Activation of dopaminergic VTA inputs to the mPFC ameliorates chronic stress‐induced breast tumor progression

Abstract Aims Chronic stress plays an important role in promoting the progression and migration of cancers. However, little is known of any direct impact on tumor progression related to the regulation of emotion‐related circuitry. The aim of this study was to explore the neural‐circuit mechanisms underlying stress‐induced progression of cancers and the impact of emotion‐related regulation of circuitry on tumor growth. Methods Optogenetic manipulation was applied to unpredictable chronic mild stress (UCMS)–treated mice bearing breast tumor cell. The stress‐related hormones, tumor‐related cytokines, the tyrosine hydroxylase (TH)–positive neurons and their fibers, dopamine receptor–positive cells, and anxiety level were measured using ELISA, immunohistochemical staining, fluorescence in situ hybridization, and behavioral test, respectively. Results By investigating breast cancer mouse models with a chronic mild stress model, optogenetic stimulation, and behavioral analysis, we show that chronic stress induced anxiety‐like behavior in mice and increased serum concentration of norepinephrine and corticosterone, hormones closely related to stress and anxiety. Optogenetic activation of VTA TH terminals in the mPFC rescued anxiety‐like behavior induced by chronic stress. Chronic stress resulted in marked progression of breast tumors, and repetitive optogenetic activation of VTA TH terminals in the mPFC significantly attenuated stress‐induced progression of breast cancers and reduced serum concentration of norepinephrine and corticosterone. Furthermore, there was a positive correlation between serum norepinephrine or corticosterone concentration and tumor size. Conclusions These findings indicate a positive role of an emotion regulation circuit on the progression of breast cancer and reveal a link between stress, emotion regulation, and the progression of breast cancers. Our findings provide new insights pertinent to therapeutic interventions in the treatment of breast cancers.


| INTRODUC TI ON
Exposure to stress is an important risk factor for psychiatric disorders, such as anxiety and depression. 1,2 One patient population that typically live with chronic stress is cancer patients. Deleterious consequences include stress-induced psychiatric disorders, which adversely affect the progression of cancers and treatment outcomes. [3][4][5][6][7] Chronic stress markedly affects the incidence and progression of many kinds of cancers in both human and animal studies. 5,[8][9][10] Despite the evidence that cancer progression is differentially modulated by stress-related disorders through various mechanisms and conditions, little is known about the neural-circuit mechanisms underlying stress-induced progression of cancers and the impact of emotion-related regulation of circuitry on tumor growth.
The medial prefrontal cortex (mPFC) plays a critical role in regulating responses to stress. 11 A large body of evidence has shown that the mPFC receives dopaminergic afferents emanating from the ventral tegmental area (VTA) and that this pathway is critical in neural circuits that have been implicated in the regulation of emotion, 12 cognition and memory, 13 and attentional processes. 14 Study using mice has shown that optogenetic stimulation of dopamine receptor 1 (D1) cells in the mPFC has an antidepressant effect and infusion of a D1 receptor agonist into the mPFC produces antidepressant symptoms. 12 Furthermore, dopamine (DA), which is a stress-related hormone, is an important part of an adaptive response to chronic stress 15 through the regulation of tumor progression and cancer cell proliferation through dopamine receptors. [16][17][18] However, whether the activation of DA neurons within the VTA-mPFC structure plays a role in the regulation of tumor progression has not been determined.
In this study, we investigated the neural-circuit mechanism underlying stress-induced progression of cancers. To study the role of chronic stress on cancer progression, an unpredictable chronic mild stress (UCMS) mouse model protocol was used and anxiety levels were evaluated using behavioral analysis and detection of stress-related hormones. The UCMS model was then used to determine the effect of stress on tumor progression in nude mice bearing breast tumor cells. To investigate the impact of emotional regulation by activation of a neural circuit on tumor growth, we injected AAV-TH-Cre and AAV-DIO-ChR2/mCherry viruses into the mouse VTA and implanted optic fibers into the mPFC. This approach allowed the selective activation of dopaminergic VTA neurons projecting to the mPFC. Acute stimulation of the VTA TH -mPFC circuit during behavioral tests was then used to verify the role of this dopaminergic projection in regulating emotional state. Chronic repetitive stimulation of this circuit was used to examine the relationship between chronic stress, reduced anxiety levels, and tumor progression. Using a combination of breast cancer mouse models, chronic repetitive optogenetic stimulation, and behavioral analysis, we reveal that breast cancer progression under anxious states was markedly increased.
We also found that optogenetic stimulation of VTA TH terminals within the mPFC rescued chronic stress-induced anxiety-like behavior and also significantly attenuated the increased progression of breast tumors.
Our data demonstrate a previously undescribed cell-and circuit-level mechanism, which adds to our understanding of stress and the progression of the breast cancers.

