Skip to main content
eLife logoLink to eLife
. 2021 Sep 16;10:e67535. doi: 10.7554/eLife.67535

Stimulation of hypothalamic oxytocin neurons suppresses colorectal cancer progression in mice

Susu Pan 1,2, Kaili Yin 1,2, Zhiwei Tang 3, Shuren Wang 4, Zhuo Chen 1,2, Yirong Wang 3, Hongxia Zhu 4, Yunyun Han 2,5, Mei Liu 4,, Man Jiang 3,, Ningzhi Xu 4,, Guo Zhang 1,2,
Editors: Ernie Blevins6, Mone Zaidi7
PMCID: PMC8536257  PMID: 34528509

Abstract

Emerging evidence suggests that the nervous system is involved in tumor development in the periphery, however, the role of the central nervous system remains largely unknown. Here, by combining genetic, chemogenetic, pharmacological, and electrophysiological approaches, we show that hypothalamic oxytocin (Oxt)-producing neurons modulate colitis-associated cancer (CAC) progression in mice. Depletion or activation of Oxt neurons could augment or suppress CAC progression. Importantly, brain treatment with celastrol, a pentacyclic triterpenoid, excites Oxt neurons and inhibits CAC progression, and this anti-tumor effect was significantly attenuated in Oxt neuron-lesioned mice. Furthermore, brain treatment with celastrol suppresses sympathetic neuronal activity in the celiac-superior mesenteric ganglion (CG-SMG), and activation of β2 adrenergic receptor abolishes the anti-tumor effect of Oxt neuron activation or centrally administered celastrol. Taken together, these findings demonstrate that hypothalamic Oxt neurons regulate CAC progression by modulating the neuronal activity in the CG-SMG. Stimulation of Oxt neurons using chemicals, for example, celastrol, might be a novel strategy for colorectal cancer treatment.

Research organism: Mouse

eLife digest

Colorectal (or ‘bowel’) cancer killed nearly a million people in 2018 alone: it is, in fact, the second leading cause of cancer death globally. Lifestyle factors and inflammatory bowel conditions such as chronic colitis can heighten the risk of developing the disease. However, research has also linked to the development of colorectal tumours to stress, anxiety and depression. This ‘brain-gut’ connection is particularly less-well understood.

One brain region of interest is the hypothalamus, an almond-sized area which helps to regulate mood and bodily processes using chemical messengers that act on various cells in the body. For instance, Oxt neurons in the hypothalamus produce the hormone oxytocin which regulates emotional and social behaviours. These cells play an important role in modulating anxiety, stress and depression.

To investigate whether they could also influence the growth of colorectal tumours, Pan et al. used various approaches to manipulate the activity of Oxt neurons in mice with colitis-associated cancer. Disrupting the Oxt neurons in these animals increased anxiety-like behaviours and promoted tumour growth. Stimulating these cells, on the other hand, suppressed cancer progression.

Further experiments also showed that treating the mice with celastrol, a plant extract which can act on the hypothalamus, stimulated Oxt neurons and reduced tumour growth. In particular, the compound worked by acting on a nerve structure in the abdomen which relays messages to the gut.

These preliminary findings suggest that the hypothalamus and its Oxt-producing neurons may influence the progression of colorectal cancer in mice by regulating the activity of an abdominal ‘hub’ of the nervous system. Modulating the activity of Oxt-producing neurons could therefore be a potential avenue for treatment.

Introduction

Colorectal cancer (CRC) is the third most commonly diagnosed malignant tumor and the second leading cause of cancer death globally. There were 1.8 million new cases, and 900,000 patients died of CRC annually worldwide (Bray et al., 2018). It is estimated that there were more than 1.5 million people living with a previous CRC diagnosis in the United States in 2019 (Miller et al., 2019), and approximately 147,950 new cases will be diagnosed and 53,200 individuals will die of CRC in 2020 (Siegel et al., 2020). Besides, prevalence of CRC is rapidly rising in developing countries. For instance, incidence and mortality of CRC rank third and fifth in both men and women among all cancers in China (Cao et al., 2020). Thus, it is imperative to understand the mechanism(s) of CRC development. Negative moods, including anxiety, stress, and depression, are frequently associated with the occurrences of cancers (Antoni et al., 2006; Lillberg et al., 2003). Anxiety is linked to a greater damage of adaptive immunity (Lutgendorf et al., 2008) and impaired quality of life among cancer patients (Delgado-Guay et al., 2009). Stress is related to the incidence or mortality of CRC in women (Kikuchi et al., 2017; Kojima et al., 2005; Nielsen et al., 2008). Although negative mood is associated with the development of cancer, the underlying neural mechanism remains poorly understood.

The hypothalamus is a key brain region in mood regulation (Price and Drevets, 2010; Schindler et al., 2012). Oxytocin (Oxt) neuropeptide-producing neurons in the paraventricular nucleus (PVN) of the hypothalamus are critical in the regulation of anxiety, stress, and depression (Neumann, 2008; Neumann and Landgraf, 2012). Previous work demonstrated that Oxt was anxiolytic when administered to humans (Heinrichs et al., 2003) and rodents (Blume et al., 2008; Ring et al., 2006; Windle et al., 1997), whereas disruption of Oxt gene elevated anxiety level in mice (Amico et al., 2004; Mantella et al., 2003). Hence, Oxt plays a crucial role in mood control. Recent work indicated that nerve fibers of the autonomous nervous system are critically involved in the progressions of prostate (Magnon et al., 2013), stomach (Hayakawa et al., 2017), and breast cancers (Kamiya et al., 2019). Furthermore, the central nervous system (CNS), in particular the hypothalamus, was shown to regulate peripheral tumor progression (Cao et al., 2010). However, the neuronal population(s) involved in this process remain unclear. In this work, by combining genetic, chemogenetic, pharmacological, and electrophysiological approaches, we show that Oxt neurons in the PVN regulate tumor progression in a CRC mouse model.

Results

Depletion of Oxt neurons promotes CAC progression

Dysregulation of mood is frequently associated with the occurrences of cancer (Antoni et al., 2006; Lillberg et al., 2003), while Oxt produced in the hypothalamus has an anxiolytic effect (Neumann, 2008; Neumann and Landgraf, 2012), suggesting that modulation of Oxt neurons may impact tumor progression in the periphery. To address this possibility, we crossed the OxtCre (Wu et al., 2012) with the Rosa26DTA176 knockin (Wu et al., 2006) mice (Figure 1A). By doing so, we obtained OxtCre and the littermate OxtCre;Rosa26DTA176 (OxtCre;DTA) mice, in which the Oxt-producing neurons in the brain had been depleted (Figure 1B and C). To confirm the importance of Oxt neurons in anxiety modulation, we analyzed the anxiety-like behavior of OxtCre and OxtCre;DTA mice. In the open field test, OxtCre;DTA mice spent less time in the central region than that of the OxtCre mice (Figure 1—figure supplement 1A). In the elevated plus maze test, lesion of Oxt neurons decreased the time spent in the open arms (Figure 1—figure supplement 1B). Moreover, in the light/dark box test, depletion of Oxt neurons significantly shortened the time spent in the light box (Figure 1—figure supplement 1C). Thus, lesion of Oxt neurons elevates anxiety level in mice.

Figure 1. Oxytocin (Oxt) neurons modulate the progression of azoxymethane (AOM)/dextran sodium sulfate (DSS)-induced colitis-associated cancer (CAC) in mice.

(A) A schematic diagram showing the strategy of generating OxtCre;DTA mice. When Cre recombinase is present, loxP-flanked Stop cassette is excised, therefore allowing the expression of DTA176 in Oxt neurons. Triangles represent loxP sites. Ires, internal ribosome entry site. pA, simian virus 40 polyadenylation signal. (B) The CAC was induced in the 2-month-old OxtCre and OxtCre;DTA mice using AOM and DSS (see also Figure 1—figure supplement 1D). After completing the experiment, immunofluorescent staining for Oxt (green) indicated that Oxt neurons had been depleted in the paraventricular nucleus (PVN) of OxtCre;DTA mice. Cell nuclei were counterstained with DAPI (blue). Scale bars, 100 μm. (C) The number of Oxt-positive cells in the PVN. n = 4 mice per group. (D and E) The CAC was induced in the 2-month-old OxtCre and OxtCre;DTA mice using AOM and DSS. Tumor number (D) and diameter (E) in mice treated with AOM/DSS are shown. n = 6 (OxtCre) or 5 (OxtCre;DTA) mice per group. (F) The density of proliferating cell nuclear antigen (PCNA)-positive cells in the tumor tissues of AOM/DSS-treated OxtCre and OxtCre;DTA mice. n = 4 mice per group. (G) The density of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-positive cells in tumor tissues. n = 4 mice per group. (H) Schematic diagrams showing that the indicated adeno-associated viruses (AAVs) were injected into mouse PVN. (I) Adult male OxtCre mice were injected with AAV-hSyn-GFP (control) or AAV-hSyn-DIO-hM3Dq-mCherry (hM3Dq) viruses into the PVN, and were then administered with AOM and DSS. The mice were i.p. injected with clozapine-N-oxide (CNO) every other day for 3 weeks (see also Figure 1—figure supplement 2D). Two hours after the final dose of CNO, mice were perfused with 4% paraformaldehyde (PFA). For control, we carried out double immunofluorescence staining for c-Fos (gray) and Oxt (red). For hM3Dq, immunostaining for c-Fos (green) was performed, and Oxt neurons were identified using hM3Dq-mCherry (red). DAPI staining is in blue. Scale bars, 50 μm. (J) The percentage of OxtPVN neurons expressing c-Fos. n = 4 mice per group. (K and L) Male OxtCre mice (2 months of age) were injected with the indicated AAV into PVN, and were then treated with AOM and DSS. Subsequently, mice were i.p. administered with CNO every other day for 3 weeks. The animals were then sacrificed and tumor number (K) as well as diameter (L) were assessed. n = 6 mice per group. (M) The density of PCNA-positive cells in tumor tissues. n = 4 (control) or 3 (hM3Dq) mice. (N) The density of TUNEL-positive cells in tumor tissues. n = 3 mice per group. Data are shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, two-tailed Student’s t-test (C–G, J–N).

Figure 1—source data 1. Source data for Figure 1, panels C-G and J-N.

Figure 1.

Figure 1—figure supplement 1. Depletion of oxytocin (Oxt) neurons increases anxiety level and promotes colitis-associated cancer (CAC) development in mice.

Figure 1—figure supplement 1.

(A) Open field test. The time spent in the central and peripheral regions of OxtCre and OxtCre;DTA mice at 2 months of age. Solid and dotted lines indicate medians and quartiles. n = 11 mice per group. (B) Elevated plus maze test. The time spent in the open and closed arms of the indicated mice. n = 10 mice per group. (C) Light/dark box test. The time spent in the light and dark boxes. n = 10 mice per group. (D) Schematic diagram of the azoxymethane/dextran sodium sulfate (AOM/DSS) protocol. (E and F) Body weight (E) and food intake (F) in the mice under AOM/DSS treatment. n = 6 (OxtCre) or 5 (OxtCre;DTA) mice per group. (G) The plasma Oxt levels in mice at the end of the experiment. n = 10 mice per group. (H) Representative images of colon and rectum collected from the AOM/DSS-treated OxtCre and OxtCre;DTA mice. White eclipses indicate individual tumor. (I) Colorectal length. n = 6 (OxtCre) or 5 (OxtCre;DTA) mice per group. (J) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of the tumor tissues collected from the AOM/DSS-treated mice. Scale bars, 50 μm. (K) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) labeling (red) of tumor tissues. Cell nuclei were counterstained with DAPI (blue). Scale bars, 20 μm. (L and M) The plasma samples of OxtCre and OxtCre;DTA mice after the treatment were collected. Plasma adrenocorticotropin (ACTH) (L) and corticosterone (M) levels were then measured. n = 7 (OxtCre, corticosterone) or 8 (all other groups) mice per group. Data are presented as means ± SEM (E–G, I, L, M). *p < 0.05, **p < 0.01, ***p < 0.001, two-tailed Student’s t-test (A–C, G, I, L, M).
Figure 1—figure supplement 2. Excitation of OxtPVN neurons inhibits colitis-associated cancer (CAC) progression.

Figure 1—figure supplement 2.

(A) Open field test. The control and hM3Dq adeno-associated viruses (AAVs) (hM3Dq) were injected into the paraventricular nucleus (PVN) of OxtCre mice. These mice were i.p. administered with clozapine-N-oxide (CNO) (3 mg kg–1) for 2 weeks, and then the open field test was performed. The time spent in the central and peripheral regions were recorded. Solid and dotted lines indicate medians and quartiles. n = 9 mice per group. (B) In elevated plus maze test, the time spent in the open and closed arms were assessed. n = 9 mice per group. (C) The time spent in the light and dark boxes in the light/dark box test. n = 9 mice per group. (D) Schematic diagram of the experimental design. The male OxtCre mice (2 months of age) were injected with the indicated AAV into the PVN, and were then treated with azoxymethane (AOM) and dextran sodium sulfate (DSS). Subsequently, CNO was i.p. administered every other day for 3 weeks. (E) The plasma oxytocin levels in mice at the end of the experiment. n = 10 mice per group. (F) Body weight (top) and food intake (bottom) in mice throughout the experiment. n = 6 mice per group. (G) Representative images of colorectal tissue. White eclipses indicate individual tumor. (H) Colorectal length. n = 6 mice per group. (I) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissues. Scale bars, 50 μm. (J) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) labeling (red) of tumor tissues. Cell nuclei were counterstained with DAPI (blue). Scale bars, 20 μm. (K and L) Plasma adrenocorticotropin (ACTH) (K) and corticosterone (L) levels at the end of the experiment. n = 8 (control, corticosterone) or 7 (all other groups) mice per group. Data are presented as means ± SEM (E, F, H, K, L). *p < 0.05, two-tailed Student’s t-test (A–C, E, K, L).
Figure 1—figure supplement 3. Density of immune cells in tumor tissues.

Figure 1—figure supplement 3.

(A–E) The male OxtCre mice (2 months of age) were injected with the indicated adeno-associated viruses (AAVs) into the paraventricular nucleus (PVN), and were then treated with azoxymethane (AOM) and dextran sodium sulfate (DSS). Thereafter, clozapine-N-oxide (CNO) was i.p. injected every other day for 3 weeks. Immunohistochemical stainings for CD8α, CD4, B220, NK1.1, and CD11b of tumor tissue sections were carried out. Representative images are shown. Scale bars, 50 μm. (F–J) The density of immune cells in tumor tissue. n = 4 (control, B220 or control, NK1.1) or 5 (all other groups) mice per group. Data are presented as mean ± SEM. *p < 0.05, two-tailed Student’s t-test (F).

Next, we administered azoxymethane (AOM) and dextran sodium sulfate (DSS) into the adult male OxtCre and OxtCre;DTA mice to induce colitis-associated cancer (CAC) in the colon and rectum (Figure 1—figure supplement 1D). Depletion of Oxt neurons did not significantly impact the body weight or food intake in mice fed a normal chow diet (Figure 1—figure supplement 1E,F). After the treatment, colorectal tissues and plasma samples were collected. Indeed, the plasma Oxt levels in OxtCre;DTA mice were barely detectable (Figure 1—figure supplement 1G), suggesting the disruption of Oxt-producing neurons. Notably, the number and diameter of CAC were both increased in the OxtCre;DTA mice (Figure 1D and E; Figure 1—figure supplement 1H), while colorectal length was not significantly affected (Figure 1—figure supplement 1I). Depletion of Oxt neurons promoted cell proliferation in the CAC, as demonstrated by the increased number of cells positive for proliferating cell nuclear antigen (PCNA), a marker for proliferating cell (Figure 1F; Figure 1—figure supplement 1J). Moreover, lesion of Oxt neurons inhibited cell apoptosis in the tumors as revealed by the reduced number of cells positive for terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) (Figure 1G; Figure 1—figure supplement 1K). Together, these data indicate that depletion of Oxt neurons promotes CAC development in mice.

Given that depletion of Oxt neurons elevated anxiety level in mice, and that the dysregulation of hypothalamic-pituitary-adrenal (HPA) axis can elicit stress, next, we assessed the circulating adrenocorticotropin (ACTH) and corticosterone levels in OxtCre and OxtCre;DTA mice with AOM/DSS-induced CAC. Plasma ACTH and corticosterone levels were evidently increased in the OxtCre;DTA mice comparing with the OxtCre mice (Figure 1—figure supplement 1L,M). Thus, the dysregulation of the HPA axis may contribute to the CAC development in the OxtCre;DTA mice.

