Resolution of cisplatin-induced fatigue does not require endogenous interleukin-10 in male mice

Abstract

Based on previous results showing a pivotal role of endogenous interleukin-10 (IL-10) in the recovery from cisplatin-induced peripheral neuropathy, the present experiments were carried out to determine whether this cytokine plays any role in the recovery from cisplatin-induced fatigue in male mice. Fatigue was measured by decreased voluntary wheel running in mice trained to run in a wheel in response to cisplatin. Mice were treated with a monoclonal neutralizing antibody (IL-10na) administered intranasally during the recovery period to neutralize endogenous IL-10. In the first experiment, mice were treated with cisplatin (2.83mg/kg/day) for five days and IL-10na (12 μg/day for three days) five days later. In the second experiment, they were treated with cisplatin (2.3 mg/kg/day for 5 days twice at a five-day interval) and IL10na (12 μg/day for three days) immediately after the last injection of cisplatin. In both experiments, cisplatin decreased body weight and reduced voluntary wheel running. However, IL-10na did not impair recovery from these effects. These results show that the recovery from the cisplatin-induced decrease in wheel running does not require endogenous IL-10 in contrast to the recovery from cisplatin-induced peripheral neuropathy

Introduction

The objective of cancer therapy is to kill rapidly proliferating cancer cells while sparing healthy cells. This is not easy to achieve and many of the side effects of cancer therapy stem from the cytotoxicity of the drugs that are used to target tumor cells. Cisplatin, a widely used chemotherapeutic drug administered alone or in combination with other adjuvant treatments such as radiotherapy, is a good example. The cytotoxicity of cisplatin is primarily due to its ability to form nuclear DNA adducts that cannot be repaired and block DNA replication and transcription, which leads to apoptosis in rapidly dividing cells [1]. However, it also affects post-mitotic cells in several organs including the kidney, liver, and nervous system because of its toxicity to mitochondrial DNA and the generation of mitochondrial-dependent radical oxygen species [2–4]. These effects are responsible for the neurotoxicities that develop in cancer patients and result in ototoxicity, peripheral neuropathy, and cognitive impairment [5–7].

The prevention of chemotherapy-induced neurotoxicities is difficult to manage as there is always the risk that interfering with the cytotoxicity of chemotherapeutic agents will decrease their ability to kill rapidly proliferating cells. To deal with this difficulty, the concept that has emerged over recent years has been to focus on promoting repair mechanisms rather than on protecting post-mitotic cells. This has been made possible thanks to a better understanding of the mechanisms that promote recovery from the adverse side effects of chemotherapeutic drugs on the peripheral nervous system and the brain. In the case of chemotherapy-induced cognitive impairment, the strategy has mainly consisted in replacing damaged mitochondria by fresh mitochondria transferred to neurons by mesenchymal stem cells injected intranasally [8] and, more recently, by the direct administration of healthy mitochondria [9, 10]. In the context of chemotherapy-induced peripheral neuropathy, recovery has been found to be dependent on the local recruitment at the level of the dorsal root ganglia of CD8-positive T cells that decrease the excitability of sensory neurons in an interleukin-10 (IL-10) dependent manner [11]. However, this mechanism is not restricted to the dorsal root ganglia as it also intervenes at the level of the brain and spinal cord meninges to mediate the recovery from chemotherapy-induced peripheral neuropathy in response to intranasal administration of mesenchymal stem cells [12] and functionalized mitochondria [10].

The involvement of IL-10 in these effects is different from what is known about the classical anti-inflammatory activity of this cytokine as macrophages are not the sole source of endogenous IL-10. IL-10 is also produced by T cells and can have pro- or anti-inflammatory activities depending on its source, the cell target, and the concentration at which it is released [13]. IL-10 has been found to be produced by dendritic cells to protect against cisplatin-induced nephrotoxicity [14]. IL-10 can still be produced by macrophages in the meninges and in the dorsal root ganglia for the resolution of neuropathic pain but in this case, its expression depends on the production of IL-13 by infiltrating CD8-positive T cells [15]. In addition, the cellular target of IL-10 is not represented by activated macrophages but by sensory neurons that express the IL-10 receptor [16]. Although not yet explored at the molecular level, this could be due to the ability of IL-10 to switch cellular metabolism from aerobic glycolysis to oxidative phosphorylation, thanks to its mitochondrial protectant activity [17, 18]. Such a protective mechanism is important in view of the role played by cisplatin-induced disruption of mitochondrial bioenergetics in the neurotoxicity of this agent [19, 20].

