Pancreatic cancer cachexia promotes cardiac dysfunction through altered adrenergic signalling in the heart

Abstract

Abstract

Cancer cachexia is a metabolic syndrome commonly observed in patients with pancreatic ductal adenocarcinoma (PDAC), characterized by wasting of skeletal and cardiac muscle. This condition is associated with the neurohormonal stress response and activation of the sympathetic nervous system. However the impact of cachexia on cardiac adrenergic signalling remains poorly understood. Here we used a preclinical model of PDAC cachexia to investigate how sympathetic input to the heart is altered. We found desensitization of β1‐adrenergic receptor (β1‐AR) in the heart, along with evidence of increased adrenergic tone. Pharmacological blockade with the β1‐selective antagonist metoprolol partially restored adrenergic responsiveness in PDAC mice, suggesting that elevated adrenergic drive contributes to β1‐AR desensitization. The impaired β1‐AR sensitivity blunted the chronotropic response to dobutamine and reduced inotropic reserve under adrenergic stress. Together these findings uncover disruption of cardiac adrenergic signalling in PDAC cachexia.

Key points

• Pancreatic cancer‐associated cachexia promotes β1‐adrenergic receptor desensitization in the heart.

• Pancreatic cancer‐induced β1‐adrenergic receptor desensitization is partly mediated by increased adrenergic transmission to the heart.

• β1‐adrenergic receptor desensitization in mice with pancreatic cancer reduces contractile reserve under adrenergic stress.

Abstract figure legend The β1‐adrenergic receptor is the predominant adrenergic subtype expressed on cardiac myocytes. Upon stimulation it activates downstream signalling pathways that increase the heart rate and cardiac output to meet metabolic demands. Here we show that pancreatic cancer cachexia not only induces cardiac wasting but also drives downregulation and desensitization of cardiac β1‐adrenergic receptors. This remodelling impairs the chronotropic response to dobutamine and reduces contractile reserve when under chronic adrenergic stress. Our data further demonstrate that this desensitization phenotype is mediated, in part, by increased adrenergic signalling and provide new mechanistic insight into cardiac sequelae of pancreatic cancer cachexia.

Introduction

Cancer‐associated cachexia is a debilitating metabolic syndrome observed in advanced malignancies, especially common in pancreatic and gastrointestinal cancers, where it affects up to 80% of individuals (Argiles et al., 2023). Clinically it manifests with unintentional weight loss, anorexia, fatigue, elevated energy expenditure and progressive wasting of both skeletal and cardiac muscle (Anandavadivelan & Lagergren, 2016; Argiles et al., 2023; Fearon et al., 2011; Lena et al., 2023). Despite its prevalence and impact there are currently no Food and Drug Administration (FDA)‐approved treatments to reverse or halt its progression. Importantly cachexia‐induced cardiac wasting occurs independent of chemotherapy status and is a predictor of poor survival in patients (Lena et al., 2023). In addition to driving tissue wasting in the heart emerging data suggest that cachexia also directly affects cardiac function, including alterations in calcium cycling (Law & Metzger, 2021) and mild changes in ejection fraction, stroke volume and global longitudinal strain (Cramer et al., 2014; Labib et al., 2021; Lena et al., 2023; Tadic et al., 2018). These observations raise the possibility that cancer cachexia actively reprogrammes the heart in ways that are clinically meaningful but poorly understood.

Although the mechanisms driving widespread tissue remodelling in cancer cachexia remain incompletely defined, accumulating evidence points to dysregulated neurohormonal signalling, particularly the activation of the sympathetic nervous system, as a key mediator of metabolic dysfunction. In preclinical models enhanced adrenergic drive to adipose tissue promotes lipolysis and browning (Petruzzelli et al., 2014; Xie et al., 2022), contributing to systemic metabolic derangement. Additionally growth and differentiation factor‐15 (GDF15), a key regulator of anorexia and tissue catabolism in cancer cachexia, exerts its effects partly through the activation of sympathetic output to peripheral tissue (Suriben et al., 2020; Wang et al., 2023). In parallel pharmacological blockade of β‐adrenergic receptors alleviates several clinical and metabolic manifestations of cachexia in both patients and tumour‐bearing animals, including anorexia, and loss of fat mass, lean mass and body weight (Hyltander et al., 2000; Petruzzelli et al., 2014; Potsch et al., 2020; Springer et al., 2014; Springer et al., 2023; Stewart Coats et al., 2016; Yuan et al., 2022).