Chemogenetic activation of OxtPVN neurons suppresses CAC progression

Next, we asked whether stimulation of Oxt neurons in the PVN (OxtPVN) inhibits CAC progression. To do so, we employed the designer receptor exclusively activated by designer drug (DREADD) (Roth, 2016; Smith et al., 2016) approach to manipulate these neurons. Specifically, OxtCre mice were bilaterally injected with adeno-associated virus (AAV) carrying GFP (AAV-hSyn-GFP), or Cre-dependent hM3Dq-mCherry into the PVN (Figure 1H). To validate the DREADD system, CAC was induced in virus-injected mice. These animals were then intraperitoneally (i.p.) administered with a synthetic ligand, clozapine-N-oxide (CNO) every other day for 3 weeks. Two hours after the final dose of CNO, the mice were perfused with 4% paraformaldehyde (PFA), and then brain tissues were harvested. Immunofluorescent staining showed that treatment with CNO elicited a robust c-Fos expression in the OxtPVN neurons of hM3Dq AAV-injected mice compared with the controls (Figure 1I and J), suggesting the activation of these neurons. Mirrored with the results of Oxt neuron depletion, activation of OxtPVN neurons significantly relieved anxiety-like behavior in mice (Figure 1—figure supplement 2A-C). Thereafter, control and hM3Dq-mCherry AAVs were injected into the PVN of OxtCre mice. CAC was induced in these mice using AOM and DSS, and then CNO was i.p. administered every other day for 3 weeks (Figure 1—figure supplement 2D). After the treatment, plasma Oxt level was elevated, whereas body weight and food intake had not been significantly affected in hM3Dq AAV-infected mice (Figure 1—figure supplement 2E,F). Notably, the elevation of plasma Oxt level following chemogenetic excitation of Oxt neurons has been observed previously (Grund et al., 2019). Both tumor number and tumor diameter were reduced in mice whose OxtPVN neurons had been excited (Figure 1K and L; Figure 1—figure supplement 2G), whereas colorectal length was not impacted (Figure 1—figure supplement 2H). In agreement with the reduction in tumor size, the number of proliferating cells, revealed by the immunostaining for PCNA, was significantly decreased in hM3Dq AAV-injected mice compared with the controls (Figure 1M; Figure 1—figure supplement 2I). Besides, the TUNEL assay showed that the number of apoptotic cells was evidently increased (Figure 1N; Figure 1—figure supplement 2J). Thus, activation of OxtPVN neurons inhibits CAC progression by suppressing cell proliferation and promoting cell apoptosis.

Our assays indicated that plasma ACTH and corticosterone levels were markedly decreased in the hM3Dq AAV-injected mice (Figure 1—figure supplement 2K,L), implying that the reduced activity of HPA axis may contribute to the tumor suppression effect of OxtPVN neuron activation.

The activation of the anti-tumor immunity is crucial for cancer treatment, hence, we asked whether any of the immune cells contributes to the anti-tumor effect of OxtPVN neuron activation. To address this question, we assessed these cells in the tumor tissues. Indeed, the number of CD8+ T cells was markedly increased in hM3Dq AAV-injected mice compared with controls (Figure 1—figure supplement 3A,F), and there was no significant change in CD4+ T cells, B cells, NK cells, or macrophages (Figure 1—figure supplement 3B-E, and G-J). Hence, activation of OxtPVN neurons may enhance the anti-tumor immunity by increasing the number of CD8+ T cells.

The anti-tumor effect of OxtPVN neuron activation is dependent on its action in the CNS

Oxt neurons regulate peripheral physiology via both the neural and the endocrinal pathways (Zhang et al., 2021). Next, we asked whether the central action is important for OxtPVN neuron activation to suppress CAC progression. To this end, we elected to centrally block Oxt receptor using L-368,899, an Oxt receptor (OTR) antagonist. Specifically, adult male OxtCre mice were bilaterally injected with control or hM3Dq AAV into the PVN, and then CAC was induced using AOM and DSS. Subsequently, these mice were i.p. administered with CNO and i.c.v. injected with aCSF (artificial cerebrospinal fluid) or L-368,899 every other day for 3 weeks (Figure 2—figure supplement 1A). After the treatment, these mice were perfused with 4% PFA, and then brain tissues were sectioned. Immunofluorescent staining showed that treatment with CNO elicited a dramatic c-Fos expression in the OxtPVN neurons of hM3Dq AAV-injected mice compared with the controls (Figure 2A and B), suggesting the excitation of OxtPVN neurons.

Figure 2. Brain oxytocin (Oxt) receptor is crucial for OxtPVN neuron activation to suppress colitis-associated cancer (CAC).

(A) Adult male OxtCre mice (2 months of age) were injected with AAV-hSyn-GFP (control) or AAV-hSyn-DIO-hM3Dq-mCherry (hM3Dq) viruses into the paraventricular nucleus (PVN), and then colitis-associated cancer (CAC) was induced using azoxymethane (AOM) and dextran sodium sulfate (DSS). Subsequently, these mice were administered with clozapine-N-oxide (CNO) (i.p.), and artificial cerebrospinal fluid (aCSF) or L-368,899 (i.c.v.), an Oxt receptor antagonist (OTR anta), every other day for 3 weeks. Mice were then perfused with 4% paraformaldehyde (PFA). For control, double immunofluorescence staining for c-Fos (gray) and Oxt (red) was performed. For hM3Dq, immunofluorescent staining for c-Fos (green) was performed, and Oxt neurons were identified using hM3Dq-mCherry (red). Cell nuclei were counterstained with DAPI (blue). Scale bars, 50 μm. (B) The percentage of Oxt neurons expressing c-Fos in the PVN. n = 7 (hM3Dq, OTR anta) or 6 (all other groups) mice per group. (C) The OxtCre mice (2 months of age) were injected with indicated adeno-associated viruses (AAVs) into the PVN, and then CAC was induced using AOM and DSS. Subsequently, these mice were administered with CNO (i.p.), as well as aCSF or L-368,899 (i.c.v.), the OTR antagonist (OTR anta), every other day for 3 weeks (see also Figure 2—figure supplement 1A). Representative images of colorectal tissue after the treatments are shown. White eclipse outlines the individual tumor. (D and E) Tumor number (D) and diameter (E). ns, not significant. n = 7 (hM3Dq, OTR anta) or 6 (all other groups) mice per group. (F) Colorectal length. n = 7 (hM3Dq, OTR anta) or 6 (all other groups) mice per group. (G and H) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissues. Representative images (G) and the density of PCNA-positive cells (H) are shown. Scale bars, 50 μm. ns, not significant. n = 4 mice per group. (I and J) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay of tumor tissues. Representative images (I) and the density of TUNEL-positive cells (J) are shown. TUNEL labeling is in red. Cell nuclei were counterstained with DAPI (blue). Scale bars, 20 μm. ns, not significant. n = 4 mice per group. Data are presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Bonferroni’s post hoc test.

Figure 2—source data 1. Source data for Figure 2, panels B, D-F, H and J.

Figure 2.

Figure 2—figure supplement 1. Body weight and food intake in mice.

Figure 2—figure supplement 1.

(A) A schematic diagram of the experimental design. The control and hM3Dq adeno-associated virus (AAV) were injected into the paraventricular nucleus (PVN) of OxtCre mice (2 months of age). These animals were then induced colitis-associated cancer (CAC) using azoxymethane (AOM) and dextran sodium sulfate (DSS). Subsequently, these mice were administered with clozapine-N-oxide (CNO) (i.p.) and artificial cerebrospinal fluid (aCSF) or L-368,899 (i.c.v.), an oxytocin (Oxt) receptor antagonist (OTR anta), every other day for 3 weeks. (B and C) Body weight (B) and food intake (C) in mice throughout the experiment. n = 8 (control+ CNO + OTR anta) or 7 (all other groups) mice per group. Data are presented as means ± SEM.

Treatment with CNO and L-368,899 did not significantly impact the body weight or food intake in mice (Figure 2—figure supplement 1B,C). As anticipated, activation of OxtPVN neurons inhibited CAC progression in mice (Figure 2C–E). Notably, brain treatment with L-368,899 significantly abrogated this effect (Figure 2C–E). Colorectal length remained not impacted in the mice administered with CNO and L-368,899 (Figure 2F). Moreover, the immunostaining for PCNA revealed that excitation of OxtPVN neurons inhibited cell proliferation, however, this effect was markedly attenuated when the mice were administered with L-368,899 (Figure 2G and H). Furthermore, the TUNEL assay showed that the effect of activation of OxtPVN neurons on cell apoptosis was diminished when the mice were administered with L-368,899 (Figure 2I and J). Collectively, these data suggest that the tumor suppressive effect of OxtPVN neuron activation is dependent on its action in the CNS.

OxtPVN neurons regulate the neuronal activities in the sympathetic CG-SMG

The CNS is known to control peripheral physiology via both the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS). Besides, the sympathetic celiac-superior mesenteric ganglion (CG-SMG) predominantly innervates colon and rectum. Hence, we examined the effect of OxtPVN neuron activation on CG-SMG neuronal activity. To do this, adult male OxtCre mice were injected with control and hM3Dq AAV into the PVN. After recovery, these mice were i.p. administered with CNO. Two hours later, CG-SMG was dissected and fixed in 4% PFA. Double immunofluorescence staining for c-Fos and tyrosine hydroxylase (TH), a marker of catecholamine neuron, revealed that the activities of the sympathetic neurons in CG-SMG were significantly inhibited following the activation of OxtPVN neurons (Figure 3A and B).

Figure 3. Surgical removal of celiac-superior mesenteric ganglion (CG-SMG) attenuates the tumor-promoting effect of oxytocin (Oxt) neuron depletion.

(A) Adult male OxtCre mice were injected with control or hM3Dq adeno-associated viruses (AAVs) (hM3Dq) into the paraventricular nucleus (PVN). After surgical recovery, these mice were administered with clozapine-N-oxide (CNO). Two hours later, CG-SMG were dissected and fixed in 4% paraformaldehyde (PFA). Double immunofluorescence staining for c-Fos (red) and tyrosine hydroxylase (TH, in green) of the CG-SMG was performed. Cell nuclei were counterstained with DAPI (blue). Scale bars, 50 μm. (B) The percentage of TH-positive cells expressing c-Fos in the CG-SMG. n = 4 mice per group. (C) Representative images showing mouse abdominal cavity with (left panel), or without (right panel) CG-SMG (following the resection). (D) A schematic diagram of experimental design. The colitis-associated cancer (CAC) was induced in adult OxtCre and OxtCre;DTA mice using azoxymethane (AOM) and dextran sodium sulfate (DSS). After the first cycle of DSS treatment, sham operation and CG-SMG resection were performed in mice. (E) Representative images of colorectal tissue after the treatment. White eclipse was used to outline the individual tumor. (F and G) Tumor number (F) and diameter (G). ns, not significant. n = 8 (OxtCre, sham) or 7 (all other groups) mice per group. (H and I) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissue. Representative images (H) and the density of PCNA-positive cells (I) are shown. ns, not significant. Scale bars, 50 μm. n = 4 (OxtCre, sham) or 5 (all other groups) mice per group. (J and K) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay of tumor tissue. Representative images (J) and the density of TUNEL-positive cells (K) are shown. TUNEL labeling is in red. Cell nuclei were counterstained with DAPI (blue). ns, not significant. Scale bars, 20 μm. n = 5 (OxtCre) or 4 (OxtCre;DTA). Data are presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, two-tailed Student’s t-test (B) or one-way ANOVA with Bonferroni’s post hoc test (F, G, I, K).

Figure 3—source data 1. Source data for Figure 3, panels B, F, G, I and K.

Figure 3.

Figure 3—figure supplement 1. Transection of the preganglionic fiber of CG-SMG abolishes the inhibitory effect of Oxt neuron activation.

Figure 3—figure supplement 1.

(A) Schematics of in vivo single-unit recordings in celiac-superior mesenteric ganglion (CG-SMG) with the transection of preganglionic fiber. Clozapine-N-oxide (CNO) was injected via an infusion cannula directed to third ventricle (3V). (B and C) Representative images displaying the CG-SMG with (B) or without (C) the preganglionic nerve fiber (following the transection). (D) Normalized firing rate of recorded CG-SMG neurons in response to CNO infusion in sham (top) and preganglionic fiber transection groups (bottom). Dashed line indicates the time point of CNO delivery. i.c.v., intracerebroventricular injection. n = 30 cells from 4 mice (sham) or 34 cells from 4 mice (transection). (E) Statistics of average firing frequency of CG-SMG neurons in response to CNO infusion in sham and preganglionic fiber transection groups. Solid and dotted lines indicate medians and quartiles, respectively. ns, not significant. n = 30 cells from 4 mice (sham) or 34 cells from 4 mice (transection). (F and G) Correlation of firing rate before and after CNO infusion in sham (F) and preganglionic fiber transection groups (G). Green filled circles represent individual units with significant lower firing frequency after CNO infusion. Red triangles represent the units with higher firing rate. Gray squares indicate neurons without significant change in firing rate. Inset: proportions of CG-SMG neurons with or without change in firing rate following CNO infusion. n = 30 cells from 4 mice (sham) or 34 cells from 4 mice (transection). Solid and dotted lines indicate medians and quartiles (E). *p < 0.05, one-way ANOVA with Bonferroni’s post hoc test (E).
Figure 3—figure supplement 2. Body weight and food intake in mice.

Figure 3—figure supplement 2.

(A and B) Body weight (A) and food intake (B) in mice throughout the experiment. n = 9 (OxtCre, sham), 10 (OxtCre, resection), 7 (OxtCre;DTA, sham), or 9 (OxtCre;DTA, resection) mice per group. (C) Colorectal length. n = 8 (OxtCre, sham) or 7 (all other groups) mice per group. Data are presented as means ± SEM.

To confirm this OxtPVN neuron -> THCG-SMG neuron pathway, we cut the preganglionic nerve fiber of CG-SMG, and then assessed the neuronal activity in this ganglion using in vivo single-unit recordings. Specifically, adult male OxtCre mice were injected with control and hM3Dq AAV into the PVN, and were also implanted with infusion cannula directed to the third ventricle. After recovery, these animals were performed sham operations, or the transection of the preganglionic fiber of CG-SMG (Figure 3—figure supplement 1A-C). Subsequently, the 6 min control (1% DMSO in aCSF) spiking activity was acquired before CNO (1 µg per mouse) application through the pre-implanted cannula. Single-unit spikes from 30 (sham) and 34 (transection) CG-SMG neurons were isolated, and the firing rates were compared before and after CNO infusion (Figure 3—figure supplement 1D). Group data showed that i.c.v. administration of CNO significantly reduced the firing frequency of CG-SMG neurons, however, transection of preganglionic fiber significantly abolished this effect (Figure 3—figure supplement 1E). Scatterplot of mean firing frequency of individual CG-SMG neuron revealed a mixed modulation following OxtPVN neurons activation (Figure 3—figure supplement 1F,G). The majority of CG-SMG neurons (67%) displayed a decreased firing frequency after CNO infusion. Only a small proportion of neurons (16%) showed an increased firing frequency. The remainder (17%) maintained their activity level after CNO infusion. Yet, after the transection of the preganglionic fiber, the majority of CG-SMG neurons (65%) maintained their activity level after CNO infusion. Hence, following OxtPVN neuron activation, the signal that leads to the suppression of CG-SMG neurons is transmitted through the preganglionic fiber.

The CG-SMG is required for lesion of Oxt neurons to promote CAC development

Next, we assessed the OxtPVN neuron -> THCG-SMG neuron connection using the CAC mouse model. To this end, CAC was induced in the adult OxtCre and OxtCre;DTA mice using AOM and DSS. After the first cycle of DSS treatment, CG-SMG resection and sham surgeries were performed in mice (Figure 3C and D). These manipulations did not significantly impact body weight or food intake in mice (Figure 3—figure supplement 2A,B). While depletion of Oxt neurons led to the increasing of CAC number and diameter, CG-SMG resection markedly attenuated these effects (Figure 3E–G). We noted that colorectal length was not affected in these mice (Figure 3—figure supplement 2C). In agreement with the data of tumor number and size, the effects on cell proliferation and cell apoptosis were both attenuated when CG-SMG were removed from these mice (Figure 3H–K). Taken together, the promotion of CAC development owing to Oxt neuron deficiency is mediated by the sympathetic CG-SMG.

Celastrol enhances OxtPVN neuron excitability by increasing their input resistance

Celastrol is a pentacyclic triterpenoid initially extracted from the root of thunder god vine. A recent study showed that treatment with celastrol decreased the body weight in obese mice, but not mice with normal weight (Ma et al., 2015). A following study suggested that hypothalamus is critical for celastrol to regulate energy balance (Liu et al., 2015). Therefore, we assessed the effect of i.c.v. administered celastrol on hypothalamic neuronal activity. The data showed that the number of c-Fos-positive cells was increased in the PVN, but not other nuclei (Figure 4—figure supplement 1A,B), suggesting that brain treatment with celastrol stimulates neurons in the PVN. Oxt neurons in the PVN play a critical role in energy balance control, therefore, we asked whether its activity is modulated by celastrol. To answer this question, we analyzed Oxt neuron excitability after bath application of celastrol via slice electrophysiology. The hypothalamic slices were obtained from OxtCre;Rosa26-LSL-EYFP (OxtCre;EYFP) mice, in which enhanced yellow fluorescent protein (EYFP) was expressed in Oxt neurons (Figure 4A). In response to 500 ms current steps, Oxt neurons fired more action potentials (AP) across increasing current injections in celastrol condition, suggesting an enhanced neuronal excitability (Figure 4B and C). We also analyzed the AP waveforms, and found that celastrol increased the size of afterhyperpolarization (Figure 4D and E), but did not impact AP threshold, AP amplitude, AP half-width, or AP area (Figure 4D and F; Figure 4—figure supplement 1C-E). Moreover, celastrol increased input resistance of Oxt neurons, which might increase neuronal excitability (Figure 4G). These data implicate that celastrol enhances Oxt neuron firing.

Figure 4. Celastrol enhances the excitability of OxtPVN neurons, and its administration in the brain inhibits colitis-associated cancer (CAC) progression.