In contrast to the amount of data that has accumulated during recent years on the mechanisms that mediate the development of chemotherapy-induced peripheral neuropathy and cognitive dysfunction and their resolution, little is known about the symptom of chemotherapy-induced fatigue. Cancer patients experience fatigue as a distressing, persistent, subjective sense of physical emotional, and/or cognitive tiredness or exhaustion related to cancer and/or cancer treatment that is not proportional to recent activity, not relieved by sleep, and interferes with their usual functioning. [21]. Using murine models of cancer we have shown that cancer-induced fatigue is the result of the competition between the energy requirements of the tumor and those of skeletal muscles [22, 23] and that the development of fatigue and its recovery in response to cisplatin are dependent on the release of the growth differentiation factor 15 (GDF15) in the general circulation and the activation of its brain receptor GDNF family receptor alpha like (GFRAL) [24]. As GDF15 is a mitokine that is produced in response to mitochondrial stress, the present study was undertaken to determine whether a mechanism involving endogenous IL-10 contributes to the resolution of cisplatin-induced fatigue as it does for peripheral neuropathy. For this purpose, we chose to block endogenous IL-10 by intranasal administration of a neutralizing antibody against IL-10 as we had already done successfully in other experiments [25], and we assessed the effects of this intervention on fatigue measured by the decrease of wheel running induced in mice by cisplatin. The results were essentially negative, showing no evidence of a contribution of endogenous IL-10 in the recovery from fatigue induced by cisplatin.

Animals and methods

The experiments were carried out in male C57BL/6J mice purchased from Jackson Laboratories at ten weeks of age and maintained in a standard laboratory environment at 22°C with food and water ad libitum, and lights turned off from 7:00 PM to 7:00 AM. We used only male mice as there are no major sex differences in the neurotoxicities of cisplatin measured by peripheral allodynia, decreased wheel running, or cognitive impairment [15, 26, 27]. Mice were habituated to their housing environment for a minimum of ten days before being given a running wheel (Med Associates, Fairfax, VT) in their home cage two weeks before the start of the experiment. All experimental protocols were approved by the Animal Care and Use Committee at The University of Texas MD Anderson Cancer Center.

The general toxicity of cisplatin was measured by body weight losses [28]. Fatigue-like behavior was assessed by decreases in the voluntary wheel running activity [22, 26]. Wheel running data is presented starting the night prior to the first injection (day 0) and the number of wheel revolutions during the 12 h dark phase serves to determine a baseline for each mouse. The percent baseline wheel running for a given day is therefore the number of wheel revolutions a mouse performed during the 12 h dark phase that starts on that day divided by the baseline value for that same mouse. The time occurrence of wheel rotations is recorded by the Wheel Manager software. As the smallest bin size that this software allows is 1 min, active time spent running is calculated as the number of minutes during the 12 h dark phase that a mouse performs an arbitrarily selected number of wheel revolutions (more than 3). Wheel running velocity is calculated as the average revolutions per minute for all minutes in the 12 h dark phase during which the mouse is actively using the wheel.

To assess the effects of cisplatin on wheel running activity mice were injected intraperitoneally with a regimen of daily intraperitoneal injections of cisplatin at the dose of 2.83 or 2.3 mg/kg for five consecutive days [20, 26]. Cisplatin was injected around noon during the light phase of the light-dark cycle. The dose of 2.83 mg/kg was used when mice were given only one 5-day cycle of cisplatin. As this dose becomes toxic when repeated for longer than 5 days, cisplatin was injected at 2.3 mg/kg/day when mice were given two 5-day cycles of cisplatin separated by a 5-day rest interval [20, 26]. Control mice were injected intraperitoneally with phosphate-buffered saline (PBS).

PBS or goat anti-murine IL-10 monoclonal antibody IL-10na, Sigma-Aldrich, #15415) dissolved in PBS at the concentration of 1 μg/μl was delivered daily to each nostril at the dose of 2 × 3 μl for four to five days resulting in a total daily dose of 12 μg per mouse [25]. This dose was selected based on its ability to neutralize endogenous IL-10 produced in response to systemic inflammation induced by intraperitoneal injection of an infra septic dose of the cytokine inducer lipopolysaccharide [25]. IL-10na was administered 30 min after intranasal instillation of 100 units of hyaluronidase (Sigma-Aldrich, Saint-Louis, MO) in 3 μl PBS per nostril [25]. The instillation was done in fully awake animals. that were restrained manually with their head back in a supine position. While restrained manually with their head back in a supine position, mice were administered one intranasal instillation of 3 μl of hyaluronidase followed 30 min later by two 3 μl of IL-10na or PBS administered via a 10 μl pipette.