Under physiological conditions sympathetic signalling through β1‐adrenergic receptor supports normal cardiac rhythm and function. Dysregulation of this pathway, as observed in various cardiomyopathies, contributes to electrical instability, structural remodelling and reduced cardiovascular reserve. Emerging clinical observations suggest that similar adrenergic dysregulation also occurs in the context of cancer cachexia. Patients with advanced cancer exhibit elevated resting heart rate and reduced heart rate variability across both low‐ and high‐frequency bands compared to healthy controls, physiological signatures consistent with changes in autonomic input to the heart (Anker et al., 2016; Anker et al., 2020; Chauhan et al., 2012; Cramer et al., 2014; Lee et al., 2016). Additionally a higher prevalence of ventricular arrhythmias is reported in patients with cancer and cancer cachexia (Anker et al., 2021). In support of these data β‐blocker administration reduced resting heart rate and energy expenditure in patients and improved systolic function and survival in animal models of cancer cachexia (Hyltander et al., 2000; Poetsch et al., 2023; Springer et al., 2014). However it remains unknown whether these cardiovascular changes reflect direct alterations in cardiac adrenergic transmission, particularly given that the heart undergoes atrophy rather than hypertrophy in the setting of cancer cachexia.

In this study we investigated the impact of pancreatic ductal adenocarcinoma (PDAC)‐associated cachexia on cardiac adrenergic signalling using an established PDAC cachexia model. We hypothesized that tumour‐induced systemic stress reprogrammes sympathetic input to the heart, resulting in functional changes and impairing the heart's adaptive response. To test this we assessed pre‐ and postsynaptic components of cardiac sympathetic signalling. We also evaluated the cardiac compensatory response to adrenergic stress. We uncovered a previously unrecognized change in cardiac adrenergic signalling and provide new mechanistic insight into the cardiac sequelae of cancer cachexia.

Results

PDAC cachexia desensitizes the β1‐adrenergic receptors in the heart

Given the well‐documented cardiovascular alterations observed in both cancer patients and preclinical models, as well as the protective effects of β‐blockers in this context, we hypothesized that cancer cachexia is associated with increased adrenergic drive to the heart. Chronic activation of β1‐adrenergic receptor (β1‐AR) is known to trigger compensatory downregulation of receptor gene expression and upregulation of G‐protein–coupled receptor kinases (GRKs), particularly GRK2 and GRK5, as well as β‐arrestins, which, in turn, promote receptor desensitization and internalization (Mangmool et al., 2018). This adaptive mechanism limits sustained adrenergic stimulation and cardiac damage elicited by elevated adrenergic tone. Therefore to determine whether PDAC‐associated cachexia leads to increased adrenergic transmission to the heart we evaluated the molecular signature for β1‐AR desensitization in a murine model of PDAC cachexia. We first confirmed key cachectic phenotypes in tumour‐bearing mice, including reduced food intake and significant cardiac atrophy (Fig. 1A–C ). We then performed quantitative PCR analysis on cardiac tissue and found a significant downregulation of Adrb1, the gene encoding the β1‐AR, along with significant upregulation of Grk5 and β‐arrestin isoforms. However Grk2 expression remained unchanged (Fig. 1D ). This gene expression pattern is consistent with chronic β1‐adrenergic stimulation and receptor desensitization (Mangmool et al., 2018). In addition to β1‐AR we assessed expression of other adrenergic receptors and found a significant downregulation of Adra1d, which encodes the ⍺1D‐AR (Fig. 1E ). Given that α1D‐ARs are predominantly expressed in coronary vasculature (Jensen et al., 2009) this change may reflect altered sympathetic input to the vascular compartment as well.

