Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jul;20(7):934-942.
doi: 10.1038/nn.4574. Epub 2017 Jun 5.

Cancer-induced Anorexia and Malaise Are Mediated by CGRP Neurons in the Parabrachial Nucleus

Affiliations
Free PMC article

Cancer-induced Anorexia and Malaise Are Mediated by CGRP Neurons in the Parabrachial Nucleus

Carlos A Campos et al. Nat Neurosci. .
Free PMC article

Abstract

Anorexia is a common manifestation of chronic diseases, including cancer. Here we investigate the contribution to cancer anorexia made by calcitonin gene-related peptide (CGRP) neurons in the parabrachial nucleus (PBN) that transmit anorexic signals. We show that CGRPPBN neurons are activated in mice implanted with Lewis lung carcinoma cells. Inactivation of CGRPPBN neurons before tumor implantation prevents anorexia and loss of lean mass, and their inhibition after symptom onset reverses anorexia. CGRPPBN neurons are also activated in Apcmin/+ mice, which develop intestinal cancer and lose weight despite the absence of reduced food intake. Inactivation of CGRPPBN neurons in Apcmin/+ mice permits hyperphagia that counteracts weight loss, revealing a role for these neurons in a 'nonanorexic' cancer model. We also demonstrate that inactivation of CGRPPBN neurons prevents lethargy, anxiety and malaise associated with cancer. These findings establish CGRPPBN neurons as key mediators of cancer-induced appetite suppression and associated behavioral changes.