For statistical analysis, body weights and numbers of wheel rotations per night were expressed as a percent of the baseline measured on the day before the first injection of cisplatin. Body weights averaged 22 g at baseline. The number of wheel rotations per night at baseline averaged 30,000. The results were analyzed by two-way ANOVAs (treatment × time) with repeated measurements on the time factor. Post-hoc comparisons of group means were done using Bonferroni’s correction for multiple comparisons.

Results

In the first experiment, mice were randomized to 3 treatment groups (no cisplatin + PBS, cisplatin + PBS, and cisplatin + IL-10na, n=6 mice per group) with cisplatin injected daily at the dose of 2.83 mg/kg for five days. IL-10na was administered daily for four days starting five days after the last injection of cisplatin as this time corresponds to the beginning of wheel running recovery. There was no control group for IL-10na as we already showed that administration of IL-10na did not affect behavior in healthy mice [11, 16, 25]. Body weight decreased over time in cisplatin-treated mice (interaction treatment × day F(8,68)=21.7, p<0.001). This decrease was significant on the last day of cisplatin treatment (p<0.05) (Fig. 1A). Wheel running activity expressed as a percent of baseline, decreased over time in cisplatin-treated mice (treatment factor F(2,19)=38.5, p<0.001; interaction treatment × day F8,76)=9.10, p<0.001). This decrease was significant on days 3–5 of cisplatin treatment (p<0.001) (Fig. 1B). Wheel running activity returned slowly to control values after cessation of cisplatin treatment. A 2-way ANOVA on wheel running data during the three days of IL-10na treatment and the two following days (day 10–14) revealed a significant effect of the treatment factor (F(2,18)=19.8, p<0.001) and the treatment × day interaction (F(8,72)=6.99, p<0.001). IL-10na treatment did not modify the depressing effect of cisplatin on wheel running activity. This decrease was highly significant during the IL-10na treatment (day 10–12, p<0.001). These findings indicate that intranasal administration of IL-10na did not affect recovery from cisplatin-induced decreases in wheel running activity.

Fig. 1 –

Immunoneutralization of IL-10na does not alleviate recovery of wheel running activity from one 5-day cycle of cisplatin. The horizontal axis represents days and the vertical axis body weight (A) and running wheel activity (B). (A) Effects of cisplatin on body weight measured in grams during cisplatin treatment before initiation of administration of IL-10 neutralizing antibody (IL10-na). (B) Effects of cisplatin and IL-10na on nightly wheel running expressed as a percentage of baseline. Means + SEM, n=6/group.

A second experiment was initiated to ensure these negative findings were not due to the relatively fast recovery of wheel running from the decrease induced by only one cycle of cisplatin and the delayed IL-10na treatment. In this experiment, mice were submitted to two five-day cycles of cisplatin administration spaced five-day apart [20]. The daily dose of cisplatin was reduced from 2.83 mg/kg to 2.3 mg/kg to minimize toxicity. IL-10na treatment was administered daily for five days starting 24 h after the termination of cisplatin. Mice were randomized to 3 treatment groups (PBS only, cisplatin + PBS and cisplatin + IL-10na, n=8 mice per group). Changes in body weight and in wheel running activity were analyzed separately for the cisplatin treatment period (fifteen days extending from day 1 to day 15), and for the recovery period (twenty-six days extending from day 16 to day 42). Cisplatin decreased body weight during the treatment period (treatment factor F(2,20)=8.92, p<0.01; interaction treatment × day F(18,180)=20.5, p<0.001) (Fig. 2A). This effect was significant during the last two days of the first cycle of treatment (p<0.05) and the last three days of the second cycle (p<0.001) (Fig. 2A). Body weights did not significantly differ between treatment groups at the end of the recovery period (F(2,20)=2.79, p<0.10) (Fig. 2B).

Fig. 2 –

Immunoneutralization of IL-10 does not alleviate recovery of wheel running activity from two 5-day cycles of cisplatin. The horizontal axis represents days for Fig. 2A, C, and D, and the vertical axis body weight (A and B) and running wheel activity (C and D). (A) Effects of cisplatin on body weights measured in grams during each cycle of cisplatin treatment and before administration of IL-10 neutralizing antibody (IL-10na). (B) Effects of cisplatin and IL10na on body weights measured in grams at the end of the experiment. (C) Effects of cisplatin on nightly wheel running measured during cisplatin treatment and expressed as percent of baseline. (D) Effects of cisplatin and IL-10na on nightly wheel running measured during recovery and expressed as percent of baseline. Means + SEM, n=8/group.