Figure 1. Cardiac β1‐adrenergic receptor (β1‐AR) is downregulated in mice with pancreatic ductal adenocarcinoma (PDAC) cachexia.

A, schematic illustrating study design. B, cumulative food intake over study duration. N = 6 male mice per group. C, heart mass normalized to initial body mass. N = 7 male sham mice and 5 male PDAC mice. D and E, quantitative PCR analysis of cardiac tissue. The eukaryotic 18S rRNA was used as the endogenous control. N = 6–12 male mice per group. F and G, Time course analysis of heart mass and Adrb1 expression in mice with PDAC. N = 5 male sham mice and 6 male PDAC mice per group. H and I, Pair‐feeding study to assess heart mass and β1‐adrenergic receptor desensitization gene signature in PDAC mice. N = 7 male mice per group. J and K, western blot analysis of β1‐AR protein (55 kDa) levels in ventricular and atrial tissue normalized to GAPDH (37 kDa). N = 7 male mice per group. L and M, western blot assessment of β‐arrestin‐1,2 (50 kDa) and GRK5 (68 kDa) in ventricular tissue of mice with PDAC cachexia. N = 7 male mice per group. Data were analysed by unpaired Student's t test (C, D, E, K and M), one‐way ANOVA (H and I) or two‐way ANOVA (B, F and G). Error bars denote SD for all graphs.

To define the temporal relationship between β1‐AR downregulation and cardiac wasting during PDAC progression we conducted a time course study and found that β1‐AR downregulation coincides with the onset of cardiac wasting (Fig. 1F and G ). We next asked whether reduced food intake, a hallmark of cancer cachexia, drove the adrenergic desensitization gene expression signature, or whether these changes resulted from tumour‐specific effects. We conducted a pair‐feeding study, where a cohort of sham‐operated control mice was calorically matched to the intake of tumour‐bearing mice. Interestingly both groups demonstrated cardiac atrophy, but only mice with PDAC cachexia exhibited a gene signature indicative of β1‐AR desensitization in the heart. In contrast sham‐operated pair‐fed mice showed no significant change in the expression of β1‐AR, Grk5 or β‐arrestins compared to sham mice fed ad libitum (Fig. 1H and I ). The presence of cardiac wasting in sham pair‐fed mice without accompanying β1‐AR downregulation suggests that cardiac wasting alone is insufficient to induce β1‐AR downregulation. Moreover these findings indicate that metabolic changes associated with reduced food intake do not drive AR downregulation but instead suggest that the tumour is required to initiate this remodelling programme in the heart.

To further validate the downregulation of β1‐adrenegic receptor at the protein level we performed western blot analysis on both ventricular tissue and atrial tissue, which includes the sinoatrial node region. Consistent with our gene expression data we observed a significant downregulation of β1‐AR protein level in the ventricular compartment (Fig. 1J and K ). Although we did not see a significant change in β1‐AR protein in atrial tissue, variable levels of β1‐AR in PDAC mice compared to sham controls perhaps indicate ongoing reprogramming (Fig. 1J and  K ). We also assessed the protein level for β‐arrestins and GRK5 in the ventricles, which demonstrated no change in GRK5, but the protein level for β‐arrestin‐1,2 was upregulated and approaching statistical significance (Fig. 1L and M ).

We next investigated whether the molecular downregulation of β1‐AR corresponded to functional desensitization. To address this we measured the heart rate responses to dobutamine, a selective β1‐AR agonist, and propranolol, a non‐selective β‐blocker (Fig. 2A ). Interestingly mice with PDAC cachexia exhibited a significantly blunted chronotropic response to dobutamine, indicating impaired β1‐AR sensitivity (Fig. 2B ). Following a 15 min washout period we administered a bolus of propranolol, which elicited comparable heart rate reductions in both PDAC and sham animals (Fig. 2C ). The comparable response to propranolol with a blunted response to dobutamine is consistent with increased norepinephrine (NE) release driving reduced β1‐AR sensitivity and expression. We did not see a difference in baseline heart rate between sham and PDAC mice, which is not surprising given the pre‐ and postsynaptic changes related to heart rate control (Fig. 2D ). However ECG analysis of baseline heart rate revealed prolongation of PR interval in mice with PDAC cachexia, with no change in P‐wave duration or QRS duration (Fig. 2E–H ). Together these data demonstrate β1‐AR desensitization and reprogramming of cardiac conduction in tumour‐bearing animals. Given the observed downregulation of cardiac Adra1d we also assessed blood pressure to evaluate whether this molecular change had functional consequences on systemic haemodynamics. However blood pressure remained unchanged between sham and PDAC mice (Fig. 2I and J ).