| Unpredictable chronic mild stress procedure (UCMS)
The UCMS protocol was performed as previously described, 19 with minor modifications. The stress group was randomly assigned to one of the following environmental stressors during each daily session for 4 weeks: (a) overnight illumination where mice were exposed to regular room light during the night period; (b) squeezing, where 8-12 mice were placed in one box that was normally used to hold mice food for 2 hours; (c) wet bedding, where mice were placed inside a homecage on a pad that was suitably wet with drinking water for the duration of the night period; (d) restraint, where mice were placed for 1 hour in a 50-mL plastic tube with both ends open to allow air; (e) inverted light cycle, where regular room light was off during the day and on during the night for 2 days, and (f) tilted cage, where homecages were tilted at a 45° angle for 2 hours. Bodyweight was monitored every three days.

| Behavioral tests
For all animal experiments, littermate mice were randomly assigned to the experimental groups and were identifiable by a unique identification number. Experimenters were blind to experimental group allocation. In addition, data analysts were blind to experimental conditions. Groups of mice were age-matched (8-14 weeks). All mice were handled for 15-30 minutes per day for three days prior to behavioral assays to reduce stress introduced by contact with experimenter. Behavioral tests were used to evaluate anxiety levels in mice.

| Elevated plus-maze test (EPM)
A plastic elevated plus maze consisting of a central platform (5 × 5 cm) with two white open arms (25 × 5×25 cm) and two white closed arms (25 × 5×25 cm) extending from it in a plus shape was used. The maze was elevated 65 cm above the floor. Mice were individually placed in the center with their heads facing a closed arm. The number of entries and the amount of time spent in each arm type were recorded.

| Open-field test (OFT)
A plastic open-field square chamber (50 × 50 cm) was used and conceptually divided into a central field (25 × 25 cm) and a peripheral field for analysis. Each mouse was placed in a random corner of the arena at the start of each test. The number of entries and the amount of time spent in the center and corners were recorded, as was the total distance traveled. Locomotor and exploratory activity was assessed using the OFT. Distance traveled was used as a measure of locomotor activity.

| Hematoxylin and eosin staining
Hematoxylin and eosin (H&E) staining was conducted according to a routine procedure. 21 Briefly, after fixation and dehydration, freshly collected breast tumor tissues were made into 5-μm paraffin sections for H&E staining. After deparaffinization and rehydration, tumor sections were stained with hematoxylin solution for 5 minutes, rinsed in running tap water for 5-10 seconds followed by differentiation using 0.1% hydrochloric acid-ethanol for 1-3 seconds, and then rinsed in running tap water for 10-30 seconds. Then, the sections were blued for 10-30 seconds and rinsed in running tap water for 10-15 minutes.
Next, the sections were stained with eosin solution for 1-3 minutes followed by dehydration with graded alcohol and clearing in xylene.
The mounted slides were then examined and photographed using a normal fluorescence microscope (Axio Scope.A1, ZEISS).

| Surgery and virus injection
For all stereotaxic surgeries, animals were anesthetized with 1% pentobarbital sodium (20 mg/kg) prior to surgery. Once anesthetized, hair at the incision site was removed and eyes were coated with ophthalmic ointment. Next, the animals were head-fixed in a stereotaxic apparatus (RWD, China) and the incision site was steri- with the same virus as the control group but also received UCMS treatment and optogenetic stimulation. This meant that the control and stress-mCherry groups did not express the channelrhodopsin (ChR2), a light-sensitive channel protein that can be used to activate neurons following optogenetic stimulation. Mice in the stress-ChR2 group were injected with a virus cocktail including AAV-TH-Cre and AAV-DIO-ChR2-mCherry and received UCMS treatment and optogenetic stimulation. Those mice implanted with an optical fiber in the mPFC were given a one-week recovery period prior to experiments.

