treated with cisplatin followed by nasal administration of mesenchymal stem cells. Cisplatin reduced the percentage of correct responses due to an increase in omissions, indicating attention deficits. Mesenchymal stem cell treatment reversed these cisplatin-induced deficits in attention. Cisplatin also induced abnormalities in markers of synaptic integrity in the prefrontal cortex. Specifically, cisplatin decreased expression of the global presynaptic marker synaptophysin and the glutamatergic presynaptic marker vGlut2. Expression of the presynaptic GABAergic marker vGAT increased. Nasal mesenchymal stem cell administration normalized these markers of synaptic integrity. In conclusion, cisplatin induces long-lasting attention deficits that are associated with decreased synaptic integrity in the prefrontal cortex. Nasal administration of mesenchymal stem cells reversed these behavioural and structural deficits.

SCieNTifiC RePoRTS | (2018) 8:17400 | DOI: 10.1038/s41598-018-35919-x of arm entries in the Y-maze, indicating that the deficit cannot be explained by a general decrease in activity 21,22 . More importantly, we showed that intranasal application of mesenchymal stem cells (MSCs) 48 and 96 hours after completion of cisplatin treatment completely reversed these cisplatin-induced memory and orientation deficits in young adult mice (Chiu et al. 23 ). As mentioned above, patients with chemobrain not only show signs of memory deficits, but also show impairment in attention and executive functioning. The 5-choice serial reaction time task (5CSRTT) has been developed to assess attention deficits in rats and mice [24][25][26][27][28] . We used this task in adult mice (7-8 months old) to examine the effects of cisplatin treatment on these translationally relevant aspects of chemotherapy-induced deficits in cognitive function. We also assessed the effects of nasal MSC application on performance in the 5CSRTT. To get more insight into the structural abnormalities associated with impaired performance in this behavioural task, we examined the effects of cisplatin and MSCs on synaptic integrity in the prefrontal cortex.

Methods
Animals. The  Experiments were performed with male C57BL/6J mice (Jackson Laboratory, Bar Harbor, ME) housed at 22 ± 2 °C, on a 12/12-hour reverse dark-light cycle (dark 830-2030 hours) with water and food ad libitum. Mice were group-housed for the duration of the study. At an age of 5-6 months, all mice were food restricted to 80-85% of their free-feeding body weight, as is standard for operant testing protocols. Mice were trained in the operant chamber as described below until all mice had reached the 4-second stimulus time advancement criteria. This approach was selected to ensure that all mice had learned the task before exposure to cisplatin. Mice were then returned to free feeding and treated with 2 cycles of 5 daily intraperitoneal injections of cisplatin (2.3 mg/kg/day; Fresenius Kabi USA, Lake Zurich, IL) or phosphate-buffered saline (PBS), followed by a 5-day rest without injections. This schedule was selected on the basis of our previous work 21 .
Experimental Plan. We performed 2 separate experiments. In Experiment 1 (saline n = 5 mice; cisplatin n = 7 mice), there were no further interventions. In Experiment 2, mice were treated with saline (n = 11) or cisplatin (n = 20) followed by nasal administration of MSCs or PBS (saline/PBS: n = 5; saline/MSCs: n = 6; cisplatin/ PBS: n = 10; cisplatin/MSCs: n = 10) 48 hours and 96 hours after completion of cisplatin treatment. We have not detected differences between the saline/PBS and saline/MSCs groups; therefore, we pooled the results from these 2 groups for data analysis and labelled the pooled group "control".
Mouse MSCs (GIBCO Mouse C57BL/6; Invitrogen, Carlsbad, CA) were cultured in Dulbecco's Modified Eagle's medium/F12 medium with GlutaMax-I, supplemented with 10% MSC-qualified foetal bovine serum and 5 µg/mL gentamycin (all from Invitrogen). Before administration of MSCs, 3 µL per nostril of hyaluronidase in PBS (total 100 U per mouse; Sigma-Aldrich, St. Louis, MO) was administered to each nostril to increase the permeability of the nasal mucosa 29,30 . Thirty minutes later mice received MSCs applied as 2 doses of 3 µL to each nostril (total volume of 12 µL with 10 6 cells per mouse per day 29,30 ) or PBS. We selected this schedule on the basis of our previous work in models of ischemic brain damage [30][31][32][33] . Three days before returning to the 5CSRTT training, mice were put back on the food-restriction schedule.