Figure 2 Pancreatic ductal adenocarcinoma (. PDAC) desensitizes cardiac β1‐adrenergic receptor in cachectic mice.

A, schematic illustrating study design. B and C, assessment of heart rate response to dobutamine and propranolol treatment. N = 10 sham mice and 11 PDAC mice. One PDAC mouse was excluded from propranolol analysis due to poor injection quality. D, measurement of baseline heart rate in PDAC and sham mice. E–H, representative ECG traces and quantification of PR interval, P duration and QRS duration from baseline heart rate. I and J, assessment of baseline systolic and diastolic blood pressure. N = 4 male sham and 6 male PDAC mice. K, bulk RNA sequencing and unsupervised clustering of atrial and ventricular samples from PDAC and sham mice, focusing on genes involved in cardiac adrenergic signalling. N = 3 male mice per group. L, gene set enrichment analysis of pathways related to cardiac contraction and conduction. Data were analysed by unpaired Student's t test (B–J). Error bars denote SD for all graphs.

To further understand the molecular changes in cardiac adrenergic signalling we performed bulk RNA sequencing on atria and ventricles collected from PDAC and sham mice. Unsupervised clustering of genes involved in cardiac adrenergic pathways distinctly separated sham and PDAC samples, with each group exhibiting a consistent and unique gene expression signature (Fig. 2K ). In addition to β1‐AR several other genes downstream of β1‐AR signalling were downregulated, including multiple genes encoding adenylate cyclase proteins (Adcy1, Adcy3 and Adcy7) and genes involved in calcium handling, including Cacna2d2, Cacnb1, Cacnb2, Camk2a, Camk2b and Camk2d. We also observed altered gene expression of protein phosphatases and their regulatory subunits (Fig. 2K ), which play a central role in dampening adrenergic signalling by dephosphorylating key proteins downstream of β1‐AR signalling. Such shifts in gene expression could have broad effects on calcium dynamics, conduction and contractility. In support of this gene set enrichment analysis of our RNA‐sequencing data revealed significant downregulation of pathways related to cardiac muscle contraction and cardiac conduction in both atrial and ventricular tissues, processes known to be modulated by adrenergic input (Fig. 2L ) (Bers, 2002). Collectively these findings demonstrate that PDAC cachexia profoundly reprogrammes cardiac adrenergic receptor signalling, which beyond reducing β1‐AR sensitivity may also compromise the heart's contractile reserve. Although this reprogramming may result from sustained adrenergic stimulation of the heart, it is possible that alternative mechanisms acting independently of the sympathetic nervous system also modulate cardiac adrenergic signalling.

PDAC cachexia alters neurotransmitter regulation in sympathetic nerves innervating the heart