Cisplatin decreased significantly wheel running activity (interaction treatment × day F(26,247)=3.53, p<0.001) (Fig. 2C). This effect was significant during the last two days of the first cycle of cisplatin treatment (p<0.001), the two days following it (p<0.05), and the last two days of the second cycle of cisplatin treatment (p<0.001). The detrimental effect of cisplatin on wheel running activity was present during the entire recovery period (treatment factor F(2, 19)=8.00, p<0.01; treatment × day F(44, 418)=0.57, not significant) (Fig. 2D). Comparison of group means according to the treatment factor showed that control mice ran more than cisplatin-treated mice (p<0.05) independently of the IL10na treatment. Mice treated with cisplatin and IL-10na did not differ from mice treated with cisplatin and PBS during the recovery period.

The reduction in wheel running could be due to a decrease in the time spent in the wheel, a reduction in running speeds, or a combination of these two factors. To determine which of these factors was responsible for the decrease in wheel running, we calculated the time spent on the wheel and the running velocity of mice during cisplatin treatment and during recovery (Fig. 3). Cisplatin and IL-10na had no effect on running velocity measured during treatment (Fig. 3A) (treatment factor F(2, 19)=0.71, not significant; treatment × time interaction F(26, 247)=1.02, not significant) or recovery (treatment factor F(2, 19)=1.65, not significant; interaction treatment × time F(46, 437)=0.59, not significant) (Fig. 3B). Time spent running decreased during cisplatin treatment (Fig. 3C) (treatment factor F(2, 19)=2.37, not significant; treatment × time interaction F(26, 247)=6.06, p<0.001) (Fig. 3C) and during recovery (treatment factor F(2, 19)=5.20, p<0.05; treatment × time interaction F(46, 437)=1.80, p<0.05) (Fig. 3D). Post hoc comparison of group means on a day-to-day basis revealed that the time spent running in the wheel was lower in both cisplatin-treated groups than in the control mice on days 16, 17, and 20 to 22 (p<0.05). However, this difference was no longer significant from day 25. There was evidence that mice treated with cisplatin and administered intranasal IL-10na spent less time running than control mice on days 18, 19, 23, and 25 (p<0.05) whereas mice treated with cisplatin and administered intranasal PBS did not differ significantly from control mice on those days. However, the two groups of mice treated with cisplatin did not differ from each other during the recovery period whether they received IL-10na or PBS. These data indicate that mice treated with cisplatin spent less time on the wheel. Treatment with IL-10na made this worse but in a non-consistent manner during the recovery period. As mice ran at the same speed as usual when they were on the wheel, their fatigue was unlikely to be due to motor impairment.

Fig. 3 –

Cisplatin treatment reduces time spent on the wheel but not velocity and these effects are not modified by IL-10na. The horizontal axis represents days and the vertical axis running velocity (A and B) and time spent running (C and D). Running velocity measured as numbers of wheel revolutions per min during cisplatin treatment (A) and after cisplatin treatment (B) in response to intranasal administration of IL-10 neutralizing antibody (IL-10na). Time spent running was measured by the total number of minutes during which mice were minimally active in the running wheels at night during cisplatin treatment (C) and after cisplatin treatment and in response to intranasal IL-10na (D). Means + SEM, n=8/group. (*) indicates time points at which mice treated with cisplatin and IL-10na but not mice treated with cisplatin and PBS differed from control mice (p<0.05).

Discussion

The present results show that intranasal administration of IL-10na does not impair recovery from cisplatin-induced decreased wheel running whether the neutralizing antibody is administered after one or two cycles of cisplatin. These results do not corroborate what has been learned about the involvement of endogenous IL-10 in the recovery from chemotherapy-induced peripheral neuropathy.

Cisplatin induces mitochondrial dysfunction in neurons in the peripheral nervous system and in the central nervous system [29, 30]. The involvement of IL-10 in recovery from cisplatin neurotoxicities has been mainly studied in the peripheral nervous system. CD8-positive T cells infiltrating the meninges recognize the cellular damage caused by cisplatin via their TIM3 receptor and transmit this information to IL-10-producing macrophages in dorsal root ganglia via IL-13 [15]. By acting on IL-10 receptors expressed in peripheral sensory neurons [16], this sequence of events promotes the resolution of cisplatin-induced peripheral neuropathy. The involvement of endogenous IL-10 in these effects has been confirmed by the lack of resolution of cisplatin-induced peripheral neuropathy in mice in which the gene for IL-10 or its receptor is deleted as well as in mice that have been treated intrathecally with IL-10na [11, 30].