5-Choice Serial Reaction Time Task.
The 5-CSRTT was administered by a single investigator as described previously 28 and using the schedule presented in Fig. 1A. The operant testing boxes (Modular test chamber, model ENV-307W) were obtained from Med Associates INC (Fairfax, VT). Mice were exposed to 1 session per day from Monday through Friday. They were food restricted from Sunday until end of day on Friday, and then had free access to food until Sunday morning.
Chamber Habituation, Nose Poke, and Food Magazine Pre-Training. Before starting 5CSRTT training, mice were exposed to the reward (Chocolate-flavored Dustless Precision Pellets, Med Associates, St Albans, VT) in their home cage. The next day, nose-poke training was started by introducing each mouse individually to the chamber for 20 minutes, with all 5 stimulus lights and the food magazine light illuminated. Two chocolate pellets were placed in each of the 5 light apertures, and 10 chocolate pellets were placed in the magazine. Daily training sessions were repeated until the mouse reliably consumed all pellets provided. During the subsequent food magazine training sessions, a light in 1 of the 5 stimulus holes was kept on until the mouse made a nose poke and consumed the pellet in the food magazine. This session was repeated 2-3 times, until the mouse could finish up to 50 trials (consume 50 pellets) within the 20-minute session time.
5CSRTT Training. During the 5CSRTT training, a nose poke into the illuminated hole was required to obtain the reward. Training was initiated by completion of a magazine entry, triggering an intertrial interval (ITI) of 5-10 seconds during which the animal waited for illumination of 1 of the 5 holes at the back of the chamber (Fig. 1A). A nose poke to the illuminated hole completed a correct trial and triggered delivery of the reward; the cycle was then repeated. The initial stimulus time was 16 seconds. Mice were kept at the same stimulus time for 2 consecutive sessions after meeting the advancement criterion (stimulus time of 16 seconds or 8 seconds: ≥30 correct trials and 50 trials in total over 2 consecutive days; stimulus time ≤4 seconds: >50% correct trials and 100 trials). Then timing was reduced to 8 seconds, 4 seconds, 2 seconds, and 1 second.
Only a correct response during the stimulus time and a following 2-second limit hold time initiated the reward. If the mouse nose poked correctly during the stimulus time plus the 2-second limit hold, then the food reinforcer was delivered. If the mouse nose poked a non-illuminated hole (incorrect response) or failed to respond within the stimulus time plus the 2-second limit hold period (omitted response), the chamber light was illuminated for 5 seconds (time out), after which the cycle restarted. If the mouse nose poked during the 5-10-second ITI (premature response), a 5-second time out was presented. A session was completed when the maximum number of trials (50 trials for the 16-second or 8-second stimulus time, or 100 trials for stimulus times ≤4 seconds), or the maximum session duration (30 minutes) was reached. After a mouse met advancement criterion at a stimulus time of 4 seconds, we continued training it with a 4-second stimulus time for at least 8 days. Subsequently, the mouse was returned to free feeding, with cisplatin treatment initiated 3 days later. After completion of cisplatin treatment, mice were food restricted for 1 week, after which 5CSRTT was restarted at a stimulus time of 4 seconds. Mice continued the 5CSRTT until they met advancement criteria at a stimulus time of 1 second.
To determine whether there is spontaneous recovery of cisplatin-induced attention deficits, the mice in Experiment 1 were re-tested in the 5CSRTT starting 4 months after completion of treatment, at a 4-second stimulus time. We again tested the mice when they reached advancement criteria at the 1-second stimulus time.
The number and percentage of correct, incorrect, and omitted trials, latency times to correct and incorrect responses, and latency to collect the reward were recorded.
Statistical analysis. Data are expressed as mean ± standard error of the mean. Statistical analysis was performed using 1-way or 2-way analysis of variance (ANOVA) followed by Tukey analysis. P < 0.05 was considered statistically significant.