| Immunohistochemistry
After completion of experiments, mice were deeply anesthetized with 10% chloral hydrate (400 mg/kg) and transcardially perfused with PBS, and then ice-cold 4% paraformaldehyde (PFA) in PBS (wt/vol). Brains were dissected and postfixed at 4°C in 4% PFA overnight and then equilibrated in 30% sucrose in PBS. Next, 30-μm coronal slices were cut using a freezing microtome. Slices were stored in a cryoprotection solution at 4°C until processed further. Immunohistochemistry was performed to map the co-localization of ChR2-mCherry virus expression in the VTA with TH immunoreactivity. Antibody staining was performed on single-well floating tissue sections. Sections were incubated overnight in primary antibodies at 4°C followed by 1-h incubation with secondary antibodies at room temperature (RT). The primary antibody used in this study was anti-tyrosine hydroxylase antibody (Millipore, MAB318; 1:200). Suitable secondary antibodies were chosen to reveal the co-localization with different fluorescent colors. For counterstaining, sections were incubated for 5 minutes with 4', 6-diamidino-2-phenylindol (DAPI, 0.4 mg/mL, Sigma). All images were captured with a Zeiss LSM880 confocal microscope.

| Fluorescence in situ hybridization
We used single-probe in situ hybridization (ISH) on fixed-frozen tumor tissue sections. For single probes, coding region fragments of D1 and D2 receptors were isolated from tissue cDNA using PCR and cloned into the pCR4 Topo Vector (Thermo Fisher). Digoxigenin Biotech) and then imaged using an Zeiss Axio Imager Z2 microscope.

| Statistics
Statistical analyses were conducted, and graphs made, using GraphPad Prism 7. All results are presented as the mean ± SEM of at least five parallel assessments (refer to detailed explications in figure legends). The statistical significance of group differences was evaluated by Student's t tests for the comparison of two groups or analysis of variances (ANOVA) followed by the Bonferroni post hoc test for the comparison of three and more groups. Kolmogorov-Smirnov test was used to assess data distribution. When data followed a normal distribution but did not obey homogeneity of variance, Welch's t tests were used. When data did not follow a normal distribution and did not obey homogeneity of variance, the nonparametric Mann-Whitney test for comparison of two groups and the Kruskal-Wallis test for comparison of three or more groups were used to calculate the statistical significance of group differences. All statistical parameters for specific analyses are described in the appropriate figure legend. A value of P < .05 was considered statistically significant.

| Chronic stressors induce anxiety-like behavior in C57BL/6J (C57) mice
We adopted a chronic stress model to investigate the impact of UCMS stressors on the regulation of bodyweight and emotional behavior, such as anxiety. Two groups of C57 mice, (a) standard diet without stressors (control) and (b) standard diet with stressors (stress), were compared in these experiments ( Figure 1A). We found no significant difference in locomotor activity ( Figure 1B). Neither was there a significant difference in bodyweight ( Figure 1C) or bodyweight gain ( Figure 1D) between the control and stress groups throughout the

| Chronic stressors increase stress-related hormones in both in BLAB/c nude mice and in C57 mice
The stress-related hormones corticosterone (CORT) and norepinephrine (NE) were both markedly higher in the serum of stressed mice than in controls ( Figure 2B-C). We also found that both CORT and NE were significantly higher in stressed nude mice than in naive nude control mice ( Figure 2D-E).

| Chronic stressors induce loss of bodyweight and accelerate tumor progression
To investigate whether UCMS promotes the development of breast cancer, we measured tumor volume in two breast cancer models with or without stressors ( Figure 3A

| Optogenetic stimulation of VTA TH inputs to mPFC rescues anxiety-like behavior induced by chronic stressors
To investigate the involvement of VTA-mPFC in the regulation of anxiety-like behavior induced by chronic stressors, we optogenetically activated VTA TH terminals in the mPFC of C57 mice ( Figure 4A-B).
Immunostaining revealed that 94.95 ± 0.45% of THimmunopositive cells were positive for ChR2, and 96.06 ± 0.78% of ChR2-expressing cells, were co-labeled with TH ( Figure 4C-D), which demonstrates that the virus was effectively expressed and stable in the VTA. Also, robust ChR2 terminal expression was found in the mPFC ( Figure 4E