(A) Electrophysiology of paraventricular nucleus (PVN) slice of 4-month-old OxtCre;EYFP mice. Left, a differential interference contrast (DIC) image of the recorded neuron (arrow). Middle, expression of enhanced yellow fluorescent protein (EYFP) (green) in the same cell suggests that it is an oxytocin (Oxt) neuron. Right, merged image. Scale bar, 10 μm. (B) Voltage response of Oxt neuron in response to 100 and –50 pA current injection in control and celastrol (5 μM in artificial cerebrospinal fluid [aCSF]) conditions. (C) Bath application of celastrol increased the number of action potentials (AP) fired across increasing current injections. n = 20 cells from five mice (control or celastrol). (D) Representative AP traces from control and celastrol conditions. Arrowhead indicates the AP threshold. (E–G) The size of afterhyperpolarization (AHP) (E), AP threshold, (F) and input resistance (Rm) (G) in control and celastrol conditions. Solid and dotted lines indicate medians and quartiles, respectively. n = 23 cells (E, F) or 27 cells (G) from five mice (control) or 28 cells from five mice (celastrol). (H and I) The CAC was induced in male C57 BL/6 mice (2 months of age) using azoxymethane (AOM) and dextran sodium sulfate (DSS). These animals were then i.c.v. administered with control versus celastrol every other day for 3 weeks. After the treatment, tumor number (H) and diameter (I) were determined. n = 10 (control) or 8 (celastrol) mice per group. (J) Colorectal length. n = 10 (control) or 8 (celastrol) mice per group. (K) The density of proliferating cell nuclear antigen (PCNA)-positive cells in tumor tissue. n = 5 mice per group. (L) The density of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-positive cells in tumor tissue. n = 5 mice per group. Data are presented as means ± SEM (C, H–L). *p < 0.05, **p < 0.01, ***p < 0.001, two-way ANOVA with Bonferroni’s post hoc test (C), or two-tailed Student’s t-test (E, G, H, I, K, L).

Figure 4—source data 1. Source data for Figure 4, panels C and E-L.

Figure 4.

Figure 4—figure supplement 1. Celastrol excites neurons in the PVN and promotes Oxt release from PVN.

Figure 4—figure supplement 1.

(A) Male C57 BL/6 mice (2 months of age) were i.c.v. administered with vehicle or celastrol (0.5 µg). Two hours later, the mice were anesthetized, and were then perfused with 4% paraformaldehyde (PFA). Immunofluorescent staining for c-Fos (red) of brain tissue sections was carried out. Representative images display the expression of c-Fos in the paraventricular (PVN), dorsomedial (DMH), ventromedial (VMH), and arcuate (Arc) nuclei. Cell nuclei were counterstained with DAPI (blue). Scale bars, 50 μm. (B) The number of c-Fos-positive cells in the neuronal nuclei. n = 4 (celastrol, PVN) or 5 (all other groups) mice per group. (C–E) Action potential amplitude (C), half-width (D), and area (E) in control and celastrol condition (5 μM in artificial cerebrospinal fluid [aCSF]). Solid and dotted lines indicate medians and quartiles, respectively. n = 23 cells from five mice (control) or 28 cells from five mice (celastrol). (F) The PVN tissue slices of male C57 BL/6 mice (2 months of age) were dissected from the brain. Basal and celastrol-elicited oxytocin (Oxt) release rates were then determined. n = 10 mice per group. Data are presented as means ± SEM (B, F). *p < 0.05, two-tailed Student’s t-test (B, F).
Figure 4—figure supplement 2. Brain treatment with celastrol suppresses colitis-associated cancer (CAC) progression in mice.

Figure 4—figure supplement 2.

(A) Schematic diagram of the experimental design. The CAC was induced in the male C57 BL/6 mice (2 months of age) using azoxymethane (AOM) and dextran sodium sulfate (DSS). These mice were then i.c.v. administered with vehicle versus celastrol every other day for 3 weeks. (B) The plasma oxytocin (Oxt) levels at the end of the experiment. n = 8 mice per group. (C and D) The body weight (C) and food intake (D) in mice throughout the experiment. n = 10 (control) or 8 (celastrol) mice per group. (E) Representative images of the colorectal tissue. White eclipses outline individual tumor. (F) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissues. Scale bars, 50 μm. (G) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay of tumor tissues. TUNEL labeling is shown in red. Cell nuclei were counterstained with DAPI (blue). Scale bars, 20 μm. Data are presented as means ± SEM. *p < 0.05, two-tailed Student’s t-test (B).

Besides, the above data suggested that celastrol might promote Oxt release from the OxtPVN neurons. To address this possibility, we carried out an ex vivo Oxt release assay. The PVN slices were dissected from the male adult C57 BL/6 mice. These tissue slices were balanced in normal Locke’s solution, and then in the same solution supplemented with celastrol. The data showed that treatment with celastrol enhanced the rate of Oxt releasing (Figure 4—figure supplement 1F). Altogether, these data demonstrate that celastrol could excite OxtPVN neurons.

Brain treatment with celastrol suppresses CAC progression in mice

Next, we assessed the effect of brain administered celastrol on CAC progression. To this end, CAC was induced in adult male C57 BL/6 mice using AOM and DSS (Figure 4—figure supplement 2A). These mice were then implanted with a guide cannula directed to the third ventricle. After surgical recovery, vehicle and celastrol were administered into the third ventricle via the pre-implanted cannula every other day for 3 weeks (Figure 4—figure supplement 2A). Mice receiving celastrol treatment exhibited higher plasma Oxt level than that of the controls (Figure 4—figure supplement 2B), suggesting that this chronic treatment stimulated OxtPVN neurons. Consistent with the previous study (Liu et al., 2015), treatment with celastrol did not impact energy balance in CAC mice with normal body weights (Figure 4—figure supplement 2C,D). This treatment significantly reduced tumor number and diameter (Figure 4H,I; Figure 4—figure supplement 2E), while it did not affect colorectal length (Figure 4J). Besides, cell proliferation was suppressed, and cell apoptosis was enhanced in the tumor tissue of mice treated with celastrol (Figure 4K and L; Figure 4—figure supplement 2F,G). Collectively, these data indicate that brain treatment with celastrol suppresses CAC progression in mice.

Depletion of Oxt neuron abolishes the anti-tumor effect of celastrol

The above data suggested that hypothalamic Oxt neurons are important for celastrol to suppress CAC progression in mice. To address this question, the CAC was induced in the OxtCre and OxtCre;DTA mice (Figure 5A). These mice were then i.p. injected with vehicle versus celastrol every other day for 3 weeks (Figure 5A). Treatment with celastrol did not significantly impact the body weight or food intake in mice (Figure 5B and C). While celastrol inhibited CAC progression in mice, lesion of Oxt neurons could markedly abrogate this effect (Figure 5D–F). Lesion of Oxt neuron or celastrol treatment did not have noticeable effect on colorectal length (Figure 5G). Notably, the effects of celastrol on cell proliferation and cell apoptosis in CAC were both attenuated in the mice deficient for Oxt neurons (Figure 5H–K). Thus, hypothalamic Oxt neurons are required for celastrol to suppress CAC progression.

Figure 5. Depletion of oxytocin (Oxt) neurons attenuates the anti-tumor effect of celastrol.

Figure 5.

(A) A schematic diagram of experimental design. The colitis-associated cancer (CAC) was induced in the OxtCre and OxtCre;DTA mice (2 months of age), in which control solution and celastrol (Cel) were i.p. administered every other day for 3 weeks. (B and C) Body weight (B) and food intake (C) in mice throughout the experiment. n = 7 (OxtCre, Ctrl), 8 (OxtCre, Cel), or 6 (OxtCre;DTA) mice per group. (D) Representative images of colorectal tissue after the indicated treatments. White eclipses indicate individual tumor. (E and F) Tumor number (E) and diameter (F). Cre;DTA, OxtCre;DTA. ns, not significant. n = 7 (OxtCre, Ctrl), 8 (OxtCre, Cel), or 6 (OxtCre;DTA) mice per group. (G) Colorectal length. n = 7 (OxtCre, Ctrl), 8 (OxtCre, Cel), or 6 (OxtCre;DTA) mice per group. (H and I) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissue. Representative images (H) and the density of PCNA-positive cells (I) are shown. ns, not significant. Scale bars, 50 μm. n = 4 (OxtCre) or 5 (OxtCre;DTA) mice per group. (J and K) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay of tumor tissue. Representative images (J) and the density of TUNEL-positive cells (K) are shown. TUNEL labeling is in red. Cell nuclei were counterstained with DAPI (blue). ns, not significant. Scale bars, 20 μm. n = 4 mice per group. Data are presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Bonferroni’s post hoc test (E, F, I, K).

Figure 5—source data 1. Source data for Figure 5E–G,I,K.

Agonism of β2-adrenergic receptor attenuates the anti-tumor effect of OxtPVN neuron activation

Next, we interrogated whether activation of SNS target, that is, β2 adrenergic receptor (β2AR), would attenuate the anti-tumor effect of OxtPVN neuron excitation. Our data showed that isoprenaline, an agonist for β2AR, did not affect the activity of CG-SMG neurons (Figure 6—figure supplement 1A,B), suggesting that it is proper to use this drug to target CAC cells. Thereafter, adult male OxtCre mice were bilaterally injected with control and hM3Dq AAV into the PVN, and then CAC was induced. These mice were i.p. administered with CNO every other day, and were also i.p. injected with saline or isoprenaline on a daily basis. These treatments were continued for 3 weeks (Figure 6—figure supplement 1C). Subsequently, these mice were perfused with 4% PFA, and then brain tissues were sectioned. Immunofluorescent staining showed that treatment with CNO elicited a robust c-Fos expression in the OxtPVN neurons of hM3Dq AAV-injected mice compared with the controls (Figure 6A and B), suggesting the activation of these neurons.

Figure 6. Treatment with an agonist for β2 adrenergic receptor attenuates the anti-tumor effect of OxtPVN neuron activation.

(A) Control and AAV-hSyn-DIO-hM3Dq-mCherry (hM3Dq) viruses were injected into the paraventricular nucleus (PVN) of male adult OxtCre mice. Colitis-associated cancer (CAC) was then induced using azoxymethane (AOM) and dextran sodium sulfate (DSS). These mice were i.p. administered with clozapine-N-oxide (CNO) every other day and i.p. injected with saline or isoprenaline, a β2 adrenergic receptor agonist, on a daily basis. After 3 weeks of treatment, mice were perfused with 4% paraformaldehyde (PFA). For control, double immunofluorescence staining for c-Fos (gray) and oxytocin (Oxt) (red) was performed. For hM3Dq, immunofluorescent staining for c-Fos (green) was performed and Oxt neurons were identified using hM3Dq-mCherry (red). Cell nuclei were counterstained with DAPI (blue). Scale bars, 50 μm. (B) The percentage of Oxt neurons expressing c-Fos in the PVN. ISO, isoprenaline. n = 7 (control, saline) or 6 (all other groups) mice per group. (C) Adult OxtCre mice were injected with adeno-associated viruses (AAVs) into the PVN. CAC was then induced using AOM and DSS. Subsequently, these mice were i.p. administered with CNO every other day and i.p. injected with saline or isoprenaline on a daily basis. These treatments were continued for 3 weeks (see also Figure 6—figure supplement 1C). Representative images of colorectal tissue after the treatments are shown. White eclipse outlines individual tumor. (D and E) Tumor number (D) and diameter (E). n = 7 (control, saline) or 6 (all other groups) mice per group. (F) Colorectal length. n = 7 (control, saline) or 6 (all other groups) mice per group. (G and H) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissue. Representative images (G) and the density of PCNA-positive cells (H) are shown. Scale bars, 50 μm. n = 4 (saline) or 5 (ISO) mice per group. (I and J) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay of tumor tissue. Representative images (I) and the density of TUNEL-positive cells (J) are shown. TUNEL labeling is in red. Cell nuclei were counterstained with DAPI (blue). Scale bars, 20 μm. n = 4 mice per group. Data are presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Bonferroni’s post hoc test (B, D, E, H, J).

Figure 6—source data 1. Source data for Figure 6, panels B, D-F, H and J.

Figure 6.

Figure 6—figure supplement 1. Body weight and food intake in mice.

Figure 6—figure supplement 1.

(A) Male C57 BL/6 mice (2 months of age) were i.p. administered with saline or isoprenaline (10 mg kg–1). Two hours later, celiac-superior mesenteric ganglion (CG-SMG) were dissected and fixed in 4% paraformaldehyde (PFA). Double immunofluorescence staining of c-Fos (red) and tyrosine hydroxylase (TH, in green) of the CG-SMG was performed. Cell nuclei were counterstained with DAPI (blue). Scale bars, 50 μm. (B) The percentage of TH-positive cells expressing c-Fos in the CG-SMG. n = 5 (saline) or 4 (isoprenaline) mice per group. (C) A schematic diagram illustrating the experimental design. The adult OxtCre mice were injected with control or hM3Dq adeno-associated viruses (AAVs) into the paraventricular nucleus (PVN). The colitis-associated cancer was then induced using azoxymethane (AOM) and dextran sodium sulfate (DSS). Subsequently, these mice were i.p. administered with clozapine-N-oxide (CNO) (3 mg kg–1) every other day, as well as saline or isoprenaline (ISO), a β2 adrenergic receptor agonist, on a daily basis. These treatments were continued for 3 weeks. (D and E) Body weight (D) and food intake (E) in mice throughout the experiment. n = 7 (control+ CNO + saline), 6 (control+ CNO + ISO), or 5 (hM3Dq) mice per group. Data are presented as means ± SEM.

Treatment with CNO and/or isoprenaline did not impact the body weight or food intake in control and hM3Dq AAV-injected mice (Figure 6—figure supplement 1D,E). Excitation of OxtPVN neurons suppressed CAC progression in mice, however, activation of β2AR with isoprenaline significantly abolished this effect (Figure 6C–E). Colorectal length was not significantly impacted in the mice administered with isoprenaline (Figure 6F). The histological data revealed that the effects of OxtPVN excitation on cell proliferation and cell apoptosis were dramatically attenuated when isoprenaline was administered (Figure 6G–J). Hence, activation of β2AR can significantly abrogate the anti-tumor effect of OxtPVN neuron activation.

Brain OTR is crucial for centrally administered celastrol to suppress CG-SMG neuronal activity

Our data indicated that Oxt neurons are important for celastrol to restrict CAC development in mice (Figure 5). Next, we asked whether i.c.v. administered celastrol could similarly regulate CG-SMG neuronal activity. To address this question, adult male C57 BL/6 mice were implanted with a guide cannula, and were then allowed to recover from surgeries. Subsequently, the preganglionic fiber of CG-SMG was transected, or left intact (sham). These mice were i.c.v. administered with vehicle versus celastrol. Two hours later, CG-SMG was dissected and fixed in 4% PFA. Double immunofluorescence staining for c-Fos and TH revealed that administration of celastrol suppressed the activity of sympathetic neurons in the CG-SMG. Notably, this effect was markedly diminished when the preganglionic nerve fiber of CG-SMG was transected (Figure 7—figure supplement 1A-C).

Thereafter, we asked whether brain OTR is crucial for centrally administered celastrol to suppress the CG-SMG neuronal activity. To this end, adult male C57 BL/6 mice were implanted with a guide cannula directed to the third ventricle. After surgical recovery, these mice were i.c.v. administered with vehicle control or L-368,899, the OTR antagonist, an hour before in vivo single-unit recordings. Subsequently, the 6 min control spiking activity was acquired before celastrol application through the guide cannula (Figure 7A and B). Single-unit spikes from 68 CG-SMG neurons (vehicle) and 44 CG-SMG neurons (OTR antagonist) were isolated, and the firing rates were compared before and after celastrol infusion (Figure 7C and D). Group data showed that treatment with celastrol significantly reduced the firing frequency of CG-SMG neurons, however, blockade of OTR abrogated this effect (Figure 7E). Scatterplot of mean firing frequency of individual CG-SMG neuron revealed a mixed modulation by celastrol (Figure 7F). The majority of CG-SMG neurons (63%) displayed a decreased firing frequency after celastrol infusion. Only a small proportion of neurons (18%) showed an increased firing frequency. The remainder (19%) maintained their activity level during celastrol infusion. However, when L-368,899 was applied, the majority of CG-SMG neurons (57%) maintained their activity level during celastrol infusion (Figure 7G), suggesting that blockade of OTR could attenuate the inhibitory effect of celastrol on neuronal firing rate in CG-SMG. Together, these data suggest that brain OTR is crucial for centrally administered celastrol to suppress the neuronal activity in the CG-SMG.

Figure 7. Treatment with isoprenaline abolished the anti-tumor effect of celastrol.

(A) Schematics of in vivo single-unit recordings in celiac-superior mesenteric ganglion (CG-SMG). L-368,899, the Oxt receptor (OTR) antagonist (OTR anta), and celastrol were applied through a guide cannula directed to third ventricle (3V). (B) A CG-SMG image was taken during the operation. (C) Example waveform of the single unit detected. (D) Normalized firing rate of recorded CG-SMG neurons in response to celastrol infusion in vehicle (top) and OTR antagonist (bottom) groups. Dashed line indicates the time point of celastrol delivery. i.c.v., intracerebroventricular injection. n = 68 cells (vehicle) from 7 mice or 44 cells (OTR antagonist) from 6 mice. (E) Statistics of average firing frequency of CG-SMG neurons in response to celastrol infusion in vehicle and OTR antagonist groups. Solid and dotted lines indicate medians and quartiles, respectively. n = 68 cells (vehicle) from 7 mice or 44 cells (OTR antagonist) from 6 mice. (F and G) Correlation of firing rate before and after celastrol infusion in vehicle (F) and OTR antagonist (G) groups. Green filled circles represent individual units with significantly lower firing frequency after celastrol infusion. Red triangles represent the units with higher firing rates. Gray squares indicate neurons without significant difference in firing rates. Inset: proportions of CG-SMG neurons with significantly decreased rates, increased rates, or no change in rates after celastrol infusion in vehicle (F) and OTR antagonist group (G). n = 68 cells (vehicle) from 7 mice or 44 cells (OTR antagonist) from 6 mice. (H and I) Colitis-associated cancer (CAC) was induced in male C57 BL/6 mice (2 months of age). These mice were then i.c.v. administered with vehicle (control) or celastrol every other day. In the meantime, the mice were i.p. injected with saline or isoprenaline (ISO) on a daily basis. These treatments were continued for 3 weeks (see also Figure 7—figure supplement 2A). Tumor number (H) and diameter (I) are shown. ns, not significant. n = 8 (saline) or 7 (ISO) mice per group. (J) The density of proliferating cell nuclear antigen (PCNA)-positive cells in tumor tissue. ns, not significant. n = 4 mice per group. (K) The density of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)-positive cells in tumor tissue. ns, not significant. n = 4 mice per group. Data are presented as means ± SD (C) or means ± SEM (H–K). *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Bonferroni’s post hoc test (E, H–K).