Results
Effect of Cisplatin on Learning/Memory in the 5CSRTT. Before treatment with cisplatin, we trained adult male C57BL/6 mice aged 5-6 months at the start of the study in the operant chamber until they met advancement criteria at a stimulus time of 4 seconds (see Fig. 1B for schematic overview). Subsequently, mice were treated with 2 cycles of cisplatin (5 days 2.3 mg/kg/day, 5 days rest, 5 days 2.3 mg/kg/day) or vehicle and body weight was monitored regularly (Fig. 1B). In Experiment 1, mice did not receive further treatments. In Experiment 2, half the mice received MSCs via the nasal route 48 and 96 hours after completion of cisplatin treatment.
After completion of cisplatin treatment, we reintroduced the mice to the 5CSRTT at a stimulus time of 4 seconds according (Fig. 1B). On the second day, all mice had reached the advancement criteria (>50 responses, with >50% correct). However, the cisplatin-treated mice showed a lower percentage of correct responses (Fig. 2). The data in Fig. 2 indicate that this was due to an increase in omissions. There were no group differences in the percentages of incorrect responses. At this early time point after MSC administration, no improvement of performance in the 5CSRTT was observed (Fig. 2D-F).

Effect of Cisplatin and MSCs on Performance in the 5CSRTT at a Short Stimulus Time.
Consistent with what we observed at the 4-second interval time, cisplatin-treated mice continued to show fewer correct responses at the 1-second interval time, and this was due to a larger percentage of omissions at both time points (Fig. 3A-F). These findings indicate that cisplatin treatment induces a persistent attention deficit. Notably, nasal administration of MSCs to cisplatin-treated mice normalized the number of correct responses and omissions (Fig. 3D-F).
We did not detect any effect of cisplatin treatment on the number of premature responses in both experiments (Fig. 4A,B), indicating that the attention deficits (increased omissions) in cisplatin-treated mice were not associated with an increase in impulsivity. In the next set of trials in experiment 2, we increased the ITI from 5 to 10 seconds. As expected, this increased the number of premature responses (Figs 4C vs B). However, we still did not detect any effect of cisplatin on premature responses, supporting the notion that cisplatin treatment did not increase impulsivity (Fig. 4C). However, at the longer ITI we no longer detected group differences in correct We also observed an increase in latency to reward in the cisplatin-treated mice (Fig. 5A,B) that was not affected by MSC treatment. An increase in the latency to reward may result from either lack of attention or a decrease in motivation. However, the latency to correct response, a measure frequently used to assess changes in motivation [35][36][37][38] , was not affected by cisplatin (Fig. 5C,D). Collectively, our findings indicate that cisplatin induced an increase in omissions that is mediated by a decrease in attention. However, we cannot fully exclude a contribution of decreased motivation.
To determine whether there is spontaneous recovery of the cisplatin-induced attention deficits, the mice in Experiment 1 were retested in the 5CSRTT starting 4 months after completion of treatment. Retesting the mice when they reached the advancement criterion at the 1-second stimulus time (approximately 8 weeks later) revealed that spontaneous recovery of cisplatin-induced attention deficits does not occur within this time frame (Fig. 6). The cisplatin-treated mice still displayed a decrease in correct responses that was fully explained by an increase in omissions.