To determine if PDAC cachexia reprogrammes cardiac adrenergic signalling through alterations in sympathetic drive we first assessed changes in sympathetic innervation, particularly given that cardiac innervation is often disrupted in the diseased heart (Stevens et al., 1998; Fukuda et al., 2015; Wagner et al., 2023; Scalco et al., 2024). Quantification of key axon guidance molecules, including nerve growth factor (Ngf) and semaphorin‐3A (Sema3a), which are critical for sympathetic nerve development and patterning (Ieda et al., 2007; Fukuda et al., 2015), revealed no significant expression changes (Fig. 3A ). Similarly comprehensive histologic quantification of sympathetic nerve fibres across whole‐heart sections showed no change in overall cardiac innervation (Fig. 3B ). In patients with heart failure increased sympathetic transmission is accompanied by increased expression and activity of tyrosine hydroxylase (TH), the rate‐limiting enzyme for catecholamine synthesis. Therefore although elevated NE release and decreased NE reuptake contribute to depletion of NE stores in these patients, elevated TH activity helps maintain cardiac NE stores, even though overall levels remain reduced (Eisenhofer et al., 1996). Therefore we next asked whether NE biosynthesis and stores in nerves innervating the heart are altered in PDAC cachexia. To address this we measured TH protein levels in the stellate ganglia and found a small but significant increase in PDAC animals compared to sham controls (Fig. 3C ). To capture dynamic changes in cardiac NE stores in our model we performed longitudinal analysis of NE content in cardiac sympathetic nerve terminals using high‐performance liquid chromatography (HPLC) across multiple stages of cachexia progression: from precachexia to late cachexia. NE content within cardiac neurons remained unchanged throughout disease progression (Fig. 3D ), suggesting a compensatory increase in NE synthesis that is sufficient to support sustained elevated release without depleting NE stores (Alousi & Weiner, 1966).

Figure 3. Neurotransmitter cycling is disrupted in pancreatic ductal adenocarcinoma (PDAC) mice.

A, quantitative PCR analysis of cardiac axon guidance cues. N = 6–7 male mice per group. B, quantification of sympathetic nerve fibres in cardiac tissue. N = 5 male sham and 6 male PDAC mice. Scale bar: 100 µm. C, western blot analysis of tyrosine hydroxylase protein (62 kDa) levels in stellate ganglia normalized to β‐actin (45 kDa). N = 6 male mice per group. D, high‐performance liquid chromatography (HPLC) measurement of norepinephrine levels in cardiac sympathetic nerve terminals. N = 5 male sham and 10 male PDAC mice per group. E, bulk RNA sequencing of stellate ganglia showing all differentially expressed genes in PDAC mice. N = 6 male mice per group. F, gene set enrichment analysis of stellate ganglia revealing downregulation of pathways involved in neurotransmitter cycling. G, quantitative PCR validation of select differentially expressed genes from bulk RNA sequencing in stellate ganglia. H, schematic illustrating the working hypothesis that disrupted neurotransmitter cycling contributes to increased sympathetic transmission to the heart and subsequent receptor desensitization. Red arrows indicate changes observed in PDAC mice. Heart images in panels B and D indicate that the data were collected from the heart. Data were analysed using unpaired Student's t test (A, B, C and G) or two‐way ANOVA (D). Error bars denote SD for all graphs.

Although innervation density and NE content in nerve terminals remained stable, sympathetic neuronal function appears to be altered through broader reprogramming of neurotransmitter handling and synaptic transmission. To investigate this possibility and other mechanisms that may be altered we performed bulk RNA‐sequencing on stellate ganglia. Several genes involved in synaptic transmission and NE handling, including complexin‐2 (Cplx2), the NE transporter Slc6a2 (NET); the NE degradation enzyme, Maoa; and Adra2a (α2A‐AR), which inhibits neurotransmitter release from presynaptic nerve terminals, were significantly downregulated (Fig. 3E ). We also observed an upregulation in Slc18a1, which encodes the vesicular monoamine transporter‐1, essential for loading NE into presynaptic vesicles (Fig. 3E ). Gene set enrichment analysis further revealed the suppression of pathways related to NE reuptake and enzymatic degradation (Fig. 3F ).

To validate transcriptomic changes we performed quantitative PCR (qPCR) on select key genes. Consistent with our RNA‐sequencing data we found that the expression of Maoa, Slc6a2 and Adra2a was significantly reduced, whereas Slc18a1 (VMAT1) expression was significantly upregulated (Fig. 3G ). Although VMAT1 is typically restricted to small‐intensely fluorescent cells in healthy adult stellate ganglia (Schutz et al., 1998), it is possible that in the setting of cancer cachexia postganglionic sympathetic neurons also express VMAT1 to enhance neurotransmitter packaging and release. Collectively these data suggest a shift towards increased NE packaging and release, impaired NE reuptake and degradation and reduced feedback inhibition of NE release through α2A‐AR. These changes may contribute to elevated efferent signalling and β1‐AR desensitization observed in mice with PDAC cachexia, which is reminiscent of changes observed in heart failure (Eisenhofer et al., 1996) (Fig. 3H ).