Figure 7—source data 1. Source data for Figure 7, panels E-K.

Figure 7.

Figure 7—figure supplement 1. The preganglionic nerve fiber is crucial for brain administered celastrol to suppress neuronal activities in celiac-superior mesenteric ganglion (CG-SMG).

Figure 7—figure supplement 1.

(A) A schematic diagram of the experimental design. Male C57 BL/6 mice (2 months of age) were implanted with a guide cannula directed to the third ventricle. After 2 weeks of recovery, the preganglionic nerve fiber of CG-SMG was transected. The other groups of mice were administered with sham operations. Subsequently, these mice were i.c.v. administered with vehicle (control) or celastrol. Two hours later, CG-SMG was dissected and fixed in 4% paraformaldehyde (PFA). (B) Double immunofluorescence staining for c-Fos (red) and tyrosine hydroxylase (TH, in green) of the CG-SMG. Cell nuclei were counterstained with DAPI (blue). Scale bars, 50 μm. (C) The percentage of TH-positive cells expressing c-Fos in the CG-SMG. ns, not significant. n = 5 mice per group. Data are presented as means ± SEM. *p < 0.05, one-way ANOVA with Bonferroni’s post hoc test (C).
Figure 7—figure supplement 2. Activation of β2 adrenergic receptor abolishes the tumor suppression effect of centrally administered celastrol.

Figure 7—figure supplement 2.

(A) A schematic diagram of the experimental design. The colitis-associated cancer was induced in the male C57 BL/6 mice (2 months of age) using azoxymethane (AOM) and dextran sodium sulfate (DSS). Subsequently, these mice were i.c.v. administered with vehicle (control) or celastrol (Cel) every other day. In the meanwhile, these mice were i.p. injected with saline or isoprenaline (ISO), a β2 adrenergic receptor agonist, on a daily basis. These treatments were continued for 3 weeks. (B and C) Body weight (B) and food intake (C) in mice throughout the experiment. n = 8 (saline) or 7 (ISO) mice per group. (D) Representative images of colorectal tissue. White eclipses outline the individual tumor. (E) Colorectal length. n = 8 (saline) or 7 (ISO) mice per group. (F) Immunohistochemical staining for proliferating cell nuclear antigen (PCNA) of tumor tissues. Scale bars, 50 μm. (G) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay of tumor tissues. TUNEL labeling is shown in red. Cell nuclei were counterstained with DAPI (blue). Scale bars, 20 μm. Data are presented as means ± SEM.

Agonism of β2AR abrogates the tumor suppressive effect of celastrol

Lastly, we interrogated whether the activation of β2AR could attenuate the anti-tumor effect of celastrol. To do so, the AOM/DSS-induced CAC mice were implanted with a guide cannula directed to the third ventricle. After recovery, these animals were i.c.v. administered with vehicle versus celastrol every other day for 3 weeks. Besides, these mice received daily saline or isoprenaline treatment (Figure 7—figure supplement 2A). Treatment with celastrol and/or isoprenaline did not impact the body weight or food intake in mice (Figure 7—figure supplement 2B,C). As anticipated, brain treatment with celastrol suppressed CAC progression in mice. Yet, treatment with isoprenaline significantly abrogated this effect (Figure 7H,I; Figure 7—figure supplement 2D). Treatment with celastrol and/or isoprenaline did not impact colorectal length (Figure 7—figure supplement 2E). The immunohistochemistry data revealed that treatment with celastrol inhibited cell proliferation, however, this effect was markedly attenuated when the mice were administered with isoprenaline (Figure 7J; Figure 7—figure supplement 2F). Besides, the TUNEL assay showed that the effect of brain treatment with celastrol on cell apoptosis was diminished when the mice were treated with isoprenaline (Figure 7K; Figure 7—figure supplement 2G). Overall, these data suggest that activation of β2AR can significantly abolish the anti-tumor effect of centrally administered celastrol.

Discussion

Negative mood is associated with the occurrences of cancers, however, the underlying mechanisms remain less well understood. In this study, we show that excitation of OxtPVN neurons remarkably ameliorated CAC progression in mice, and that this effect was mediated by inhibiting the neuronal activities in the CG-SMG. Also, brain treatment with celastrol suppressed the progression of CAC, and this effect required hypothalamic Oxt neurons. Moreover, we show that β2AR was involved in these processes. Together, our current work demonstrates that modulating hypothalamic Oxt neurons can impact the CAC progression in mice.

Negative moods, such as anxiety, depression, and stress, are implicated in tumor progression. As for CRC, a recent study has revealed a significant association of perceived stress with the incidences of rectal cancer (Kikuchi et al., 2017). Perceived stress at work and stressful life events elevated the risk of CRC (Azizi and Esmaeili, 2015; Blanc-Lapierre et al., 2017). Besides, stress is one of the key contributing factors to the onset and development of spontaneous colitis in humans (Mitchell and Drossman, 1987; Salem and Shubair, 1967). This association, in particular the one between chronic stress and colitis, was further confirmed in murine models (Gao et al., 2018; Reber et al., 2006; Reber et al., 2008). Moreover, chronic psychosocial stress was shown to result in the deterioration of CAC progression in mice (Peters et al., 2012). Hence, these findings suggest that stress is critical for CRC progression. Previous studies showed that Oxt has an anxiolytic effect in both humans (Heinrichs et al., 2003) and rodents (Blume et al., 2008; Ring et al., 2006; Windle et al., 1997). Conversely, our current and others’ previous studies (Amico et al., 2004; Mantella et al., 2003) demonstrated that disruption of Oxt neuron or Oxt gene increased anxiety level in mice. Importantly, we show that depletion of Oxt neuron promoted tumor progression in CAC mice, which agrees with the previous findings showing that increased stress level could promote colorectal tumor progression. Remarkably, our data indicated that chronic excitation of OxtPVN neurons or treatment with celastrol could significantly inhibit CAC progression in mice. These results are consistent with previous reports displaying that social support reduced the risk of colon cancer (Ikeda et al., 2013; Kinney et al., 2003). Besides, recent work demonstrated that Oxt has a prosocial role in humans (Kosfeld et al., 2005) and rodents (Lukas et al., 2011; Teng et al., 2013). Altogether, these findings suggest that the anxiolytic property of Oxt is critically important in its anti-tumor effect.

Previous studies unveiled a crucial role for nerve fiber in the tumorigenesis of various organs and tissues. For instance, both the densities of SNS and PNS nerve fibers were correlated with the aggressiveness of human prostate cancer (Magnon et al., 2013). Intriguingly, blockade of SNS activity suppressed the development of prostate cancer, whereas blockade of PNS activity inhibited the invasion and metastasis of prostate cancer in mice (Magnon et al., 2013). A further study indicated that norepinephrine released from SNS nerves drove angiogenesis in prostate cancer (Zahalka et al., 2017). Besides, a recent study showed that vagal innervation contributed to the development of stomach cancer via muscarinic acetylcholine M3 receptor (Zhao et al., 2014). Infiltration of nerve fibers was associated with the aggressiveness of breast cancer (Pundavela et al., 2015). The sensory neurons were able to facilitate the initiation and progression of pancreatic ductal adenocarcinoma in mice (Saloman et al., 2016). Together, these findings underscore an important role for nerve fiber of the autonomous nervous system in the initiation, invasion, or metastasis of cancers in peripheral organs, and hence the term ‘cancer neuroscience’ was coined (Demir et al., 2020; Monje et al., 2020). However, whether the CNS is similarly important remains largely unknown. In this work, we show that stimulation of OxtPVN neurons could suppress CAC progression in mice. Thus, in concert with other evidence (Cao et al., 2010; Liu et al., 2014), our current study implicates a critical role for the CNS, in particular the hypothalamus, in peripheral tumor development.

In summary, our current study indicates that chemogenetic stimulation of OxtPVN neurons or brain treatment with celastrol can suppress CAC progression in mice. The anti-tumor effect of celastrol requires hypothalamic Oxt neurons. Overall, these results suggest that modulating Oxt neuronal activity might be a relevant strategy for the treatment of CRC.

Materials and methods

Key resources table.

Reagent type (species) or resource Designation Source or reference Identifiers Additional information
Genetic reagent (Mus. musculus) OxtCre Jackson Laboratory 024234 PMID:23028821
Genetic reagent (Mus. musculus) Rosa26DTA176 PMID:16407399
Genetic reagent (Mus. musculus) Rosa26-LSL-EYFP PMID:11299042
Strain, strain background (AAV) AAV-hSyn-GFP Obio Technology AOV062
Strain, strain background (AAV) pAAV-hSyn-DIO-hM3Dq-mCherry Obio Technology HYMBH2482
Antibody (Rabbit polyclonal) anti-c-Fos Santa Cruz Biotechnology Cat# sc-7202; RRID:AB_2106765 IF, (1:150)
Antibody (Goat polyclonal) anti-c-Fos Santa Cruz Biotechnology Cat# sc-52-G; RRID:AB_2629503 IF, (1:25)
Antibody (Mouse monoclonal) anti-TH Santa Cruz Biotechnology Cat# sc-25269; RRID:AB_628422 IF, (1:200)
Antibody (Mouse monoclonal) anti-CD4 Santa Cruz Biotechnology Cat# sc-19641; RRID:AB_10554681 IHC, (1:50)
Antibody (Mouse monoclonal) anti-CD11b Santa Cruz Biotechnology Cat# sc-53086; RRID:AB_628894 IHC, (1:100)
Antibody (Rabbit polyclonal) anti-c-Fos Abcam Cat# ab190289; RRID:AB_2737414 IF, (1:2000)
Antibody (Rabbit polyclonal) anti-Oxt Immunostar Cat# 20068; RRID:AB_572258 IF, (1:400)
Antibody (Mouse monoclonal) anti-PCNA Boster Biological Cat# BM0104 IHC, (1:200)
Antibody (Rabbit polyclonal) anti-CD8α Bioss Cat# bs-0648R; RRID:AB_10857537 IHC, (1:250)
Antibody (Rat monoclonal) anti-B220 BD Biosciences Cat# 553087; RRID:AB_394617 IHC, (1:300)
Antibody (Mouse monoclonal) anti-NK1.1 BD Biosciences Cat# 550627; RRID:AB_398463 IHC, (1:400)
Commercial assay or kit Oxytocin EIA kit Enzo Life Sciences Cat# ADI-900–153 A; RRID:AB_2815012
Commercial assay or kit Corticosterone ELISA kit Enzo Life Sciences Cat# ADI-900–097; RRID:AB_2307314
Commercial assay or kit ACTH ELISA kit Aviva Systems Biology Cat# OKEH00628
Commercial assay or kit In Situ Cell Death Detection Kit, TMR red Sigma-Aldrich Cat# 12156792910
Commercial assay or kit SABC-POD kit Boster Biological Cat# SA1021
Chemical compound, drug Azoxymethane Sigma-Aldrich Cat# A5486
Chemical compound, drug Avertin Sigma-Aldrich Cat# T48402
Chemical compound, drug Isoprenaline Sigma-Aldrich Cat# I5627
Chemical compound, drug Proteinase K Sigma-Aldrich Cat# 3115879001
Chemical compound, drug Dextran sulfate sodium TdB Labs Cat# 9011-18-1
Chemical compound, drug CNO MedChemExpress Cat# HY-17366
Chemical compound, drug Celastrol Mengry Bio-Technology Cat# MR80328
Chemical compound, drug L-368,899 Santa Cruz Biotechnology Cat# sc-204037
Softwares, algorithm Pclamp 10 acquisition Molecular Devices
Softwares, algorithm OmniPlex neural recording data acquisition system Plexon
Softwares, algorithm Offline Sorter V4.0 Plexon
Softwares, algorithm Neuroexplorer V5.0 Plexon
Softwares, algorithm Matlab R2019b MathWorks
Softwares, algorithm Photoshop Adobe
Softwares, algorithm Prism 8 GraphPad Software RRID:SCR_002798
Softwares, algorithm ImageEP software PMID:19229173
Softwares, algorithm ImageLD software PMID:18704188
Softwares, algorithm ImageOF software https://cbsn.neuroinf.jp/modules/xoonips/detail.php?id=ImageOF

Mice

The OxtCre (Wu et al., 2012) mouse line was purchased from the Jackson Laboratory (Bar Harbor, ME). Rosa26DTA176 (Wu et al., 2006) and Rosa26-LSL-EYFP (Srinivas et al., 2001) mice have been described previously. We generated the OxtCre;Rosa26DTA176 mice by crossing the OxtCre with the Rosa26DTA176 mice, and the OxtCre;Rosa26-LSL-EYFP (OxtCre;EYFP) mice by crossing the OxtCre with the Rosa26-LSL-EYFP mice. C57 BL/6 mice were purchased from the Vital River Laboratory Animal Technology (Beijing, China). Rodent chow diet was purchased from HFK Bioscience (Beijing, China). All mice were housed in a 12-hr light/12-hr dark cycle in a temperature-controlled room (22–24°C).

Antibodies and chemicals

Rabbit and goat anti-c-Fos, mouse anti-TH, anti-CD4, and anti-CD11b antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti-c-Fos antibody was purchased from Abcam (Cambridge, UK). Rabbit anti-Oxt antibody was obtained from Immunostar (Hudson, WI). Mouse anti-PCNA antibody was purchased from Boster Biological (Wuhan, China). Rabbit anti-CD8α antibody was purchased from Bioss (Woburn, MA). Rat anti-B220 and mouse anti-NK1.1 antibodies were obtained from BD Biosciences (San Diego, CA). Alexa Fluor (AF) 488 goat anti-rabbit, AF 555 donkey anti-rabbit, AF 633 donkey anti-goat, and AF 488 donkey anti-mouse secondary antibodies were purchased from Thermo Fisher (Waltham, MA).

Azoxymethane, isoprenaline, and Avertin were purchased from Sigma-Aldrich (St Louis, MO). Dextran sulfate sodium was obtained from TdB Labs (Uppsala, Sweden). CNO was purchased from MedChemExpress (Monmouth Junction, NJ). Celastrol was obtained from Mengry Bio-Technology (Shanghai, China). L-368,899 was purchased from Santa Cruz Biotechnology.

AOM/DSS-induced CAC mouse model

Male mice were i.p. injected with the azoxymethane (12.5 mg kg–1). A week later, mice were administrated with two cycles of 5-day oral exposure to DSS (2.5% in drinking water) and then 16-day normal drinking water (Neufert et al., 2007).

Stereotaxic surgery

Third ventricle cannulation: The procedures have been described before (Wu et al., 2017; Zhang et al., 2008). Briefly, mice were anesthetized with Avertin (300 mg kg–1) and were then placed on an ultra-precise stereotaxic instrument (David Kopf, Tujunga, CA). Next, a guide cannula (RWD Life Science, Shenzhen, China) was placed directed to third ventricle (coordinates: A/P –2.0 mm posterior to bregma, D/V –5.0 mm). Mice were allowed to fully recover from surgeries.

For AAV injection, mice were anesthetized and placed on the stereotaxic instrument. With the help of a guide cannula, viral solution was injected bilaterally into the PVN (coordinates: A/P, –0.85 mm posterior to bregma, M/L, ± 0.2 mm, D/V, –4.8 mm).

Chemogenetics

AAVs carrying GFP (AAV-hSyn-GFP) or Cre-dependent hM3Dq-mCherry (AAV-hSyn-DIO-hM3Dq-mCherry) were purchased from Obio Technology (Shanghai, China). Adult male OxtCre mice were bilaterally injected with AAVs into the PVN, and were then allowed to recover from surgeries. After the induction of CAC, mice were i.p. administered with CNO (3 mg kg–1, every other day for 3 weeks) to activate the hM3Dq-expressing Oxt neurons.

Treatments

Treatment with CNO and L-368,899: The control and hM3Dq AAVs were injected into the PVN of adult OxtCre mice. CAC was induced using AOM and DSS. These mice were i.p. injected with CNO and i.c.v. administered with vehicle or L-368,899 (2 µg per mouse) every other day for 3 weeks. Body weight and food intake in mice were recorded throughout the experiment.

Celastrol: Adult male C57 BL/6 mice bearing AOM and DSS-induced CAC were implanted with a guide cannula directed to the third ventricle, and were then allowed to recover from surgeries. aCSF and celastrol (0.5 µg per mouse) was i.c.v. administered every other day for 3 weeks. In a separate experiment, adult male and female OxtCre and OxtCre;DTA mice were administered with AOM and DSS to induce CAC, and were then i.p. injected with vehicle (1% DMSO in saline) or celastrol (1 mg kg–1) every other day for 3 weeks. Body weight and food intake were regularly assessed throughout the experiment.