Effect of Cisplatin and MSCs on Synaptic Integrity.
To determine the effects of cisplatin and MSCs on synaptic integrity, we quantified the expression of the presynaptic markers synaptophysin, vGAT, and vGlut2, the postsynaptic glutamatergic marker PSD95, and the γ-aminobutyric acid (GABA)ergic marker gephyrin in the prefrontal cortex after completion of the behavioural assessments 4 months after the last dose of cisplatin, when the animals were 11-12 months old. Treatment of mice with cisplatin sharply reduced the level of synaptophysin (Fig. 7) in the prefrontal cortex. This reduction in synaptophysin was associated with a decrease in the presynaptic glutamatergic marker vGLUT2 and an increase in the presynaptic GABAergic marker vGAT (Fig. 7).
The change in the expression of the postsynaptic glutamatergic marker PSD95 did not reach statistical significance, while expression of gephyrin was increased in cisplatin-treated mice (Fig. 8). Nasal application of MSCs normalized the expression of all markers.

Discussion
Attention deficits and problems concentrating are frequently reported by patients during and after cancer treatment [2][3][4][5][6][7] . Here, we examined the effects of cisplatin on the performance of adult male mice in the 5CSRTT to assess deficits in attention and response inhibition [24][25][26][27][28] . Our results clearly demonstrate that treatment of adult (7-8 months old) male mice with cisplatin induces a deficit in attention, demonstrated by increased omissions in the 5CSRTT. Importantly, we also show that nasal administration of MSCs after completion of cisplatin treatment reversed the attention deficit. The attention deficits as detected in the 5CSRTT were associated with abnormalities in the expression of markers of synaptic integrity in the prefrontal cortex, as characterized by a reduction in expression of the presynaptic marker synaptophysin and the glutamatergic marker vGlut2 and an increase in expression of the presynaptic GABAerigc marker vGAT. MSC treatment normalized expression of these markers of synaptic integrity in the prefrontal cortex of cisplatin-treated mice. Further studies are needed to determine whether cisplatin and MSCs have similar effects in female mice. We recently demonstrated that treatment of young adult male or female mice (8-10 weeks old at start of treatment) with cisplatin reduced performance in a number of behavioural tasks that assess spatial orientation and short-term memory, including the social recognition task and the novel object and place recognition task, and the spontaneous alternations in the Y-maze test 21,22 (and Chiu et al. 23 ). In addition, we showed that cisplatin decreases the performance of young adult mice in the puzzle box, a task developed to assess executive functioning (Chiu et al. 23 ). We also showed that nasal application of MSCs after completion of cisplatin treatment, when mice were 11-13 weeks-old, normalized cognitive functioning in these tasks (Chiu et al. 23 ). In the current study, we chose to use mice who fully had reached adulthood as this is more relevant from a translational viewpoint. We show here that cisplatin treatment of 7-8-month-old mice induces deficits in attention evidenced by an increase in omissions in the 5CSRTT. Without further intervention, the attention deficit lasts until at least 4 months after completion of cisplatin treatment. Notably, we also show that nasal administration of MSCs after completion of cisplatin treatment (when the mice are approximately 8 months old) is sufficient to reverse the behavioural deficits. This is an important finding because it shows that MSCs are capable of promoting repair not only of the young adult brain, but also of the brains of mice that have fully reached adulthood.
Our data show that cisplatin treatment decreased the percentage of correct responses without changing accuracy, but rather by increasing omissions. This increase in omissions could be due to a decrease in motivation; however, we did not detect changes in latency to correct response, which would be expected if motivation were reduced. In addition, we do not have evidence for an increase in omissions toward the end of a session, when    The prefrontal cortex plays a key role in the regulation of attention 24,39 , and our data show that impaired attention in cisplatin-treated mice is associated with changes in markers of synaptic integrity in the prefrontal cortex. More specifically, we detected a reduction in the expression of the presynaptic marker synaptophysin and in the glutamatergic presynaptic marker vGlut2. In contrast, cisplatin increased expression of the presynaptic GABAergic marker vGAT. These findings indicate a change in the inhibitory/excitatory balance in the prefrontal cortex of cisplatin-treated mice. MSC treatment prevented this change in expression of markers of synaptic integrity. Studies in rats using specific GABA-A receptor agonists or antagonists have shown that both can lead to impaired attention characterized by increased omissions 40 , indicating that both increases as well as decreases in GABAergic activity could lead to the behavioural phenotype observed in our cisplatin-treated mice.