Further analysis of our stellate ganglia transcriptomic data revealed upregulation of genes and pathways associated with inflammation, suggesting an increase in the inflammatory landscape of sympathetic ganglia in PDAC cachexia. Prior studies in models of heart failure and myocardial infarction have demonstrated that inflammatory signalling, particularly interleukin‐6 (IL‐6), can drive cholinergic transdifferentiation of adrenergic neurons innervating the heart (Kanazawa et al., 2010; Olivas et al., 2016). To investigate whether a similar phenomenon occurs in PDAC we assessed the expression of choline acetyltransferase (ChAT), the enzyme responsible for acetylcholine synthesis, in postganglionic sympathetic neurons. Immunostaining revealed ChAT expression restricted to preganglionic nerve terminals, with postganglionic sympathetic neurons expressing only TH, indicating a preserved noradrenergic phenotype (Fig. 4A ). In parallel qPCR analysis of stellate ganglia showed no significant difference in Chat mRNA levels between PDAC and sham animals (Fig. 4B ). Taken together these findings indicate that cholinergic transdifferentiation does not occur in cardiac innervating sympathetic neurons during PDAC cachexia, despite a pro‐inflammatory transcriptional environment.

Figure 4 Pancreatic ductal adenocarcinoma (. PDAC) cachexia does not induce cholinergic transdifferentiation of sympathetic neurons in the stellate ganglia.

A, confocal immunofluorescence imaging of stellate ganglia stained for tyrosine hydroxylase (TH) and choline acetyltransferase (ChAT), acquired using a 40× objective. Scale bar: 50 µm. B, quantitative PCR analysis of Chat expression in stellate ganglia. Data were analysed using unpaired Student's t test. Error bars denote SD.

β1‐receptor blockade partially mitigates the desensitization phenotype

Although elevations in plasma NE levels are well documented in patients with heart failure, similar increases are not consistently observed in animal models and do not reliably reflect cardiac sympathetic activity (Scalco et al., 2024). Therefore to definitively determine whether changes in the stellate ganglia enhance cardiac adrenergic transmission and contribute to β1‐AR desensitization we conducted a pharmacological study with metoprolol, a selective β1‐AR antagonist. We implanted osmotic mini pumps at the time of tumour implantation to deliver 50 mg/kg/day of metoprolol. Two days after implantation resting heart rate was significantly reduced in both PDAC and sham mice receiving metoprolol compared to their respective untreated controls, confirming effective β1‐AR blockade (Fig. 5A and B ). At the study endpoint tumour burden, food intake and heart mass were comparable between metoprolol‐treated and untreated PDAC mice, indicating equivalent disease burden (Fig. 5C–E ). These findings also imply that cardiac atrophy in PDAC occurs through mechanisms independent of β‐adrenergic input.

Figure 5. β1‐adrenergic receptor blockade partially attenuated cardiac adrenergic receptor desensitization in pancreatic ductal adenocarcinoma (PDAC) mice.

A, experimental schematic illustrating study design and continuous administration of 50 mg/kg/day of metoprolol via osmotic mini pump. B, resting heart rate measured 2 days post‐tumour and pump implantation in sham and PDAC mice. C, tumour mass at study endpoint. D, cumulative food intake measured at study endpoint. E, heart mass normalized to initial body weight. F–I, quantitative PCR (qPCR) to assess expression of β1‐adrenergic receptor (Adrb1), G‐protein receptor kinase‐5 (Grk5), β‐arrestin‐1 (Arrb1) and β‐arrestin‐2 (Arrb2) in the heart. N = 6–8 male mice per group. One of the sham control hearts was excluded from qPCR analysis due to low RNA quality. Data were analysed using two‐way ANOVA or unpaired Student's t test (C). Error bars denote SD for all graphs.