Treatment with CNO and isoprenaline: The control and hM3Dq AAVs were injected into the PVN of male OxtCre mice, in which CAC was then induced. These mice were i.p. administered with CNO (3 mg kg–1) every other day for 3 weeks. During this period, saline and isoprenaline (10 mg kg–1) were i.p. administered on a daily basis. Body weight and food intake in mice were assessed.

Treatment with celastrol and isoprenaline: Adult male C57 BL/6 mice bearing CAC were i.c.v. administered with vehicle or celastrol (0.5 µg per mouse) every other day for 3 weeks. In the meanwhile, these mice were i.p. injected with saline or isoprenaline (10 mg kg–1) on a daily basis. Body weight and food intake in mice were measured.

Removal of CG-SMG, and the transection of its preganglionic nerve fiber

Mice were anesthetized using Avertin, and then the abdomen was cut open. Abdominal viscera were gently pulled out and held in warm sterile saline-soaked gauze. The intersection of the descending aorta and the left renal artery was identified, where the superior mesenteric artery was located. The CG-SMG is wrapped around the superior mesenteric artery and associated lymphatic vessels. Fine forceps and microdissection scissor were used to remove CG-SMG or transect its preganglionic nerve fiber.

Slice electrophysiology

The OxtCre;EYFP mice (4 months of age) were euthanized with an overdose of sodium pentobarbital (40 mg kg–1, i.p.). Coronal PVN slices (300 μm in thickness) were cut in a solution containing (in mM): 228 sucrose, 26 NaHCO3, 11 glucose, 2.5 KCl, 1 NaH2PO4, 7 MgSO4, and 0.5 CaCl2, and recovered in aCSF containing (in mM): 119 NaCl, 26 NaHCO3, 11 glucose, 2.5 KCl, 1 NaH2PO4, 1.3 MgSO4, and 2.5 CaCl2. Recordings were performed in a submerged-style chamber mounted under an infrared-differential interference contrast microscope (BX-51 WI, Olympus, Tokyo, Japan). Slices were constantly perfused with heated aCSF (35°C) and bubbled continuously with 95% O2 and 5% CO2. Oxt neurons were identified by EYFP epifluorescence. Whole-cell recordings were achieved using a Multiclamp 700B amplifier (Molecular Devices, San Jose, CA). Signals were filtered at 10 kHz, and then sampled by Digidata 1550B4 (Molecular Devices) at 20 kHz using Clampex 10 acquisition software. The pipette resistance was about 4–6 MΩ with an internal solution containing (in mM): 135 K-gluconate, 8 KCl, 10 HEPES, 0.25 EGTA, 2 MgATP, 0.3 Na3GTP, 0.1 spermine, 7 phospho-creatine (pH 7.25–7.3; osmolarity 294–298). For celastrol condition, celastrol (5 μM) was added to the incubation chamber 20 min prior to recording and was added in bath aCSF throughout recording. Liquid junction potential (16 mV) has been corrected in the text and figures.

In vivo single-unit recordings

Male mice (8 weeks of age) were implanted with a guide cannula directed to the third ventricle. Two weeks later, in vivo single-unit recordings were performed and analyzed as described previously (Tseng et al., 2011). The guide tubes housed 16-channel electrodes using 25.4 μm formvar-insulated nichrome wire (761500, A-M System, Sequim, WA). The final impedance of the electrodes was 700–800 kΩ. On the recording day, the CG-SMG located at the intersection of the descending aorta and left renal artery was identified, and the 16-channel electrodes were manually placed into CG-SMG. A sterile cotton swab was dipped in saline solution, and was then placed by the CG-SMG to maintain tissue humidity during recording. Spiking activities were digitized at 40 kHz, bandpass-filtered from 250 to 8000 Hz, and stored on a PC for further offline analysis.

For administration of celastrol and L-368,899, the C57 BL/6 mice were implanted with an infusion cannula directed to third ventricle and were then singly housed to allow recovery from surgeries. On the recording day, aCSF and L-368,899 were applied through the pre-implanted cannula 1 hr before recordings. The 6 min control (5% DMSO in aCSF) spiking activity was acquired before celastrol (0.5 µg per mouse) application through the infusion cannula.

In the CG-SMG preganglionic nerve fiber transection experiment, adult OxtCre mice were injected with control or hM3Dq AAV into the PVN. These mice were also implanted with an infusion cannula directed to third ventricle. After recovery, the preganglionic nerve fiber of CG-SMG was transected before recording. In the control group, sham operations were carried out before recording. Subsequently, the 6 min control (1% DMSO in aCSF) spiking activity was acquired before CNO (1 µg per mouse) application through the infusion cannula.

In vivo single-unit recordings data analysis

The single-unit spike sorting was performed with Offline Sorter V4.0 (Plexon, Dallas, TX). Spikes were detected when a minimum waveform reached an amplitude threshold of –4.50 standard deviation greater than the noise amplitude. Principal component analysis and automatic scan were employed to separate single-unit waveforms into individual clusters. Manual checking was then performed to ensure that the cluster boundaries were clearly separated. All isolated single units exhibited recognizable refractory periods (>1 ms) in the inter-spike interval histograms. Only well-isolated units (L ratio <0.2, isolation distance >15) were included in the data analysis.

The response of single unit was analyzed with Neuroexplorer V5.0 (Plexon). Well-separated units were used to analyze the responses before (baseline) and after celastrol or CNO infusion. Firing rates of neurons during baseline, 10 and 20 min after celastrol or CNO infusion were compared to determine the significance of difference in firing rates (paired Student’s t-test, 95% confidence interval). For heatmap analysis, z-score of each bin (10 s) was calculated by the following equation: z = (x-μ)/σ, in which x is the raw firing rate, μ is the mean firing rate during the baseline period, and σ is the corresponding standard deviation. Further normalization was utilized for better presentation. All of the single-unit z-scores were plotted using Matlab R2019b (Natick, MA).

Immunofluorescence

The detailed procedures have been described previously (Shen et al., 2020). Mice were anesthetized using Avertin, and were then transcardially perfused with 4% PFA. Mouse brains were removed, post-fixed in 4% PFA, and infiltrated with 20–30% sucrose solutions. Brain tissues were sectioned using a cryostat. Tissue sections were washed with phosphate buffered saline (PBS), blocked with 5% serum/0.3% Triton X-100/PBS for 30 min, incubated with primary antibodies at 4 °C overnight, and fluorophore-conjugated secondary antibodies at room temperature for 1 hr. Cell nuclei were counterstained with DAPI.

Immunofluorescence staining of CG-SMG: Mice were euthanized, and then the CG-SMG were dissected, fixed in 4% PFA for 10 min. The tissues were infiltrated with 75–100% ethanol, and were then embedded in paraffin and sectioned (thickness: 3 μm). The tissue sections were deparaffinized and rehydrated using graded ethanol. Antigen retrieval was then performed. Tissue sections were washed with 1× PBS, blocked with 5% serum/0.3% Triton X-100/PBS for 30 min, incubated with primary antibodies at 4°C overnight, and fluorophore-conjugated secondary antibodies at room temperature for 1 hr. Cell nuclei were counterstained with DAPI. Images were acquired with the LSM 780 confocal microscope (Carl Zeiss, Jena, Germany). Cells were manually counted in one representative image collected for each mouse.

Immunohistochemistry

Paraffin-embedded tissue sections were deparaffinized, rehydrated, and antigen-recovered. Sections were then blocked with 5% serum/0.3% Triton X-100/PBS for 30 min, incubated with primary antibodies at 4°C overnight and followed by a reaction using a SABC-POD kit (Boster Biological). Images were acquired using an IX71 microscope (Olympus). Cells were counted using Photoshop (Adobe, San Jose, CA).

TUNEL assay

The In Situ Cell Death Detection Kit was purchased from Sigma-Aldrich. Paraffin-embedded tissue sections were deparaffinized and rehydrated. Next, tissue sections were rinsed in distilled water, incubated with proteinase K (18.5 µg ml–1 in 10 mM Tris·HCl) at 37°C for 15 min, washed with 1× PBS, and were then incubated with TUNEL reaction mixture in the humidified chamber at 37 °C for 1 hr. Cell nuclei were counterstained with DAPI. Images were acquired with the LSM 780 confocal microscope. TUNEL-positive cells were manually counted using Photoshop.

Behavioral analyses

Open field test: Adult male OxtCre, OxtCre;DTA mice, and the OxtCre mice injected with control or hM3Dq AAV were placed in an opaque, square open field (40 cm L × 40 cm W × 40 cm H), and were then allowed to freely explore for 5 min and monitored with the ImageOF software (https://cbsn.neuroinf.jp/modules/xoonips/detail.php?id=ImageOF). The open field was divided into a peripheral region and a 13.3 cm × 13.3 cm central region. Time spent in the central versus peripheral region during the test was presented.

Elevated plus maze test: the plus maze had two closed arms (35 cm L × 6 cm W × 22 cm H) and two open arms (35 cm L × 6 cm W). The maze was elevated 74 cm from the floor. Mice were placed on the center section and allowed to explore the maze freely and monitored with ImageEP software (Komada et al., 2008). Time spent in the open versus closed arms during the 5 min period was presented.

Light/dark box test: The apparatus was comprised of a dual compartment box (20 cm L × 20 cm W × 40 cm H) with free access between them. The dark box was made of black Plexiglass and the light one was exposed to room light. The exploratory activity was monitored for 5 min using the ImageLD software (Takao and Miyakawa, 2006). Time spent in the light versus dark box was presented.

Oxt release assay

The detailed procedures have been described previously (Zhang et al., 2011). In order to determine the effect of celastrol on Oxt release, PVN tissue slices were dissected from the brain of C57 BL/6 mice and were balanced in normal Locke’s solution supplied with 95% O2 and 5% CO2 at 37 °C. The solution was changed every 5 min for 10 times and the 11th sample was collected to measure the basal Oxt release rate. The slices were then incubated in the same solution containing celastrol (5 μM) for 5 min and this solution was measured to determine the Oxt release rate under celastrol condition. An oxytocin EIA kit (Enzo Life Sciences, Farmingdale, NY) was used to determine the Oxt concentration in the solutions.

Plasma Oxt, ACTH, and corticosterone assays

The plasma was collected from mice after the completion of the experiments. Plasma Oxt and corticosterone levels were determined using the Oxt EIA kit and a corticosterone ELISA kit (Enzo Life Sciences), respectively. Plasma ACTH was assessed using an ACTH ELISA kit (Aviva Systems Biology, San Diego, CA).

Statistical analysis

All data are presented as means ± SEM unless otherwise specified. Sample sizes with sufficient power were determined according to our published studies and relevant literature. Animals were assigned to specific experimental groups without bias. Data were analyzed using Prism 8 (GraphPad Software, San Diego, CA) or Matlab R2019b. Data distribution was assumed to be normal but this was not formally tested. Two-group comparisons were assessed using two-tailed Student’s t-test. One-way and two-way analysis of variance (ANOVA) with Bonferroni’s post hoc test was used for comparisons of more than two groups. Key experiments were repeated at least twice independently. No data were excluded from the analyses. When necessary, experimental performers were blind to group information before data were obtained. A p-value of less than 0.05 was considered statistically significant.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (81573146 and 91539125 to GZ, 81972767 to ML, 31871089 to YH, and 31871028 to JM).

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

Mei Liu, Email: liumei@cicams.ac.cn.

Man Jiang, Email: manjiang@hust.edu.cn.

Ningzhi Xu, Email: xuningzhi@cicams.ac.cn.

Guo Zhang, Email: gzhang@hust.edu.cn.

Ernie Blevins, University of Washington, Seattle, United States.

Mone Zaidi, Icahn School of Medicine at Mount Sinai, United States.

Funding Information

This paper was supported by the following grants:

  • National Natural Science Foundation of China 81573146 to Guo Zhang.

  • National Natural Science Foundation of China 91539125 to Guo Zhang.

  • National Natural Science Foundation of China 81972767 to Mei Liu.

  • National Natural Science Foundation of China 31871089 to Yunyun Han.

  • National Natural Science Foundation of China 31871028 to Man Jiang.

Additional information

Competing interests

No competing interests declared.

Author contributions

Data curation, Formal analysis, Methodology, Visualization, Writing - original draft.

Data curation, Formal analysis, Methodology.

Slice electrophysiology and in vivo single-unit recordings.

Data curation, Formal analysis, Methodology.

Data curation, Formal analysis.

Slice electrophysiology and in vivo single-unit recordings.

Formal analysis, Writing – review and editing.

Formal analysis, Funding acquisition.

Formal analysis, Funding acquisition, Methodology, Writing – review and editing.

Formal analysis, Funding acquisition, Slice electrophysiology and in vivo single-unit recordings, Visualization, Writing – review and editing.

Formal analysis, Resources, Supervision, Writing – review and editing.

Conceptualization, Funding acquisition, Supervision, Writing – review and editing.

Ethics

Animal procedures were approved by the IACUC at Huazhong University of Science and Technology (#2511).

Additional files

Transparent reporting form

Data availability

All data that support the findings of this study are included in this published article and its supplementary files. Source data files have been provided for Figures 1-7.