Figures

Figure 1
Figure 1. CGRPPBN neurons are pathologically active in tumor-bearing mice
a, Illustration of PBN anatomical location in the pons. b, Quantification of Fos expression in the external lateral PBN in sham-treated or tumor-bearing mice (n = 7, Sham; n = 8, Sham PF; n = 9, LLC; one-way ANOVA: F(2,21) = 138.60, p < 0.0001). c, Coincidence of Cre:GFP-labeled CGRP neurons and Fos expression in tumor-bearing mice (n = 9, LLC). d, Representative images demonstrating Cre:GFP and Fos overlap. Data points in graphs represent each individual animal. *** P ˂ 0.001. PF, pair-fed; LLC, Lewis lung carcinoma cell implant; scp, superior cerebellar peduncle. Scale bar is 50 µm. See also Supplementary Fig. 1.
Figure 2
Figure 2. Inactivation of CGRPPBN neurons prevents cancer anorexia
a, Stereotaxic injections of AAV1-DIO-GFP:TetTox (or AAV1-DIO-GFP control virus) into the PBN of CalcaCre:GFP/+ mice prior to LLC tumor implantation. b, Cumulative food intake following tumor implantation or sham treatment in mice with intact (GFP, n = 9), bilateral inactivation (TetTox, n = 8), or unilateral inactivation (TetTox Unilat, n = 4) of CGRPPBN neurons (two-way repeated measures ANOVA: Interaction, F(42, 350) = 19.99, p < 0.0001; Time, F(14, 350) = 4188, p < 0.0001; Treatment, F(3, 350) = 10.61, p = 0.0001). c, Average daily food intake at various time epochs (two-way repeated measures ANOVA: Interaction, F(9, 72) = 7.04, p < 0.0001; Time Epoch, F(3, 72) = 24.76, p < 0.0001; Treatment, F(3, 72) = 8.65, p = 0.0005). d-e, GFP-labeled axonal processes of CGRPPBN neurons in the CeA and ovBNST. Yellow, dashed outlines represent brain areas that were quantified for Fos expression. f-h, Quantification and representative images of Fos expression in the CeA (one-way ANOVA: F(4,38) = 12.79, p < 0.0001) and ovBNST (one-way ANOVA: F(4,38) = 25.00, p < 0.0001) of sham or tumor-bearing mice with intact (GFP) or inactivated (TetTox) CGRPPBN neurons. Line graphs show mean ± SEM. Box plots show mean (+), median, quartiles (boxes) and range (whiskers) for each treatment. * P ˂ 0.05, ** P ˂ 0.01, *** P ˂ 0.001. Scale bars are 100 µm. See also Supplementary Fig. 2.
Figure 3
Figure 3. Established cancer anorexia is reversed with chemogentic inhibition of CGRPPBN neurons
a, Stereotaxic injections of AAV1-DIO-hM4Di:mCherry or AAV1-DIO-mCherry control virus into the PBN of CalcaCre:GFP/+ mice prior to LLC tumor implantation. b, Daily food intake of sham-treated or tumor-bearing mice relative to the start of CNO treatment administered ip twice-daily at 2 mg/kg and in the drinking water (mCherry LLC, n = 8; mCherry Sham, n = 8; hM4Di LLC, n = 7; hM4Di Sham, n = 8; two-way ANOVA: Interaction, F(27, 237) = 3.34, p < 0.0001; Time, F(9, 237) = 2.72, p = 0.0049; Treatment, F(3, 237) = 44.53, p < 0.0001). c, Average daily food intake during the 4 d period prior to CNO treatment (Pre CNO) and all of the days after commencement of CNO treatment (Post CNO)(two-way ANOVA: Interaction, F(3, 54) = 17.55, p < 0.0001; Time, F(1, 54) = 0.37, p = 0.55; Treatment, F(3, 54) = 36.43, p < 0.0001). d, Representative images of Fos expression in the PBN of tumor-bearing mice with intact (mCherry) or inhibited (hM4Di) CGRPPBN neurons. e-g, Quantification and representative images of Fos expression in the CeA (one-way ANOVA: F(3, 30) = 74.56, p < 0.0001) and ovBNST (one-way ANOVA: F(3, 30) = 23.76, p < 0.0001) of sham or tumor-bearing mice with intact (mCherry) or inhibited (hM4Di) CGRPPBN neurons. Line graphs show mean ± SEM. Box plots show mean (+), median, quartiles (boxes) and range (whiskers) for each treatment. ** P ˂ 0.01, *** P ˂ 0.001. Scale bars are 50 µm. See also Supplementary Fig. 3.
Figure 4
Figure 4. Role of CGRPPBN neurons in cancer-induced cachexia and increased energy expenditure
a, Time course of experiments with CalcaCre:GFP/+ mice that previously underwent stereotaxic injections of AAV1-DIO-GFP:TetTox (or AAV1-DIO-GFP control virus) into the PBN. b-d, Calorimetry measurements from GFP control mice (n = 9) before (Base) and after tumor implantation (VCO2, one-way repeated-measures ANOVA: F(3, 35) = 15.62, p < 0.0001; VO2, one-way repeated-measures ANOVA: F(3, 35) = 15.99, p < 0.0001; kcal/h, one-way repeated-measures ANOVA: F(3, 35) = 19.27, p < 0.0001). e-g, Calorimetry measurements taken from TetTox mice (n = 9) before (Base) and after tumor implantation (VCO2, one-way repeated-measures ANOVA: F(3, 35) = 15.29, p < 0.0001; VO2, one-way repeated-measures ANOVA: F(3, 35) = 18.00, p < 0.0001; kcal/h, one-way repeated-measures ANOVA: F(3, 35) = 15.16, p < 0.0001). h-i, Fat and lean body mass measurements of GFP control mice before, 17 d after tumor implantation (LLC1), and after developing anorexia (LLC2); hashtag indicates significant difference (P ˂ 0.05) between LLC1 and LLC2 (Fat Mass, one-way repeated-measures ANOVA: F(2, 26) = 57.13, p < 0.0001; Lean Mass, one-way repeated-measures ANOVA: F(3, 35) = 5.736, p < 0.01). j, Average daily food intake of GFP mice before and after tumor implant (one-way repeated-measures ANOVA: F(3, 35) = 10.94, p < 0.0001). k-l, Fat and lean body mass measurements from TetTox mice before (Base), 17 d after tumor implantation (LLC1), and after developing anorexia (LLC2)(Fat Mass, one-way repeated-measures ANOVA: F(2, 26) = 69.74, p < 0.0001; Lean Mass, one-way repeated-measures ANOVA: F(3, 35) = 10.90, p < 0.001). m, Average daily food intake of TetTox mice before and after tumor implant (one-way repeated-measures ANOVA: F(3, 35) = 3.718, p < 0.05). Box plots show mean (+), median, quartiles (boxes) and range (whiskers) for each treatment. * P ˂ 0.05, ** P ˂ 0.01, *** P ˂ 0.001; asterisks indicate significant differences compared to baseline. See also Supplementary Fig. 4 and Supplementary Fig. 5.