We next evaluated whether β1‐AR blockade altered the cardiac adrenergic receptor desensitization gene signature. As expected PDAC mice exhibited reduced β1‐AR expression and increased Grk5 expression compared to sham‐operated mice. Treatment with metoprolol attenuated the decrease in β1‐AR expression. However recovery was not complete, suggesting a second mechanism beyond adrenergic signalling is also driving β1‐AR downregulation (Fig. 5F ). This was accompanied by a significant decrease in Grk5 expression. However among metoprolol‐treated mice Grk5 expression remained higher in PDAC animals (Fig. 5G ). Grk5 expression was also significantly reduced in sham mice treated with metoprolol compared to sham controls, and expression of β‐arrestin‐1 and β‐arrestin‐2 remained elevated in PDAC mice regardless of metoprolol treatment (Fig. 3G–I ). Collectively these results indicate that increased adrenergic transmission to the heart contributes to β1‐AR desensitization in PDAC cachexia. However the incomplete recovery of β1‐AR suggests that the desensitization phenotype is not solely dependent on elevated adrenergic signalling to the heart. Moreover the similarities in food intake, heart mass and tumour burden between metoprolol‐treated and untreated PDAC mice indicate that the partial restoration of β1‐AR expression is not merely a consequence of reduced disease severity, underscoring the therapeutic potential of β1‐AR blockade in mitigating PDAC‐induced AR desensitization.

PDAC cachexia reduces the compensatory reserve in the heart

Elevated adrenergic signalling in heart failure initially serves as a compensatory mechanism to enhance contractility, but sustained stimulation becomes maladaptive, leading to β1‐AR desensitization and reduced contractile reserve. We observed a similar desensitization process in PDAC‐induced cachexia, despite the absence of overt heart failure or haemodynamic stress. This led us to hypothesize that cancer cachexia predisposes the heart to impaired contractile reserve under stressors that typically drive heart failure. To test this we employed a well‐established isoproterenol‐induced adrenergic stress model in mice with and without PDAC cachexia. Mice received isoproterenol at 20 mg/kg/day via subcutaneous mini osmotic pumps, a lesser dose than typically used to induce heart failure (30 mg/kg/day) (Fig. 6A ). This dose was chosen for two key reasons: (1) to ensure tolerability in cachectic mice and (2) to apply a submaximal stress that is insufficient to cause decompensation in control mice, thereby allowing us to determine whether PDAC cachexia sensitizes the heart to early functional decline.

Figure 6. Adrenergic receptor desensitization decreases compensatory capacity in heart failure.

A, schematic illustrating study design. B, tumour burden in pancreatic ductal adenocarcinoma (PDAC) mice at the study endpoint. C, quantitative PCR analysis of β1‐adrenergic receptor expression in cardiac tissue. D, cardiac mass measured at study endpoint and normalized to initial body weight. E, echocardiographic measurement of left ventricular mass at study endpoint. F–H, quantitative PCR (qPCR) analysis of cardiac tissue. I, representative M‐mode echocardiographic images in parasternal short‐axis view 11 days post‐tumour and pump implantation. J–R, echocardiographic assessment of heart rate, stroke volume, cardiac output, cardiac index, ejection fraction, left ventricular end‐diastolic and end‐systolic volumes and endocardial area at end diastole and systole 11 days post‐tumour and pump implantation. N = 4–5 male mice per group. Echocardiographic assessments were completed in 4 out of 5 PDAC control mice due to time constraints during the imaging session. The eukaryotic 18S rRNA was used as the endogenous control for qPCR analyses. Data were analysed using two‐way ANOVA or unpaired Student's t test (B). Error bars denote SD for all graphs.