References

  1. Amico JA, Mantella RC, Vollmer RR, Li X. Anxiety and stress responses in female oxytocin deficient mice. Journal of Neuroendocrinology. 2004;16:319–324. doi: 10.1111/j.0953-8194.2004.01161.x. [DOI] [PubMed] [Google Scholar]
  2. Antoni MH, Lutgendorf SK, Cole SW, Dhabhar FS, Sephton SE, McDonald PG, Stefanek M, Sood AK. The influence of bio-behavioural factors on tumour biology: pathways and mechanisms. Nature Reviews. Cancer. 2006;6:240–248. doi: 10.1038/nrc1820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Azizi H, Esmaeili ED. Stressful life events and risk of colorectal cancer: a case-control study of Iran. Asian Pacific Journal of Cancer Prevention. 2015;16:2403–2407. doi: 10.7314/apjcp.2015.16.6.2403. [DOI] [PubMed] [Google Scholar]
  4. Blanc-Lapierre A, Rousseau MC, Weiss D, El-Zein M, Siemiatycki J, Parent ME. Lifetime report of perceived stress at work and cancer among men: A case-control study in Montreal, Canada. Preventive Medicine. 2017;96:28–35. doi: 10.1016/j.ypmed.2016.12.004. [DOI] [PubMed] [Google Scholar]
  5. Blume A, Bosch OJ, Miklos S, Torner L, Wales L, Waldherr M, Neumann ID. Oxytocin reduces anxiety via ERK1/2 activation: local effect within the rat hypothalamic paraventricular nucleus. The European Journal of Neuroscience. 2008;27:1947–1956. doi: 10.1111/j.1460-9568.2008.06184.x. [DOI] [PubMed] [Google Scholar]
  6. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for Clinicians. 2018;68:394–424. doi: 10.3322/caac.21492. [DOI] [PubMed] [Google Scholar]
  7. Cao L, Liu X, Lin EJ, Wang C, Choi EY, Riban V, Lin B, During MJ. Environmental and genetic activation of a brain-adipocyte BDNF/leptin axis causes cancer remission and inhibition. Cell. 2010;142:52–64. doi: 10.1016/j.cell.2010.05.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cao M, Li H, Sun D, Chen W. Cancer burden of major cancers in China: A need for sustainable actions. Cancer Communications. 2020;40:205–210. doi: 10.1002/cac2.12025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Delgado-Guay M, Parsons HA, Li Z, Palmer JL, Bruera E. Symptom distress in advanced cancer patients with anxiety and depression in the palliative care setting. Supportive Care in Cancer. 2009;17:573–579. doi: 10.1007/s00520-008-0529-7. [DOI] [PubMed] [Google Scholar]
  10. Demir IE, Reyes CM, Alrawashdeh W, Ceyhan GO, Deborde S, Friess H, Gorgulu K, Istvanffy R, Jungwirth D, Kuner R, Maryanovich M, Na’ara S, Renders S, Saloman JL, Scheff NN, Steenfadt H, Stupakov P, Thiel V, Verma D, Yilmaz BS, White RA, Wang TC, Wong RJ, Frenette PS, Gil Z, Davis BM. Clinically actionable strategies for studying neural influences in cancer. Cancer Cell. 2020;38:11–14. doi: 10.1016/j.ccell.2020.05.023. [DOI] [PubMed] [Google Scholar]
  11. Gao X, Cao Q, Cheng Y, Zhao D, Wang Z, Yang H, Wu Q, You L, Wang Y, Lin Y, Li X, Wang Y, Bian JS, Sun D, Kong L, Birnbaumer L, Yang Y. Chronic stress promotes colitis by disturbing the gut microbiota and triggering immune system response. PNAS. 2018;115:E2960–E2969. doi: 10.1073/pnas.1720696115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Grund T, Tang Y, Benusiglio D, Althammer F, Probst S, Oppenlander L, Neumann ID, Grinevich V. Chemogenetic activation of oxytocin neurons: Temporal dynamics, hormonal release, and behavioral consequences. Psychoneuroendocrinology. 2019;106:77–84. doi: 10.1016/j.psyneuen.2019.03.019. [DOI] [PubMed] [Google Scholar]
  13. Hayakawa Y, Sakitani K, Konishi M, Asfaha S, Niikura R, Tomita H, Renz BW, Tailor Y, Macchini M, Middelhoff M, Jiang Z, Tanaka T, Dubeykovskaya ZA, Kim W, Chen X, Urbanska AM, Nagar K, Westphalen CB, Quante M, Lin CS, Gershon MD, Hara A, Zhao CM, Chen D, Worthley DL, Koike K, Wang TC. Nerve growth factor promotes gastric tumorigenesis through aberrant cholinergic signaling. Cancer Cell. 2017;31:21–34. doi: 10.1016/j.ccell.2016.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Heinrichs M, Baumgartner T, Kirschbaum C, Ehlert U. Social support and oxytocin interact to suppress cortisol and subjective responses to psychosocial stress. Biological Psychiatry. 2003;54:1389–1398. doi: 10.1016/s0006-3223(03)00465-7. [DOI] [PubMed] [Google Scholar]
  15. Ikeda A, Kawachi I, Iso H, Iwasaki M, Inoue M, Tsugane S. Social support and cancer incidence and mortality: the JPHC study cohort II. Cancer Causes & Control. 2013;24:847–860. doi: 10.1007/s10552-013-0147-7. [DOI] [PubMed] [Google Scholar]
  16. Kamiya A, Hayama Y, Kato S, Shimomura A, Shimomura T, Irie K, Kaneko R, Yanagawa Y, Kobayashi K, Ochiya T. Genetic manipulation of autonomic nerve fiber innervation and activity and its effect on breast cancer progression. Nature Neuroscience. 2019;22:1289–1305. doi: 10.1038/s41593-019-0430-3. [DOI] [PubMed] [Google Scholar]
  17. Kikuchi N, Nishiyama T, Sawada T, Wang C, Lin Y, Watanabe Y, Tamakoshi A, Kikuchi S. Perceived stress and colorectal cancer incidence: the Japan collaborative cohort study. Scientific Reports. 2017;7:40363. doi: 10.1038/srep40363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kinney AY, Bloor LE, Dudley WN, Millikan RC, Marshall E, Martin C, Sandler RS. Roles of religious involvement and social support in the risk of colon cancer among Blacks and Whites. American Journal of Epidemiology. 2003;158:1097–1107. doi: 10.1093/aje/kwg264. [DOI] [PubMed] [Google Scholar]
  19. Kojima M, Wakai K, Tokudome S, Tamakoshi K, Toyoshima H, Watanabe Y, Hayakawa N, Suzuki K, Hashimoto S, Kawado M, Suzuki S, Ito Y, Tamakoshi A, Group JS. Perceived psychologic stress and colorectal cancer mortality: findings from the Japan Collaborative Cohort Study. Psychosomatic Medicine. 2005;67:72–77. doi: 10.1097/01.psy.0000151742.43774.6d. [DOI] [PubMed] [Google Scholar]
  20. Komada M, Takao K, Miyakawa T. Elevated plus maze for mice. Journal of Visualized Experiments. 2008;22:1088. doi: 10.3791/1088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kosfeld M, Heinrichs M, Zak PJ, Fischbacher U, Fehr E. Oxytocin increases trust in humans. Nature. 2005;435:673–676. doi: 10.1038/nature03701. [DOI] [PubMed] [Google Scholar]
  22. Lillberg K, Verkasalo PK, Kaprio J, Teppo L, Helenius H, Koskenvuo M. Stressful life events and risk of breast cancer in 10,808 women: a cohort study. American Journal of Epidemiology. 2003;157:415–423. doi: 10.1093/aje/kwg002. [DOI] [PubMed] [Google Scholar]
  23. Liu X, McMurphy T, Xiao R, Slater A, Huang W, Cao L. Hypothalamic gene transfer of BDNF inhibits breast cancer progression and metastasis in middle age obese mice. Molecular Therapy. 2014;22:1275–1284. doi: 10.1038/mt.2014.45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Liu J, Lee J, Salazar Hernandez MA, Mazitschek R, Ozcan U. Treatment of obesity with celastrol. Cell. 2015;161:999–1011. doi: 10.1016/j.cell.2015.05.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lukas M, Toth I, Reber SO, Slattery DA, Veenema AH, Neumann ID. The neuropeptide oxytocin facilitates pro-social behavior and prevents social avoidance in rats and mice. Neuropsychopharmacology. 2011;36:2159–2168. doi: 10.1038/npp.2011.95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lutgendorf SK, Lamkin DM, DeGeest K, Anderson B, Dao M, McGinn S, Zimmerman B, Maiseri H, Sood AK, Lubaroff DM. Depressed and anxious mood and T-cell cytokine expressing populations in ovarian cancer patients. Brain, Behavior, and Immunity. 2008;22:890–900. doi: 10.1016/j.bbi.2007.12.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ma X, Xu L, Alberobello AT, Gavrilova O, Bagattin A, Skarulis M, Liu J, Finkel T, Mueller E. Celastrol protects against obesity and metabolic dysfunction through activation of a HSF1-PGC1α transcriptional axis. Cell Metabolism. 2015;22:695–708. doi: 10.1016/j.cmet.2015.08.005. [DOI] [PubMed] [Google Scholar]
  28. Magnon C, Hall SJ, Lin J, Xue X, Gerber L, Freedland SJ, Frenette PS. Autonomic nerve development contributes to prostate cancer progression. Science. 2013;341:1236361. doi: 10.1126/science.1236361. [DOI] [PubMed] [Google Scholar]
  29. Mantella RC, Vollmer RR, Li X, Amico JA. Female oxytocin-deficient mice display enhanced anxiety-related behavior. Endocrinology. 2003;144:2291–2296. doi: 10.1210/en.2002-0197. [DOI] [PubMed] [Google Scholar]
  30. Miller KD, Nogueira L, Mariotto AB, Rowland JH, Yabroff KR, Alfano CM, Jemal A, Kramer JL, Siegel RL. Cancer treatment and survivorship statistics, 2019. CA: A Cancer Journal for Clinicians. 2019;69:363–385. doi: 10.3322/caac.21565. [DOI] [PubMed] [Google Scholar]
  31. Mitchell CM, Drossman DA. Survey of the AGA membership relating to patients with functional gastrointestinal disorders. Gastroenterology. 1987;92:1282–1284. doi: 10.1016/s0016-5085(87)91099-7. [DOI] [PubMed] [Google Scholar]
  32. Monje M, Borniger JC, D’Silva NJ, Deneen B, Dirks PB, Fattahi F, Frenette PS, Garzia L, Gutmann DH, Hanahan D, Hervey-Jumper SL, Hondermarck H, Hurov JB, Kepecs A, Knox SM, Lloyd AC, Magnon C, Saloman JL, Segal RA, Sloan EK, Sun X, Taylor MD, Tracey KJ, Trotman LC, Tuveson DA, Wang TC, White RA, Winkler F. Roadmap for the emerging field of cancer neuroscience. Cell. 2020;181:219–222. doi: 10.1016/j.cell.2020.03.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Neufert C, Becker C, Neurath MF. An inducible mouse model of colon carcinogenesis for the analysis of sporadic and inflammation-driven tumor progression. Nature Protocols. 2007;2:1998. doi: 10.1038/nprot.2007.279. [DOI] [PubMed] [Google Scholar]
  34. Neumann ID. Brain oxytocin: a key regulator of emotional and social behaviours in both females and males. Journal of Neuroendocrinology. 2008;20:858–865. doi: 10.1111/j.1365-2826.2008.01726.x. [DOI] [PubMed] [Google Scholar]
  35. Neumann ID, Landgraf R. Balance of brain oxytocin and vasopressin: implications for anxiety, depression, and social behaviors. Trends in Neurosciences. 2012;35:649–659. doi: 10.1016/j.tins.2012.08.004. [DOI] [PubMed] [Google Scholar]
  36. Nielsen NR, Kristensen TS, Strandberg-Larsen K, Zhang ZF, Schnohr P, Gronbaek M. Perceived stress and risk of colorectal cancer in men and women: a prospective cohort study. Journal of Internal Medicine. 2008;263:192–202. doi: 10.1111/j.1365-2796.2007.01826.x. [DOI] [PubMed] [Google Scholar]
  37. Peters S, Grunwald N, Rümmele P, Endlicher E, Lechner A, Neumann ID, Obermeier F, Reber SO. Chronic psychosocial stress increases the risk for inflammation-related colon carcinogenesis in male mice. Stress. 2012;15:403–415. doi: 10.3109/10253890.2011.631232. [DOI] [PubMed] [Google Scholar]
  38. Price JL, Drevets WC. Neurocircuitry of mood disorders. Neuropsychopharmacology. 2010;35:192–216. doi: 10.1038/npp.2009.104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Pundavela J, Roselli S, Faulkner S, Attia J, Scott RJ, Thorne RF, Forbes JF, Bradshaw RA, Walker MM, Jobling P, Hondermarck H. Nerve fibers infiltrate the tumor microenvironment and are associated with nerve growth factor production and lymph node invasion in breast cancer. Molecular Oncology. 2015;9:1626–1635. doi: 10.1016/j.molonc.2015.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Reber SO, Obermeier F, Straub RH, Falk W, Neumann ID. Chronic intermittent psychosocial stress (social defeat/overcrowding) in mice increases the severity of an acute DSS-induced colitis and impairs regeneration. Endocrinology. 2006;147:4968–4976. doi: 10.1210/en.2006-0347. [DOI] [PubMed] [Google Scholar]
  41. Reber SO, Obermeier F, Straub RH, Veenema AH, Neumann ID. Aggravation of DSS-induced colitis after chronic subordinate colony (CSC) housing is partially mediated by adrenal mechanisms. Stress. 2008;11:225–234. doi: 10.1080/10253890701733351. [DOI] [PubMed] [Google Scholar]
  42. Ring RH, Malberg JE, Potestio L, Ping J, Boikess S, Luo B, Schechter LE, Rizzo S, Rahman Z, Rosenzweig-Lipson S. Anxiolytic-like activity of oxytocin in male mice: behavioral and autonomic evidence, therapeutic implications. Psychopharmacology. 2006;185:218–225. doi: 10.1007/s00213-005-0293-z. [DOI] [PubMed] [Google Scholar]
  43. Roth BL. DREADDs for Neuroscientists. Neuron. 2016;89:683–694. doi: 10.1016/j.neuron.2016.01.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Salem SN, Shubair KS. Non-specific ulcerative colitis in Bedouin Arabs. Lancet. 1967;1:473–475. doi: 10.1016/s0140-6736(67)91094-x. [DOI] [PubMed] [Google Scholar]
  45. Saloman JL, Albers KM, Li D, Hartman DJ, Crawford HC, Muha EA, Rhim AD, Davis BM. Ablation of sensory neurons in a genetic model of pancreatic ductal adenocarcinoma slows initiation and progression of cancer. PNAS. 2016;113:3078–3083. doi: 10.1073/pnas.1512603113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Schindler S, Geyer S, Strauss M, Anwander A, Hegerl U, Turner R, Schonknecht P. Structural studies of the hypothalamus and its nuclei in mood disorders. Psychiatry Research. 2012;201:1–9. doi: 10.1016/j.pscychresns.2011.06.005. [DOI] [PubMed] [Google Scholar]
  47. Shen Q, Chen Z, Zhao F, Pan S, Zhang T, Cheng X, Zhang L, Zhang S, Qi J, Li J, Cai D, Zhang G. Reversal of prolonged obesity-associated cerebrovascular dysfunction by inhibiting microglial Tak1. Nature Neuroscience. 2020;23:832–841. doi: 10.1038/s41593-020-0642-6. [DOI] [PubMed] [Google Scholar]
  48. Siegel RL, Miller KD, Goding Sauer A, Fedewa SA, Butterly LF, Anderson JC, Cercek A, Smith RA, Jemal A. Colorectal cancer statistics, 2020. CA: A Cancer Journal for Clinicians. 2020;70:145–164. doi: 10.3322/caac.21601. [DOI] [PubMed] [Google Scholar]
  49. Smith KS, Bucci DJ, Luikart BW, Mahler SV. DREADDS: Use and application in behavioral neuroscience. Behavioral Neuroscience. 2016;130:137–155. doi: 10.1037/bne0000135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Developmental Biology. 2001;1:4. doi: 10.1186/1471-213x-1-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Takao K, Miyakawa T. Light/dark transition test for mice. Journal of Visualized Experiments. 2006;1:104. doi: 10.3791/104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Teng BL, Nonneman RJ, Agster KL, Nikolova VD, Davis TT, Riddick NV, Baker LK, Pedersen CA, Jarstfer MB, Moy SS. Prosocial effects of oxytocin in two mouse models of autism spectrum disorders. Neuropharmacology. 2013;72:187–196. doi: 10.1016/j.neuropharm.2013.04.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Tseng WT, Yen CT, Tsai ML. A bundled microwire array for long-term chronic single-unit recording in deep brain regions of behaving rats. Journal of Neuroscience Methods. 2011;201:368–376. doi: 10.1016/j.jneumeth.2011.08.028. [DOI] [PubMed] [Google Scholar]
  54. Windle RJ, Shanks N, Lightman SL, Ingram CD. Central oxytocin administration reduces stress-induced corticosterone release and anxiety behavior in rats. Endocrinology. 1997;138:2829–2834. doi: 10.1210/endo.138.7.5255. [DOI] [PubMed] [Google Scholar]
  55. Wu S, Wu Y, Capecchi MR. Motoneurons and oligodendrocytes are sequentially generated from neural stem cells but do not appear to share common lineage-restricted progenitors in vivo. Development. 2006;133:581–590. doi: 10.1242/dev.02236. [DOI] [PubMed] [Google Scholar]
  56. Wu Z, Xu Y, Zhu Y, Sutton AK, Zhao R, Lowell BB, Olson DP, Tong Q. An obligate role of oxytocin neurons in diet induced energy expenditure. PLOS ONE. 2012;7:e45167. doi: 10.1371/journal.pone.0045167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wu L, Meng J, Shen Q, Zhang Y, Pan S, Chen Z, Zhu LQ, Lu Y, Huang Y, Zhang G. Caffeine inhibits hypothalamic A1R to excite oxytocin neuron and ameliorate dietary obesity in mice. Nature Communications. 2017;8:15904. doi: 10.1038/ncomms15904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zahalka AH, Arnal-Estape A, Maryanovich M, Nakahara F, Cruz CD, Finley LWS, Frenette PS. Adrenergic nerves activate an angio-metabolic switch in prostate cancer. Science. 2017;358:321–326. doi: 10.1126/science.aah5072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhang X, Zhang G, Zhang H, Karin M, Bai H, Cai D. Hypothalamic IKKβ/NF-κB and ER stress link overnutrition to energy imbalance and obesity. Cell. 2008;135:61–73. doi: 10.1016/j.cell.2008.07.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Zhang G, Bai H, Zhang H, Dean C, Wu Q, Li J, Guariglia S, Meng Q, Cai D. Neuropeptide exocytosis involving synaptotagmin-4 and oxytocin in hypothalamic programming of body weight and energy balance. Neuron. 2011;69:523–535. doi: 10.1016/j.neuron.2010.12.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Zhang B, Qiu L, Xiao W, Ni H, Chen L, Wang F, Mai W, Wu J, Bao A, Hu H, Gong H, Duan S, Li A, Gao Z. Reconstruction of the Hypothalamo-Neurohypophysial System and Functional Dissection of Magnocellular Oxytocin Neurons in the Brain. Neuron. 2021;109:331–346. doi: 10.1016/j.neuron.2020.10.032. [DOI] [PubMed] [Google Scholar]
  62. Zhao CM, Hayakawa Y, Kodama Y, Muthupalani S, Westphalen CB, Andersen GT, Flatberg A, Johannessen H, Friedman RA, Renz BW, Sandvik AK, Beisvag V, Tomita H, Hara A, Quante M, Li Z, Gershon MD, Kaneko K, Fox JG, Wang TC, Chen D. Denervation suppresses gastric tumorigenesis. Science Translational Medicine. 2014;6:250ra115. doi: 10.1126/scitranslmed.3009569. [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision letter

Editor: Ernie Blevins1
Reviewed by: Jeff Roizen

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This manuscript is of broad interest to gastrointestinal physiologists, cancer biologists and neuroscientists, including readers whose interests include the hypothalamic effects of anxiety as well as central effects in cancer. Pan et al., studied the consequences of manipulations of hypothalamic oxytocin (OT) neurons on pharmacologically induced colorectal cancer progression in mice and determined that celastrol, a pentacyclic triterpenoid, which should excite OT neurons, also inhibited colorectal cancer progression, an effect, which was attenuated in OT neuron-depleted mice. The authors used a series of overlapping experimental manipulations (surgical, genetic, chemogenetic, pharmacological and electrophysiological) as a way to identify the role of the sympathetic nervous system in contributing to these effects as well as to dissect circuit and to largely support the key claims of the paper.

Decision letter after peer review:

Thank you for resubmitting your work entitled "Stimulation of hypothalamic oxytocin neurons suppresses colorectal cancer progression in mice" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors and the evaluation has been overseen by Mone Zaidi as the Senior Editor. The following individual involved in the review of your submission has agreed to reveal their identity: Jeff Roizen (Reviewer #3).

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

Essential revisions:

In order to increase the impact of the paper, and to justify the conclusions drawn, especially to show causalities between OT manipulations and progression in colorectal cancer, the following points should be taken into consideration:

1) The role of the sympathetic nervous system suppression in contributing to these effects is not yet clear. The authors could first identify that sympathetic neurons were impacted using a marker specific to catecholamine neurons (tyrosine hydroxylase, for example). Secondly, the authors could assess the effects in animals that lack SNS innervation to the target tissues in question using either surgical or chemical ablation (6-OHDA) approaches as opposed to a β-2 receptor agonist. Related to this is that the sympathetic nervous system effect was only examined in the context of their novel reagent (rather than, for instance, in the DREADD-dependent model).