Figure 5
Figure 5. Inactivation of CGRPPBN neurons permits hyperphagia in “non-anorexic” Apcmin/+ genetic cancer model
a-c, Quantification and representative images of Fos expression in Cre:GFP-labeled CGRPPBN neurons of CalcaCre:GFP/+∷Apcmin/+ mice (n = 9) or wild-type CalcaCre:GFP/+ littermates (n = 7). Data points in graphs represent each individual animal (student's t-test (two-tailed): t(14) = 7.95, p < 0.0001). d-e, Cumulative and average daily food intake of male CalcaCre:GFP/+∷Apcmin/+ and wild-type CalcaCre:GFP/+ littermates that received bilateral PBN injections of AAV1-DIO-GFP:TetTox to inactivate CGRP neurons or AAV1-DIO-GFP control virus (GFP Apcmin/+, n = 7; TetTox Apcmin/+, n = 7; GFP WT, n = 5; Cumulative Food Intake, two-way ANOVA: Interaction, F(10, 92) = 6.078, p < 0.0001; Time, F(5, 92) = 662.7, p < 0.0001; Treatment, F(2, 54) = 86.95, p < 0.0001; Daily Food Intake, two-way ANOVA: Interaction, F(10, 92) = 1.059, p = 0.4019; Time, F(5, 92) = 0.68, p = 0.64; Treatment, F(2, 92) = 69.58, p < 0.0001). f-g, Cumulative and average daily food intake of female cohort (GFP Apcmin/+, n = 12; TetTox Apcmin/+, n = 7; GFP WT, n = 4; Cumulative Food Intake, two-way ANOVA: Interaction, F(10, 118) = 2.97, p = 0.0023; Time, F(5, 118) = 451.1, p < 0.0001; Treatment, F(2, 118) = 21.35, p < 0.0001; Daily Food Intake, two-way ANOVA: Interaction, F(10, 119) = 2.47, p = 0.0101; Time, F(5, 119) = 1.42, p = 0.2233; Treatment, F(2, 119) = 21.92, p < 0.0001). Line graphs show mean ± SEM. Box plots show mean (+), median, quartiles (boxes) and range (whiskers) for each treatment. * P ˂ 0.05, ** P ˂ 0.01, *** P ˂ 0.001. † symbols represent euthanasia of a GFP Apcmin/+ mouse. Scale bar is 50 µm. See also Supplementary Fig. 5 and Supplementary Fig. 7.
Figure 6
Figure 6. Inactivation of CGRPPBN neurons attenuates sickness behaviors in LLC cancer model
a-b, Hourly locomotor activity of mice with intact (GFP, n = 8) or inactivated (TetTox, n = 7) CGRPPBN neurons, prior to tumor implantation and 13 d after implantation (Pre-LLC, two-way repeated measures ANOVA: Interaction, F(22, 286) = 1.12, p = 0.3222; Time, F(22, 286) = 39.31, p < 0.0001; LLC, two-way repeated measures ANOVA: Interaction, F(1, 286) = 4.48, p < 0.0001; Time, F(22, 286) = 15.91, p < 0.0001; Treatment, F(1, 286) = 14.73, p = 0.0021). Grey-shaded region on × axis represents dark-cycle. c-d, Total distance traveled and average velocity of locomotor activity during 23-h recording period of GFP and TetTox mice prior to and after tumor implantation (Total Distance, two-way repeated measures ANOVA: Interaction, F(1, 13) = 6.21, p = 0.0270; Time, F(1, 13) = 15.91, p = 0.0096; Treatment, F(1, 13) = 4.76, p = 0.0480; Avg Velocity, two-way repeated measures ANOVA: Interaction, F(1, 13) = 2.95, p = 0.1099; Time, F(1, 13) = 4.69, p = 0.0495; Treatment, F(1, 13) = 8.05, p = 0.0140). e-g, Measurements from open-field anxiety test of GFP and TetTox mice prior to and 15 d after tumor implantation (Time in Center, two-way repeated measures ANOVA: Interaction, F(1, 13) = 7.19, p = 0.0189; Time, F(1, 13) = 2.40, p = 0.1453; Treatment, F(1, 13) = 5.64, p = 0.0337; Distance Moved, two-way repeated measures ANOVA: Interaction, F(1, 13) = 1.99, p = 0.1814; Time, F(1, 13) = 26.04, p = 0.0002; Treatment, F(1, 13) = 0.92, p = 0.3547; Time in Hunched Posture, two-way repeated measures ANOVA: Interaction, F(1, 13) = 20.23, p = 0.0006; Time, F(1, 13) = 16.40, p = 0.0014; Treatment, F(1, 13) = 15.35, p = 0.0018). h-j, Measurements from elevated-plus maze anxiety test of GFP and TetTox mice prior to and 15 d after tumor implantation (Time in Open Arm, two-way repeated measures ANOVA: Interaction, F(1, 13) = 14.60, p = 0.0021; Time, F(1, 13) = 16.74, p = 0.0013; Treatment, F(1, 13) = 1.032, p = 0.3283; Distance Moved, two-way repeated measures ANOVA: Interaction, F(1, 13) = 1.59, p = 0.2299; Time, F(1, 13) = 1.02, p = 0.3310; Treatment, F(1, 13) = 0.28, p = 0.6045; Time in Hunched Posture, two-way repeated measures ANOVA: Interaction, F(1, 13) = 9.82, p = 0.0079; Time, F(1, 13) = 5.92, p = 0.0302; Treatment, F(1, 13) = 16.26, p = 0.0014). k-l, Quantification and representative images (24-h time point) of nesting behavior before and 16 d after tumor implantation (two-way ANOVA: Interaction, F(15, 156) = 3.01, p = 0.0003; Time, F(5, 156) = 228.40, p < 0.0001; Treatment, F(3, 156) = 44.61, p < 0.001). m, Daily food intake of GFP and TetTox mice after tumor implantation (two-tailed student's t-test on day 17: t(13) = 6.36, p < 0.0001). TetTox mice were pair-fed to the intake of GFP mice until day 16, after which TetTox mice were allowed ad libitum access to food. Line graphs show mean ± SEM. Box plots show mean (+), median, quartiles (boxes) and range (whiskers) for each treatment. * P ˂ 0.05, ** P ˂ 0.01, *** P ˂ 0.001. See also Supplementary Fig. 8.