| Chronic repetitive optogenetic stimulation of VTA TH neuron inputs to the mPFC attenuates breast tumor progression induced by chronic stressors
The experimental procedure is shown in Figure 5A The expression of ChR2 was also found at the terminals of VTA TH projections in the mPFC ( Figure 5E). After chronic repetitive activation (10 Hz, 15 ms, 5-minute ON/5-min OFF, 30 minutes) of VTA TH projections in the mPFC once every second day, tumor growth induced by stressors was significantly lower in the stress-ChR2 group compared with the stress-mCherry group, suggesting that blue-light activation had an ameliorative effect ( Figure 5F).
Screening for tumor-related cytokines, such as TNF-a, VEGF, bFGF, IL-1α, and IL-6, as suggested by previous studies, 24,27,2526 was performed by an ELISA of collected serum. We found that VEGF, bFGF, and IL-6 were higher under chronic stress than control mice, while the chronic repetitive activation of VTA TH projections in the mPFC blocked stressor-induced cytokine increase ( Figure 5G-I). Serum IL-1α was significantly reduced in the stressed mice, and chronic repetitive activation of VTA TH projections in the mPFC led to a significantly higher IL-1α serum level than control mice and even higher than that of tumor-bearing mice without stressors ( Figure 5J).
These data suggest that chronic light stimulation of VTA TH -mPFC circuitry attenuated tumor growth promoted by chronic stressors.
Although dopamine D1 and D2 receptors both were expressed on cultured 4T1 breast tumor cells using immunohistochemistry ( Figure S1B) and in 4T1 tumor tissue by fluorescence in situ hybridization ( Figure S1A), there was no effect on the growth and proliferation of the cultured cells in vitro when these cells were treated with 0 to 4 μmol/L dopamine ( Figure S1C).

| Secretion of stress-related hormones is correlated with 4T1 tumor size
To establish a mechanistic link between neuromodulation and tumor growth, we investigated the impact of chronic repetitive activation of VTA dopaminergic terminals in the mPFC on neuroendocrine stress mediators, such as norepinephrine (NE) and corticosterone (CORT).
Enzyme-linked immunosorbent assay (ELISA) detected reduced NE ( Figure 6A) and CORT ( Figure 6C) levels in stressed mice expressing ChR2 following the chronic repetitive light stimulation upon VTA dopaminergic terminals in the mPFC compared with stressed mice expressing mCherry. Furthermore, the analysis of serum levels NE and CORT and tumor size showed that serum levels of NE and CORT were positively correlated with tumor volume across all groups ( Figure 6B, NE, Pearson's r = .7882, P < .0001; Figure 6D, CORT, Pearson's r = .8023, P < .0001). This suggests that the link between the optic stimulation of DA terminals in the mPFC and reduction in anxiety induced by chronic stress and the reduction in tumor growth is correlative.