2) What is missing is a proposed causal mechanism of the anticancer effect of OT neuron activation. Is it the attenuation of the activity of the HPA axis, as repeatedly shown by OT, is it the reduction in chronic stress levels mediated by OT, are anti-inflammatory effects involved or other effects on the immune system? An experiment blocking OT receptors (centrally or within selected brain regions) or an experiment manipulating corticosterone levels during OT neuronal activation or depletion might be helpful.

3) It would be useful to confirm that both chronic OT neuron depletion as well as chemogenetic activation indeed affect the activity of OT neurons by assessing a functional parameter, i.e. OT staining, or plasma OT levels. Although chemogenetic activation of PVN OT neurons has been shown to elevate peripheral and central OT concentrations (Grund et al., 2019), is this still the case after repeated acute activation over 3 weeks? To which degree are OT neurons depleted by then?

4) In this context, the authors also describe that "celastrol may regulate the performance of certain ion channels, thus enhancing Oxt neuron firing in response to physiological stimuli". In the context of their study, what is the physiological stimulus? Does celastrol activate also baseline neuronal activity? Does celastrol also trigger OT secretion in vivo? Here, answers to these questions need to be given.

5) Please provide evidence that celastrol selectively affects OT neurons in the PVN and not any other neurons in the brain, as it was administered icv, and not only into the PVN

Reviewer #1:

The authors were hoping to be able to demonstrate that oxytocin neurons in the paraventricular nucleus can impact anxiety and modulate colitis-associated cancer (CAC) progression using a mouse model. They also identified a compound, celastrol, that can activate oxytocin neurons and reduce CAC tumor progression through an oxytocin-like pathway. Lastly, they were hoping to be able to show that celastrol can suppress tumor activity by inhibiting sympathetic nervous system activation.

The major strengths to this body of work lie in the novelty of the hypotheses being examined and the number of sophisticated approaches taken to test their hypotheses. The investigators were largely able to achieve their goals given that they identified previous findings showing that impairments in oxytocin signaling can increase anxiety using a number of behavioral approaches. They were able to demonstrate an increase in CAC progression using a mouse model with depleted oxytocin signaling within the PVN. They then found that stimulation of PVN oxytocin neurons can inhibit CAC progression by suppressing cell proliferation and promoting cell apoptosis. They found that treatment with celastrol could excite PVN oxytocin neurons and that brain treatment with celastrol can suppress CAC progression in mice. In addition, they found that hypothalamic oxytocin neurons are required in the anti-tumor effect of celastrol.

The main weakness I identified has to do with their last question as to whether the SNS contributes to these effects. I think that the authors could have more fully identified that the neurons in question were actually sympathetic neurons using a marker such as tyrosine hydroxylase. In addition, I think they could have lesioned SNS innervation to the colon through a surgical or chemical (6-OHDA) to more fully determine the impact of defective SNS innervation in their model.

I think that the authors achieved the majority of their aims in question with the exception of their last aim addressing the role of the sympathetic nervous system.

This body of work will be largely impactful to gastrointestinal physiologists, cancer biologists and neuroscientists as it has high clinical relevance given the therapeutic potential of oxytocin treatment. It reveals a potential novel treatment for colitis-associated cancer and the mechanisms that may contribute to these effects. The authors used a series of compelling experimental manipulations (genetic, chemogenetic, pharmacological and electrophysiological) and the utility of these approaches in this particular context could be very useful to the scientific community. They used an elegant number of approaches as a way to dissect the circuit and to largely support the key claims of the paper.

The authors have provided an impressive body of work to dissect the relevance of oxytocin neurons in modulating colitis-associated cancer and the extent to which a compound found to reduce CAC progression works through the oxytocin pathway. The authors should be commended for their thorough examination using a variety of genetic, chemogenetic, pharmacological and electrophysiological as a way to dissect the circuit and to largely support the key claims of the paper.

I think that the main concerns I have pertain to the role of the sympathetic nervous system suppression in contributing to these effects. I think the authors could first identify that sympathetic neurons were impacted using a marker specific to catecholamine neurons (tyrosine hydroxylase, for example). Secondly, I think they could assess the effects in animals that lack SNS innervation to the target tissues in question using either surgical or chemical ablation (6-OHDA) approaches as opposed to a β-2 receptor agonist.

Reviewer #2:

Pan et al., studied the consequences of manipulations of hypothalamic oxytocin (OT) neurons on pharmacologically induced colorectal cancer progression in mice. They use genetic and chemogenetic approaches to either chronically deplete all brain OT neurons or to selectively activate PVN OT neurons. Further, treatment with celastrol, a pentacyclic triterpenoid, which should excite OT neurons and which they applied into the cerebral ventricles daily over 3 weeks also inhibited colorectal cancer progression, an effect, which was attenuated in OT neuron-depleted mice.

Furthermore, brain treatment with celastrol suppresses neuronal activity in the celiac-superior mesenteric ganglion, and activation of β2 adrenergic receptor abolished the anti-tumor effect of centrally administered celastrol. In sum, the authors intend to show, by manipulation of the OT system its contribution to colorectal cancer progression. Although the experiments give some potential insights into the role of OT neurons in the immune responsiveness and the progression of colorectal cancer, novel causal relationships between the different findings are rather missing.

In order to increase the impact of the paper, especially to show causalities between OT manipulations and progression in colorectal cancer, the following points might be taken into consideration:

– What is missing is a proposed causal mechanism of the anticancer effect of OT neuron activation. Is it the attenuation of the activity of the HPA axis, as repeatedly shown by OT, is it the reduction in chronic stress levels mediated by OT, are anti-inflammatory effects involved or other effects on the immune system ? An experiment blocking OT receptors (centrally or within selected brain regions) or an experiment manipulating corticosterone levels during OT neuronal activation or depletion might be helpful.

– It would be useful to confirm that both chronic OT neuron depletion as well as chemogenetic activation indeed affect the activity of OT neurons by assessing a functional parameter, i.e. OT staining, or plasma OT levels. Although chemogenetic activation of PVN OT neurons has been shown to elevate peripheral and central OT concentrations (Grund et al., 2019), is this still the case after repeated acute activation over 3 weeks? To which degree are OT neurons depleted by then?

– In this context, the authors also describe that "celastrol may regulate the performance of certain ion channels, thus enhancing Oxt neuron firing in response to physiological stimuli". In the context of their study, what is the physiological stimulus? Does celastrol activate also baseline neuronal activity? Does celastrol also trigger OT secretion in vivo? Here, answers to these questions need to be given.

– Please provide evidence that celastrol selectively affects OT neurons in the PVN and not any other neurons in the brain, as it was administered icv, and not only into the PVN.

– What is the evidence that celastrol-induced suppresses of the activity of sympathetic neurons in the CG-SMG ganglion is mediated by OT? Inhibition or depletion of OT neurons may affect many other systems of the brain, such as the CRF system, which may result in elevated stress levels.

– In addition to negative mood, also other factors, which are significantly regulated by OT, need to be considered such as social support and chronic stress. In fact, chronic stress in mice was repeatedly described to induce colitis and to enhance colorectal cancer by the Reber group.In contrast, social support, mediated by OT, was shown to attenuate cancerogenesis and stress responses (Heinrichs et al.,).These aspects might be thoroughly considered and discussed.

– The link to negative moods, including depression and stress, repeatedly described in the introduction and discussion, remains vague, as mice were not manipulated to induce a state of increased anxiety or chronic stress.

In order to increase the impact of the paper, and to justify the conclusions drawn, especially to show causalities between OT manipulations and progression in colorectal cancer, the following points might be taken into consideration:

– What is missing is a proposed causal mechanism of the anticancer effect of OT neuron activation. Is it the attenuation of the activity of the HPA axis, as repeatedly shown by OT, is it the reduction in chronic stress levels mediated by OT, are anti-inflammatory effects involved or other effects on the immune system ? An experiment blocking OT receptors (centrally or within selected brain regions) or an experiment manipulating corticosterone levels during OT neuronal activation or depletion might be helpful.

– It would be useful to confirm that both chronic OT neuron depletion as well as chemogenetic activation indeed affect the activity of OT neurons by assessing a functional parameter, i.e. OT staining, or plasma OT levels. Although chemogenetic activation of PVN OT neurons has been shown to elevate peripheral and central OT concentrations (Grund et al., 2019), is this still the case after repeated acute activation over 3 weeks? To which degree are OT neurons depleted by then?

– In this context, the authors also describe that "celastrol may regulate the performance of certain ion channels, thus enhancing Oxt neuron firing in response to physiological stimuli". In the context of their study, what is the physiological stimulus? Does celastrol activate also baseline neuronal activity? Does celastrol also trigger OT secretion in vivo? Here, answers to these questions need to be given.

– Please provide evidence that celastrol selectively affects OT neurons in the PVN and not any other neurons in the brain, as it was administered icv, and not only into the PVN.

– What is the evidence that celastrol-induced suppresses of the activity of sympathetic neurons in the CG-SMG ganglion is mediated by OT? Inhibition or depletion of OT neurons may affect many other systems of the brain, such as the CRF system, which may result in elevated stress levels.

– In addition to negative mood, also other factors, which are significantly regulated by OT, need to be considered such as social support and chronic stress. In fact, chronic stress in mice was repeatedly described to induce colitis and to enhance colorectal cancer by the Reber group.In contrast, social support, mediated by OT, was shown to attenuate cancerogenesis and stress responses (Heinrichs et al.,).These aspects might be thoroughly considered and discussed.

Reviewer #3:

The authors note the role of anxiety in cancer risk and hypothesize that this role might be mediated to some extent via oxytocin neurons. To examine this hypothesis the authors attempted to examine the extent to which oxytocin neurons might modulate incidence and progression of colitis induced cancer. To answer this question they looked at effects of both positive and negative manipulation of oxytocin neurons. They observed that inhibition enabled cancer progression and further that activation prevented cancer progression.

Initially the authors demonstrate that genetically enabled lesioning of oxytocin neurons allows increases colitis associated cancer progression. Then they further demonstrate that chemogenetic activation of oxytocin neurons decreases colitis associated cancer progression. To validate a novel reagent, they then demonstrate that a novel herbal isolate activates oxytocin neurons and also decreases colitis associated cancer progression. They demonstrate that lesioning of oxytocin neurons abrogates this effect. Finally, they demonstrated that their novel compound inhibited SNS outflow and that bypass of this inhibition with the β-adrenergic agonist abrogated its prevention of colitis associated cancer.

Strengths of the work demonstrating include multiple manipulations of oxytocin neuron activity on colitis associated cancer. One relatively weakness of the work is that the sympathetic nervous system effect was only examined in the context of their novel reagent.

This work provides a basis for how anxiety might alter cancer risk.

Overall this is a strong manuscript. As noted above, one weakness is the demonstration of oxytocin neuron downstream effects on the SNS and bypass by the β-adrenergic agonist only using the novel herbal reagent (rather than, for instance, in the DREADD-dependent model).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Stimulation of hypothalamic oxytocin neurons suppresses colorectal cancer progression in mice" for further consideration by eLife. Your revised article has been evaluated by Mone Zaidi (Senior Editor) and a Reviewing Editor.

The manuscript has addressed the responses of the previous reviewers and has been substantially improved. However, there is one remaining issue that needs to be addressed, as outlined below:

The authors should remove Figure 3S1-A-B as it does not provide helpful information to the paper. I have a concern over the use of TH intensity as a way to measure SNS activity in IBAT. As indicated in the paper by Vaughan and Bartness (Methods Enzymol, 537: 199-235, 2014): "NETO is used as a direct neurochemical measure of sympathetic drive; as noted above, there is no surrogate for this method of assessment except for direct measures of sympathetic nerve activity electrophysiologically". The authors should remove mention of SNS activity within IBAT unless they can provide this assessment via NETO or electrophysiology. As mentioned earlier, the information provided is not the currently accepted approach to assess SNS in animals with IBAT denervation.

eLife. 2021 Sep 16;10:e67535. doi: 10.7554/eLife.67535.sa2

Author response


Essential revisions:

In order to increase the impact of the paper, and to justify the conclusions drawn, especially to show causalilties between OT manipulations and progression in colorectal cancer, the following points should be taken into consideration:

1) The role of the sympathetic nervous system suppression in contributing to these effects is not yet clear. The authors could first identify that sympathetic neurons were impacted using a marker specific to catecholamine neurons (tyrosine hydroxylase, for example). Secondly, the authors could assess the effects in animals that lack SNS innervation to the target tissues in question using either surgical or chemical ablation (6-OHDA) approaches as opposed to a β-2 receptor agonist. Related to this is that the sympathetic nervous system effect was only examined in the context of their novel reagent (rather than, for instance, in the DREADD-dependent model).

We gratefully thank the reviewers for these comments.

1) During this revision, we have assessed the effect of excitation of OxtPVN neurons, or transection of preganglionic fiber of CG-SMG on the activities of tyrosine hydroxylase (TH)-positive neurons in the CG-SMG. The data indicate that excitation of OxtPVN neurons rapidly suppressed the activities of TH-positive neurons in the CG-SMG (Figure 3A,B). In a separate experiment, we showed that i.c.v. administration of celastrol readily suppressed the activities of TH-positive neurons in the CG-SMG, and transection of the preganglionic fiber could significantly attenuate this effect (Figure 7—figure supplement 1A-C).

2) To address the reviewer’s second question, we elected to surgically remove CG-SMG in the OxtCre and OxtCre;DTA mice (Figure 3C-K). In agreement with our early observation, depletion of Oxt neurons promoted colitis-associated cancer (CAC) development in mice (Figure 3E-K). After the resection of CG-SMG, this effect was markedly abrogated (Figure 3E-K).

3) Following the reviewer’s suggestion, we examined the relationship between OxtPVN neurons and β2-adrenergic receptor (β2AR) in the progression of CAC. The data indicate that the DREADD-mediated activation of OxtPVN neurons (Figure 6A,B) suppressed CAC progression in mice (Figure 6C-J). Notably, i.p. administration of isoprenaline, a β2AR agonist, could significantly attenuate this effect (Figure 6C-J). These data suggest that suppression of β2AR activity is crucial for OxtPVN neuron activation to restrict CAC progression.

2) What is missing is a proposed causal mechanism of the anticancer effect of OT neuron activation. Is it the attenuation of the activity of the HPA axis, as repeatedly shown by OT, is it the reduction in chronic stress levels mediated by OT, are anti-inflammatory effects involved or other effects on the immune system ? An experiment blocking OT receptors (centrally or within selected brain regions) or an experiment manipulating corticosterone levels during OT neuronal activation or depletion might be helpful.

We gratefully thank reviewer #2 for these very helpful comments.

1) In this revision, we have assessed the activity of the HPA axis. Our data indicate that depletion of Oxt neurons resulted in the elevation of circulating ACTH and corticosterone levels in mice (Figure 1—figure supplement 1L,M). Conversely, chemogenetic approach-mediated excitation of OxtPVN neurons could lead to a significant decrease of ACTH and corticosterone levels in systemic circulation (Figure 1—figure supplement 2K,L). These data suggest that the HPA axis may play a role in the modulation of tumor progression by OxtPVN neurons.

2) In agreement with the changes in the HPA axis, our assessments show that mice deficient for Oxt neurons exhibited an elevated anxiety level (Figure 1—figure supplement 1A-C), while excitation of OxtPVN neurons in OxtCre mice had an anxiolytic effect (Figure 1—figure supplement 2A-C).

3) Also, the OxtCre mice were injected with control or hM3Dq AAV into the PVN, and then were i.p. administered with CNO every other day for 3 consecutive weeks (Figure 1—figure supplement 2D). The tumor tissues were then harvested and immune cells were assessed. The data show that the number of CD8+ T cells was remarkably increased in the tumor tissue of the mice with OxtPVN neuron activation, whereas other types of immune cell were not significantly impacted (Figure 1—figure supplement 3). These data suggest that excitation of OxtPVN neurons in the brain may bestow its beneficial effect by promoting the anti-tumor immunity.

4) Following the reviewer’s suggestion, we carried out an experiment in which L-368,899, a selective Oxt receptor (OTR) antagonist, was used to block OTR in the mouse brain (Figure 2 and Figure 2—figure supplement 1). The data show that CAC progression was inhibited in the OxtCre mice in which OxtPVN neurons had been stimulated (Figure 2C-J). Notably, blockade of OTR in the brain, which was achieved by injecting L-368,899 into the third ventricle, could markedly abolish the tumor suppression effect of OxtPVN neuron activation (Figure 2C-J). These data indicate that brain OTR is crucial for activation of OxtPVN neurons to suppress CAC progression in mice.

3) It would be useful to confirm that both chronic OT neuron depletion as well as chemogenetic activation indeed affect the activity of OT neurons by assessing a functional parameter, i.e. OT staining, or plasma OT levels. Although chemogenetic activation of PVN OT neurons has been shown to elevate peripheral and central OT concentrations (Grund et al., 2019), is this still the case after repeated acute activation over 3 weeks? To which degree are OT neurons depleted by then?

We thank reviewer #2 for these valid points. To address them, we have carried out both immunofluorescent staining and Oxt EIA assays. (1) Regarding Oxt neuron depletion, the immunofluorescent staining data demonstrate that, at the end of the experiment, ⁓94% of the Oxt neurons had been lesioned in the PVN of the OxtCre;DTA mice (Figure 1B,C), in which plasma Oxt was barely detectable (Figure 1—figure supplement 1G). (2) With regard to the chemogenetic activation of OxtPVN neurons, after a 3-week treatment of CNO, the majority of OxtPVN neurons were excited (Figure 1I,J), and plasma Oxt level was elevated in the hM3Dq AAV-injected mice (Figure 1—figure supplement 2E). Together, these data indicate that the employed experimental models could work as expected. We also cited the study by Grund and colleagues in the revised manuscript.