Comment in

  • Cancer: Appetite suppressors.
    Yates D. Yates D. Nat Rev Neurosci. 2017 Aug;18(8):457. doi: 10.1038/nrn.2017.86. Epub 2017 Jul 6. Nat Rev Neurosci. 2017. PMID: 28680163 No abstract available.

Similar articles

See all similar articles

Cited by 23 articles

See all "Cited by" articles

References

    1. Hart BL. Biological basis of the behavior of sick animals. Neurosci Biobehav Rev. 1988;12:123–137. - PubMed
    1. Ezeoke CC, Morley JE. Pathophysiology of anorexia in the cancer cachexia syndrome. J Cachexia Sarcopenia Muscle. 2015;6:287–302. - PMC - PubMed
    1. Carter ME, Soden ME, Zweifel LS, Palmiter RD. Genetic identification of a neural circuit that suppresses appetite. Nature. 2013;503:111–114. - PMC - PubMed
    1. Campos CA, Bowen AJ, Schwartz MW, Palmiter RD. Parabrachial CGRP Neurons Control Meal Termination. Cell Metab. 2016;23:811–820. - PMC - PubMed
    1. Roman CW, Derkach VA, Palmiter RD. Genetically and functionally defined NTS to PBN brain circuits mediating anorexia. Nature Comm. 2016;7 - PMC - PubMed

MeSH terms

Feedback