We hypothesized that a similar mechanism might operate at the central level and facilitate the resolution of cisplatin-induced fatigue. Cisplatin induces mitochondrial dysfunction in brain neurons [20, 29]. This effect mediates the chemotherapy-induced cognitive impairment known as chemobrain, which can be repaired by intranasal administration of mesenchymal stem cells [20] or mitochondria [9, 10]. Although the role of endogenous IL-10 in these effects has not yet been tested directly, there is already evidence that the intranasal injection of mesenchymal stem cells that treats chemobrain also reverses chemotherapy-induced peripheral neuropathy by promoting the production of IL-10 in meningeal macrophages [12]. In these experiments, the proportion of IL-10 positive macrophages in the meninges increased from 15.4 to 40.8% in response to the administration of mesenchymal stem cells although it did not change significantly in response to cisplatin (19.3%) [12]. In the central nervous system, IL-10 has been shown to have neuroprotective activity independent of its anti-inflammatory activity. Cortical neurons express IL-10 receptors that mediate the neuroprotective action of IL-10 on neurons exposed to oxygen-glucose deprivation or glutamate toxicity [31, 32]. These effects are comparable to what has been described for IL-10 in spinal cord neurons [33]. However, these effects have been observed in vitro and their potential importance in vivo has not been explored systematically. In addition, the cellular source of endogenous IL-10 involved in these effects remains to be determined. Mesenchymal stem cells injected intranasally need to produce IL-10 to be active as mesenchymal stem cells from IL-10 deficient mice do not reverse cisplatin-induced peripheral neuropathy [12]. Because of all the uncertainties surrounding a possible role of endogenous IL-10 in centrally mediated cisplatin-induced neurotoxicities, we used a pharmacological approach based on the same strategy that had proven effective to impair recovery from inflammation-induced depression-like behavior, i.e., the intranasal injection of IL-10na [25]. The advantage of administering a neutralizing antibody is that its half-life which can be estimated at 6 to 8 days in mice [34, 35] is largely sufficient to cover the time during which it needs to be active to counteract the residual effect of cisplatin on wheel running. The fact that this strategy was unable to impair recovery from cisplatin-induced decreases in running wheel activity indicates a contrario that other mechanisms than those characterized for inflammation and for cisplatin-induced peripheral neuropathy mediate the role of mitochondrial dysfunction in this effect of fatigue.

The present study has several limitations. The main one is represented by the absence of direct validation of the target. However, as already mentioned in the methods section, the dose and route of administration of the IL-10 neutralizing antibody used in the present experiment were already validated in an experiment in which mice were injected with lipopolysaccharide [25]. In this experiment, the amount of IL-10 produced in the brain and peripheral organs in response to the systemic inflammation induced by lipopolysaccharide considerably exceeded the minor variations induced by the cisplatin [12]. Another limitation is the fact that Il-10na was administered only for a few days rather than for the entire recovery period. However, the same antibody administered intrathecally for only a few days after cisplatin treatment was sufficient to prevent recovery from cisplatin-induced peripheral neuropathy [16]. Still another limitation is the measurement of cisplatin-induced fatigue by only a single endpoint represented by reductions in wheel-running activity. However, this measure is much more sensitive than other measures of activity such as spontaneous locomotor activity as evidenced in tumor-bearing mice [22]. In addition, it has the advantage of being directly dependent on mitochondrial dysfunction as demonstrated by its extreme sensitivity to variations in the activity of the GDF15-GFRAL axis [24].

In conclusion, the results of the present study do not provide any support for the possibility that endogenous IL-10 plays a role in cisplatin-induced fatigue measured by decreased wheel running, in contrast to what has been observed for cisplatin-induced peripheral neuropathy.

Highlights.

• The present experiments were carried out in male mice to determine whether endogenous IL-10 that facilitates recovery from cisplatin-induced peripheral neuropathy also modulates recovery from cisplatin-induced fatigue

• Fatigue was measured by decreased voluntary wheel running

• Immunoneutralization of IL-10 was achieved by intranasal administration of a murine anti-IL-10 neutralizing antibody to target both central and peripheral IL-10

• Immunoneutralization of IL-10 did not interfere with the recovery of cisplatin-induced fatigue

• These findings indicate that the mechanisms of cisplatin-induced fatigue differ from those involved in cisplatin-induced peripheral neuropathy