4) In this context, the authors also describe that "celastrol may regulate the performance of certain ion channels, thus enhancing Oxt neuron firing in response to physiological stimuli". In the context of their study, what is the physiological stimulus? Does celastrol activate also baseline neuronal activity? Does celastrol also trigger OT secretion in vivo? Here, answers to these questions need to be given.

In the slice electrophysiology experiments, current injection ranging from 20 to 200 pA was used to test the excitability of OxtPVN neurons. Previous work indicated that physiological stimuli, such as social touch1, tactile stimuli2, feeding3 and leptin4 could lead to the excitation of Oxt neurons. Here, the electrical stimuli were utilized to mimic the excitatory inputs in response to natural stimuli mentioned above. Our results suggest that celastrol could elevate the responsiveness to the same stimuli. We apologize for not having described this clearly. Our data indicate that i.c.v. administration of celastrol could excite OxtPVN neurons (percentage of c-Fos-positive OxtPVN neurons of total OxtPVN neurons: vehicle, 12.3±2.0%; celastrol, 34.7±6.6%. P=0.01, n=5 mice per group), suggesting that it can activate these neurons.

To assess the effect of celastrol on Oxt secretion, we chose to use an ex vivo Oxt release assay, since this method has been established in our laboratory5. The PVN slices were dissected from adult male C57 BL/6 mice, and then were balanced in normal Locke’s solution. Thereafter, the tissues were incubated in the same solution supplemented with celastrol. The contents of Oxt in these solutions were then determined using an Oxt EIA kit. Indeed, treatment with celastrol could enhance Oxt secretion from the PVN slices (Figure 4—figure supplement 1F).

5) Please provide evidence that celastrol selectively affects OT neurons in the PVN and not any other neurons in the brain, as it was administered icv, and not only into the PVN

We thank reviewer #2 for this valid suggestion. To address this question, adult male C57 BL/6 mice were i.c.v. administered with celastrol versus vehicle control. Two hours later, mice were perfused with 4% paraformaldehyde, and then brain tissues were sectioned. Immunofluorescent staining for c-Fos demonstrates that treatment with celastrol triggered excitation of neurons in the PVN, but not other hypothalamic nuclei (Figure 4—figure supplement 1A,B). In combination with our electrophysiological data (Figure 4A-G), these results suggest that celastrol could selectively regulate the activities of Oxt neurons in the PVN.

Reviewer #1:

This body of work will be largely impactful to gastrointestinal physiologists, cancer biologists and neuroscientists as it has high clinical relevance given the therapeutic potential of oxytocin treatment. It reveals a potential novel treatment for colitis-associated cancer and the mechanisms that may contribute to these effects. The authors used a series of compelling experimental manipulations (genetic, chemogenetic, pharmacological and electrophysiological) and the utility of these approaches in this particular context could be very useful to the scientific community. They used an elegant number of approaches as a way to dissect the circuit and to largely support the key claims of the paper.

We gratefully thank reviewer #1 for these encouraging comments.

The authors have provided an impressive body of work to dissect the relevance of oxytocin neurons in modulating colitis-associated cancer and the extent to which a compound found to reduce CAC progression works through the oxytocin pathway. The authors should be commended for their thorough examination using a variety of genetic, chemogenetic, pharmacological and electrophysiological as a way to dissect the circuit and to largely support the key claims of the paper.

We are very grateful to reviewer #1 for these positive comments.

I think that the main concerns I have pertain to the role of the sympathetic nervous system suppression in contributing to these effects. I think the authors could first identify that sympathetic neurons were impacted using a marker specific to catecholamine neurons (tyrosine hydroxylase, for example). Secondly, I think they could assess the effects in animals that lack SNS innervation to the target tissues in question using either surgical or chemical ablation (6-OHDA) approaches as opposed to a β-2 receptor agonist.

We appreciate reviewer #1 for these valid suggestions.

1) During this revision, we have assessed the effect of the excitation of OxtPVN neurons, or the transection of preganglionic fiber of CG-SMG on the activities of tyrosine hydroxylase (TH)-positive neurons in the CG-SMG. The data indicate that excitation of OxtPVN neurons rapidly suppressed the activities of TH-positive neurons in the CG-SMG (Figure 3A,B). In a separate experiment, we showed that i.c.v. administration of celastrol readily suppressed the activities of TH-positive neurons in the CG-SMG, and the transection of the preganglionic fiber could significantly attenuate this effect (Figure 7—figure supplement 1A-C).

2) To address the reviewer’s second question, we elected to surgically remove CG-SMG in the OxtCre and OxtCre;DTA mice (Figure 3C-K). In agreement with our early observation, depletion of Oxt neurons promoted colitis-associated cancer (CAC) development in mice (Figure 3E-K). After the resection of CG-SMG, this effect was significantly abrogated (Figure 3E-K).

Reviewer #2:

In order to increase the impact of the paper, and to justify the conclusions drawn, especially to show causalilties between OT manipulations and progression in colorectal cancer, the following points might be taken into consideration:

– What is missing is a proposed causal mechanism of the anticancer effect of OT neuron activation. Is it the attenuation of the activity of the HPA axis, as repeatedly shown by OT, is it the reduction in chronic stress levels mediated by OT, are anti-inflammatory effects involved or other effects on the immune system? An experiment blocking OT receptors (centrally or within selected brain regions) or an experiment manipulating corticosterone levels during OT neuronal activation or depletion might be helpful.

We gratefully thank reviewer #2 for these very helpful comments.

1) In this revision, we have assessed the activity of the HPA axis. Our data indicate that depletion of Oxt neurons resulted in the elevation of the circulating ACTH and corticosterone levels in mice (Figure 1—figure supplement 1L,M). Conversely, chemogenetic approach-mediated excitation of OxtPVN neurons could significantly decrease ACTH and corticosterone levels in systemic circulation (Figure 1—figure supplement 2K,L). These data suggest that the HPA axis may play a role in the modulation of tumor progression by OxtPVN neurons.

2) In agreement with the changes in the HPA axis, our assessments show that mice deficient for Oxt neuron exhibited an elevated anxiety level (Figure 1—figure supplement 1A-C), while excitation of OxtPVN neurons in OxtCre mice had an anxiolytic effect (Figure 1—figure supplement 2A-C).

3) Also, the OxtCre mice were injected with control or hM3Dq AAV into the PVN, and then were i.p. administered with CNO every other day for 3 consecutive weeks (Figure 1—figure supplement 2D). The tumor tissues were then harvested and immune cells were assessed. The data show that the number of CD8+ T cells was remarkably increased in the tumor tissue of the mice with OxtPVN neuron activation, whereas other types of immune cell were not significantly impacted (Figure 1—figure supplement 3). These data suggest that excitation of OxtPVN neurons in the brain may bestow its beneficial effect by promoting anti-tumor immunity.

4) Following the reviewer’s suggestion, we carried out an experiment in which L-368,899, a selective Oxt receptor (OTR) antagonist, was used to block OTR in the mouse brain (Figure 2 and Figure 2—figure supplement 1). The data show that CAC progression was inhibited in the OxtCre mice in which OxtPVN neurons had been stimulated (Figure 2C-J). Notably, blockade of OTR in the brain, which was achieved by injecting L-368,899 into the third ventricle, could markedly abolish the tumor suppression effect of OxtPVN neuron activation (Figure 2C-J). These data indicate that brain OTR is crucial for activation of OxtPVN neurons to suppress CAC progression in mice.

– It would be useful to confirm that both chronic OT neuron depletion as well as chemogenetic activation indeed affect the activity of OT neurons by assessing a functional parameter, i.e. OT staining, or plasma OT levels. Although chemogenetic activation of PVN OT neurons has been shown to elevate peripheral and central OT concentrations (Grund et al., 2019), is this still the case after repeated acute activation over 3 weeks? To which degree are OT neurons depleted by then?

We thank reviewer #2 for these valid points. To address them, we have carried out both immunofluorescent staining and Oxt EIA assays. (1) Regarding Oxt neuron depletion, the immunofluorescent staining data demonstrate that, at the end of the experiment, ⁓94% of the Oxt neurons had been lesioned in the PVN of the OxtCre;DTA mice (Figure 1B,C), in which plasma Oxt was barely detectable (Figure 1—figure supplement 1G). (2) With regard to the chemogenetic activation of OxtPVN neurons, after a 3-week treatment of CNO, the majority of OxtPVN neurons were excited (Figure 1I,J), and plasma Oxt level was elevated in the hM3Dq AAV-injected mice (Figure 1—figure supplement 2E). Together, these data indicate that the employed experimental models could work as expected. We also cited the study by Grund and colleagues in the revised manuscript.

– In this context, the authors also describe that "celastrol may regulate the performance of certain ion channels, thus enhancing Oxt neuron firing in response to physiological stimuli". In the context of their study, what is the physiological stimulus? Does celastrol activate also baseline neuronal activity? Does celastrol also trigger OT secretion in vivo? Here, answers to these questions need to be given.

In the slice electrophysiology experiments, current injection ranging from 20 to 200 pA was used to test the excitability of OxtPVN neurons. Previous work indicated that physiological stimuli, such as social touch1, tactile stimuli2, feeding3 and leptin4 could lead to the excitation of Oxt neurons. Here, the electrical stimuli were utilized to mimic the excitatory inputs in response to natural stimuli mentioned above. Our results suggest that celastrol could elevate the responsiveness to the same stimuli. We apologize for not having described this clearly. Our data indicate that i.c.v. administration of celastrol could excite OxtPVN neurons (percentage of c-Fos-positive OxtPVN neurons of total OxtPVN neurons: vehicle, 12.3±2.0%; celastrol, 34.7±6.6%. P=0.01, n=5 mice per group).

To assess the effect of celastrol on Oxt secretion, we chose to use an ex vivo Oxt release assay, since this method has been established in our laboratory5. The PVN slices were dissected from adult male C57 BL/6 mice, and then were balanced in normal Locke’s solution. Thereafter, the tissues were incubated in the same solution supplemented with celastrol. The contents of Oxt in these solutions were then determined using an Oxt EIA kit. Indeed, treatment with celastrol could enhance Oxt secretion from the PVN slices (Figure 4—figure supplement 1F).

– Please provide evidence that celastrol selectively affects OT neurons in the PVN and not any other neurons in the brain, as it was administered icv, and not only into the PVN.

We thank reviewer #2 for this valid suggestion. To address this question, adult male C57 BL/6 mice were i.c.v. administered with celastrol versus vehicle control. Two hours later, mice were perfused with 4% paraformaldehyde, and then brain tissues were sectioned. Immunofluorescent staining for c-Fos demonstrates that treatment with celastrol triggered excitation of neurons in the PVN, but not other hypothalamic nuclei (Figure 4—figure supplement 1A,B). In combination with our electrophysiological data (Figure 4A-G), these results suggest that celastrol could selectively regulate the activities of Oxt neurons in the PVN.

– What is the evidence that celastrol-induced suppression of the activity of sympathetic neurons in the CG-SMG ganglion is mediated by OT? Inhibition or depletion of OT neurons may affect many other systems of the brain, such as the CRF system, which may result in elevated stress levels.

1) To address the reviewer’s point, male adult C57 BL/6 mice were implanted with a guide cannula directed to third ventricle. After recovery, these mice were i.c.v. injected with aCSF or L-368,899, the OTR antagonist. An hour later, the 6-min control spiking activities were acquired from neurons of the CG-SMG, before celastrol application through the pre-implanted guide cannula. This in vivo single-unit recordings data demonstrates that i.c.v. administration of celastrol decreased the firing frequency of neurons in the CG-SMG, and that pre-treatment with OTR antagonist could significantly attenuate this effect (Figure 7D-G). These results indicate that brain Oxt is important for celastrol to regulate the neuronal activity in the CG-SMG.

2) During this revision, we measured the plasma ACTH and corticosterone levels in the OxtCre and OxtCre;DTA mice. The data display that both ACTH and corticosterone levels were elevated in the OxtCre;DTA mice compared to their levels in the controls (Figure 1—figure supplement 1L,M). This increased activity of the HPA axis may lead to elevated stress level, and then contribute to the development of CAC. However, this needs further investigations.

– In addition to negative mood, also other factors, which are significantly regulated by OT, need to be considered such as social support and chronic stress. In fact, chronic stress in mice was repeatedly described to induce colitis and to enhance colorectal cancer by the Reber group.In contrast, social support, mediated by OT, was shown to attenuate cancerogenesis and stress responses (Heinrichs et al.,).These aspects might be thoroughly considered and discussed.

Following the reviewer’s suggestion, we have included the discussions of the effects of chronic stress, especially those studies by the Reber group, and social support (by Heinrichs et al.,) on colitis and colorectal cancer in the revised manuscript.

Reviewer #3:

The authors note the role of anxiety in cancer risk and hypothesize that this role might be mediated to some extent via oxytocin neurons. To examine this hypothesis the authors attempted to examine the extent to which oxytocin neurons might modulate incidence and progression of colitis induced cancer. To answer this question they looked at effects of both positive and negative manipulation of oxytocin neurons. They observed that inhibition enabled cancer progression and further that activation prevented cancer progression.

Initially the authors demonstrate that geneticly enabled lesioning of oxytocin neurons allows increases colitis associated cancer progression. Then they further demonstrate that chemogenetic activation of oxytocin neurons decreases colitis associated cancer progression. To validate a novel reagent, they then demonstrate that a novel herbal isolate activates oxytocin neurons and also decreases colitis associated cancer progression. They demonstrate that lesioning of oxytocin neurons abrogates this effect. Finally, they demonstrated that their novel compound inhibited SNS outflow and that bypass of this inhibition with the β-adrenergic agonist abrogated its prevention of colitis associated cancer.

Strengths of the work demonstrating include multiple manipulations of oxytocin neuron activity on colicitis associated cancer. One relatively weakness of the work is that the sympathetic nervous system effect was only examined in the context of their novel reagent.

This work provides a basis for how anxiety might alter cancer risk.

Overall this is a strong manuscript. As noted above, one weakness is the demonstration of oxytocin neuron downstream effects on the SNS and bypass by the β-adrenergic agonist only using the novel herbal reagent (rather than, for instance, in the DREADD-dependent model).

We thank reviewer #3 for this comment. Following the reviewer’s suggestion, we examined the relationship between Oxt neurons in the PVN and β2-adrenergic receptor (β2AR) in the progression of colitis-associated cancer (CAC). The data indicate that the DREADD-mediated activation of Oxt neurons in the PVN (Figure 6A,B) suppressed CAC progression in mice (Figure 6C-J). Notably, i.p. administration of isoprenaline, a β2AR agonist, could significantly attenuate this effect (Figure 6C-J). These data suggest that suppression of β2AR activity is crucial for activation of Oxt neurons in the PVN to restrict CAC progression in mice.

References

1. Tang, Y., et al., Social touch promotes interfemale communication via activation of parvocellular oxytocin neurons. Nat Neurosci 23, 1125-1137 (2020).

2. Okabe, S., Yoshida, M., Takayanagi, Y. and Onaka, T. Activation of hypothalamic oxytocin neurons following tactile stimuli in rats. Neurosci Lett 600, 22-27 (2015).

3. Johnstone, L.E., Fong, T.M. and Leng, G. Neuronal activation in the hypothalamus and brainstem during feeding in rats. Cell Metab 4, 313-321 (2006).

4. Blevins, J.E., Schwartz, M.W. and Baskin, D.G. Evidence that paraventricular nucleus oxytocin neurons link hypothalamic leptin action to caudal brain stem nuclei controlling meal size. Am J Physiol Regul Integr Comp Physiol 287, R87-96 (2004).

5. Wu, L., et al., Caffeine inhibits hypothalamic A1R to excite oxytocin neuron and ameliorate dietary obesity in mice. Nat Commun 8, 15904 (2017).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The authors should remove Figure 3S1-A-B as it does not provide helpful information to the paper. I have a concern over the use of TH intensity as a way to measure SNS activity in IBAT. As indicated in the paper by Vaughan and Bartness (Methods Enzymol, 537: 199-235, 2014): "NETO is used as a direct neurochemical measure of sympathetic drive; as noted above, there is no surrogate for this method of assessment except for direct measures of sympathetic nerve activity electrophysiologically". The authors should remove mention of SNS activity within IBAT unless they can provide this assessment via NETO or electrophysiology. As mentioned earlier, the information provided is not the currently accepted approach to assess SNS in animals with IBAT denervation.

We gratefully thank the editors for these valid comments. Following the suggestion, we have deleted Figure 3-figure supplement 1A,B in the previous manuscript. We have also edited the main text and other contents (eg., the Methods section and the figure legend for Figure 3-figure supplement 1) to reflect this change.

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    Figure 1—source data 1. Source data for Figure 1, panels C-G and J-N.
    Figure 2—source data 1. Source data for Figure 2, panels B, D-F, H and J.
    Figure 3—source data 1. Source data for Figure 3, panels B, F, G, I and K.
    Figure 4—source data 1. Source data for Figure 4, panels C and E-L.
    Figure 5—source data 1. Source data for Figure 5E–G,I,K.
    Figure 6—source data 1. Source data for Figure 6, panels B, D-F, H and J.
    Figure 7—source data 1. Source data for Figure 7, panels E-K.
    Transparent reporting form

    Data Availability Statement

    All data that support the findings of this study are included in this published article and its supplementary files. Source data files have been provided for Figures 1-7.


    Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

    RESOURCES