The schedule of cisplatin treatment used in our study induces signs of neuropathic pain [41][42][43] , and there is evidence that neuropathic pain is associated with changes in excitatory/inhibitory balance [44][45][46][47][48] . Moreover, creating an excitatory imbalance in the brain is sufficient to induce mechanical allodynia 49 . It is therefore possible that cisplatin-induced changes in the inhibitory/excitatory balance contribute to the signs of peripheral neuropathy.
It has been suggested that chemotherapy accelerates the aging process in the brain [50][51][52] . Synaptophysin expression decreases during aging, and although the postsynaptic protein PSD95 also is reduced during aging, it is retained longer at the synapse when age increases [53][54][55] . Moreover, in the prefrontal cortex, GABAergic activity increases with age while it lessens over time in other areas of the brain [56][57][58][59] . Thus, our finding that cisplatin induces a decrease in synaptophysin and an increase in vGAT without changes in PSD95 is consistent with the hypothesis that cisplatin treatment leads to accelerated aging of the brain. Further studies are needed to gain more insight into the effect of cisplatin and other chemotherapeutics on brain aging.
It remains to be determined via which mechanism cisplatin induces long-lasting changes in presynaptic and postsynaptic markers and how nasally administered MSCs exert their beneficial effects on performance in the 5-CSRTT and the associated deficits in expression of markers of synaptic integrity. We hypothesize that these long-lasting changes in synaptic integrity are the result of mitochondrial deficiencies, as has also been described for other neurodegenerative disorders 60,61 . This hypothesis is supported by our recent findings showing that cisplatin induces morphological abnormalities in synaptosomal mitochondria that are associated with functional mitochondrial deficits 21 (and Chiu et al. 23 ). We also presented evidence for a causal relationship between cisplatin-induced synaptosomal mitochondrial damage and cognitive impairment. Specifically, coadministration of the mitochondrial protectant pifithrin-μ prevented both mitochondrial abnormalities and cognitive deficits 21 . Moreover, the beneficial effect of MSC treatment on cognitive function in mice treated with cisplatin followed by MSCs at the age of 2-3 months was associated with normalization of synaptosomal mitochondrial function (Chiu et al. 23 ). RNAseq analysis of the effect of MSCs on gene expression in the hippocampi of mice treated with cisplatin revealed increased expression of genes involved in mitochondrial function and oxidative phosphorylation. These previous studies were performed in mice that were much younger than those in the current study (8-10 weeks versus 7-8 months) at the time of exposure to cisplatin. It remains to be determined whether the beneficial effects of MSCs that we detected in the present study are also mediated by restoration of mitochondrial function.

Conclusions
Deficits in attention, focus, and concentration are frequently reported by patients with cancer who are treated with chemotherapy. In addition, objective tests of cognitive function in patients with chemobrain confirm the presence of attention deficits or concentration problems [2][3][4][5][6][7][13][14][15][16][17][18] . A recent imaging study in patients treated with cisplatin identified structural and functional alterations in the frontal and parietal regions of the cortex in patients treated for ovarian cancer with first-line taxane/platinum chemotherapy 62 . Here, we show that mice treated with cisplatin develop abnormalities in markers of synaptic integrity in the prefrontal cortex and deficits in attention, providing support for the clinical relevance of our model. Our findings that nasal application of MSCs after completion of cisplatin treatment normalizes both the deficits in attention as well as the abnormalities in synaptic integrity warrant further investigation and, potentially, clinical translation.

Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.