| D ISCUSS I ON
Breast cancer has the highest incidence rate of all cancers worldwide and has the second highest cancer death rate in the United States and fifth highest death rate worldwide. 28,29 Clinical studies have shown that, compared with the general population, cancer patients, and in particular, breast cancer and lung cancer patients, have a higher risk of anxiety and depression. 30,31 This may adversely affect cancer progression, outcome of treatment, and recovery in addition to quality of life and survival. 32 Stress-related disorders, such as anxiety and depression, can reprogram dendritic cells to suppress antitumor immune responses, resulting in poor efficacy of chemotherapy and immunotherapy, 33 and the activation of the brain's reward system can modulate antitumor immunity. 34  In breast cancer studies, female mice are often used as model mice due to high incidences of breast cancer in females. Although the incidence of male breast cancer (MBC) is lower, male breast cancer is biologically similar to female breast cancer and the incidence of MBC is increasing. 35,36 In our study, we focused on the effects of neuronal regulation on cancer development under chronic stress.
Of all cancer types, breast cancer patients have some of the highest levels of anxiety observed clinically. 30 However, in studies related to stress or anxiety, male mice are often used as subjects because their endocrine levels are relatively stable in comparison with females, who have an estrous cycle. For example, one study based on canine inflammatory mammary cancer (IMC) and human inflammatory breast cancer (IBC) suggests that IBC and IMC male mouse models are a useful tool for modeling human inflammatory breast cancer. 37 Similarly, our work ( Figure 2) also showed that male mice bearing breast tumor cells were effective in our study.
Our bodies constantly adapt to both mental stress and physiological stress to sustain homeostasis during acute stress, without obvious lasting impact. However, chronic stress has many adverse effects on our body, including an increase in adiposity and eating behaviors that are detrimental to health. For example, evidence from rats indicates that reduced food intake and loss of bodyweight are criteria for stress severity. 38 In humans, however, stress has a bidirectional effect on eating. Some individuals may eat less and lose weight during or after stress, and some may increase food intake and bodyweight during stress. 39,40 Studies using mice have shown that chronic stress combined with a high-fat diet leads to significantly higher bodyweight than chronic stress with a standard diet. However, there was no difference in bodyweight between a chronic stress with standard diet group and a no stress with standard diet group. 22 This is consistent with our results ( Figure 1C-D). However, we found that bodyweight decreased following daily UCMS in BALB/c nude mice bearing MCF-7 ( Figure 3C) and MDA-MB-231 ( Figure 3E). This suggests that tumor growth and progression may affect bodyweight, whereas daily UCMS does not. In summary, these data indicate that daily UCMS over a period of four weeks promoted breast tumor growth and development ( Figure 3).
Previous work shows that the circuitry from dopaminergic VTA neurons to medial prefrontal cortex (mPFC, one of largest projections from dopaminergic VTA neurons) is implicated in rewarding and aversive behavior, reinforcing processes and cognition function. [41][42][43] Although photostimulation of lateral habenula (LHb) neurons that synapse mainly onto VTA dopamine neurons projecting to mPFC induces aversive behavior, 41 and that excitation of the projection from dopaminergic VTA neurons onto the mPFC elicits positive/reinforcement behavior during reward, 42 it remains unclear which type of neurons within the mPFC is innervated by the dopaminergic VTA neuron subset involved in reward/aversive behavior because of the complexity of dopaminergic VTA neuronal subsets and functions. 44 Neither is it known what role VTA DA-mPFC circuitry plays during or after stress. Much evidence, however, shows that the brain-reward system is related to stress. For example, the mesocortical DA system can be selectively activated by stress 45  role by generating a hedonic impact due to the effect of rewards. 49 Furthermore, optical activation of D1 cells in the mPFC by ChR2 in mice has an antidepressant and anxiolytic effect and infusion of D1 receptor agonist into the mPFC also produce antidepressant and anxiolytic responses. 12 In accordance with this evidence, we hypothesize that the activation of the projection from VTA DA into the mPFC has an antistress/anxiolytic effect. Here, we demonstrate that specific activation of the terminals within the mPFC from VTA TH neurons by optogenetic manipulation in mice rescues anxiety-like behavior induced by chronic stress (Figure 4). This suggests that the VTA TH-mPFC circuitry has an antistress/anxiolytic role.
It has long been known that the effects of stress on the body's immune function are like that of a double-edged sword. 50,51 Although acute stress may have a positive effect for the organism, 52 chronic stress is generally adverse and leads to serious health complications, including dysfunction of immunity. 5,25,33,53,54 In our work, we found that serum VEGF, bFGF, and IL-6, which are associated with tumor growth and progression, 25,27 were increased ( Figure 5G-H and J) and IL-1a, which promotes antitumor immune effects, 26 and was lower in BALB/c nude mice bearing 4T1 after chronic stress ( Figure 5I).
Chronic repetitive activation of VTA TH projections in the mPFC can rescue these effects induced by chronic stress. This is consistent with the trend for change in tumor volume ( Figure 5F). Moreover, there is evidence that VEGF and IL-6 are upregulated by norepinephrine (NE), a hormone that rises in response to the stress 55 in cultured tumor cells. 25 In the present study, we found that NE was higher after chronic stress than without stress ( Figure 2).
Administration of dopamine can ameliorate tumor growth and progression by inhibiting angiogenesis in a nude mice model of gastric cancer. 56 However, an in vitro experiment suggests that dopamine has no direct effect on the survival and proliferation of tumor cells. 57 This is consistent with our result showing no difference in the survival and proliferation of breast tumor cells treated with different concentrations of dopamine ( Figure S1). This implies that the activation of dopaminergic neural circuits may affect tumor growth through mechanisms other than affecting tumor cells directly. The more detailed mechanisms underlying the role of dopaminergic neural circuitry in tumor growth require further investigation.
In summary, we have demonstrated that specific activation of VTA TH terminals within the mPFC by optogenetic manipulation ameliorated stress-induced anxiety-like behavior, and furthermore, we found that chronic repetitive optogenetic stimulation of VTA TH inputs to the mPFC markedly attenuated stress-induced progression of breast cancer in BALB/c nude mice. These results strongly suggest an impact of emotional status on tumor growth. These findings reveal a link between stress, regulation of emotion, and progression of breast cancer, and provide new insight on therapeutic interventions in breast cancer treatment.

CO N FLI C T O F I NTE R E S T
The authors claim that there are no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
JT conceived the study. X-RX, Q.X, Y-C.H, and X-Y. Z. performed experiments. YL advised regularly. JT and X-R.X analyzed data. JT and X-RX wrote the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request (jie.tu@siat.ac.cn).