Despite the added adrenergic stimulation tumour mass remained unchanged between PDAC control mice and those treated with isoproterenol (Fig. 6B ). At the molecular level cardiac β1‐AR expression was reduced to the same degree in PDAC control mice and in sham mice receiving isoproterenol, and combined exposure to PDAC and isoproterenol did not result in further downregulation of β1‐AR (Fig. 6C ). As anticipated analysis of cardiac mass revealed significant atrophy in PDAC control mice and pronounced hypertrophy in sham mice treated with isoproterenol (Fig. 6D ). Interestingly the hearts of PDAC mice treated with isoproterenol exhibited an intermediate phenotype, with cardiac mass significantly lower than that of isoproterenol‐treated sham mice but higher than that of PDAC controls (Fig. 6D ). Echocardiographic assessment of left ventricular (LV) mass also revealed a similar pattern (Fig. 6E ). qPCR analysis of cardiac tissue demonstrated upregulation of the atrophy markers Bnip3 and Ctsl, as well as the growth signalling marker Igf1, in PDAC mice treated with isoproterenol (Fig. 6F–H ). These findings indicate that although the hearts of PDAC mice treated with isoproterenol continue to undergo wasting, this process is partially offset by the hypertrophic effects of adrenergic stimulation. As such cardiac wasting may be masked in patients with pre‐existing cardiac hypertrophy or cardiovascular disease.

Echocardiographic analysis revealed a significant increase in heart rate in sham mice treated with isoproterenol. However PDAC mice receiving isoproterenol exhibited a blunted heart rate increase, mirroring the attenuated response observed in our dobutamine challenge experiments (Fig. 6I and J ). We observed a similar pattern of blunted response in stroke volume, cardiac output and cardiac index, further implicating β1‐AR desensitization in PDAC mice (Fig. 6K–M ). Ejection fraction remained unchanged between sham control mice and sham mice treated with isoproterenol, suggesting that the administered dose and duration of isoproterenol elicited an adaptive hypertrophic response to preserve systolic function. In contrast PDAC mice treated with isoproterenol showed a significant reduction in ejection fraction relative to PDAC controls, along with a trend towards reduced ejection fraction compared to isoproterenol‐treated sham mice (Fig. 6N ). Analysis of LV dimensions revealed increased LV end‐diastolic volume (LVEDV) in both sham and PDAC mice treated with isoproterenol, indicating comparable chamber dilatation and diastolic relaxation (Fig. 6O ). However LV end‐systolic volume (LVESV) was significantly elevated in PDAC mice receiving isoproterenol compared to PDAC controls, whereas no difference was observed between sham control and sham‐isoproterenol group (Fig. 6P ). Furthermore LV end‐systolic volume also trended higher in PDAC mice treated with isoproterenol compared to sham‐isoproterenol group, consistent with the observed reduction in ejection fraction. Isoproterenol treatment also increased LV endocardial area at end‐diastole and end‐systole in PDAC mice, whereas sham‐treated animals showed no change (Fig. 6Q and R ). Together these findings demonstrate that PDAC‐induced loss of β1‐AR responsiveness reduces contractile reserve under adrenergic stress.

Discussion

In this study we investigated cardiac adrenergic signalling in a mouse model of PDAC cachexia and provided evidence that the β1‐AR undergoes downregulation and desensitization in response to PDAC. These findings align with clinical observations in patients, where cardiac autonomic signalling and heart rate variability are changed, and importantly these changes are associated with poor survival in patients (Cramer et al., 2014; Guo et al., 2015; Anker et al., 2016; Lee et al., 2016; Zhou et al., 2016; Anker et al., 2020). To our knowledge this is the first report to demonstrate reduced cardiac β1‐AR expression in the context of cancer cachexia. Prior studies in a mouse model of Lewis lung carcinoma did not detect changes in β1‐AR expression, suggesting that β1‐AR downregulation may not be a universal component of cancer cachexia. Although β1‐AR expression remained unchanged in Lewis lung carcinoma, the authors observed a significant downregulation of cardiac Ngf expression, as well as a significant reduction in sympathetic axonal length (Muhlfeld et al., 2011). In contrast despite significant atrophy of the heart and changes in adrenergic signalling in our PDAC model we found no change in sympathetic nerve density or in the cardiac expression of axonal guidance cues. The intact innervation phenotype in our model is also in stark contrast to many other forms of cardiomyopathies, where cardiac sympathetic denervation and remodelling of autonomic input are common features (Stevens et al., 1998; Wagner et al., 2023; Scalco et al., 2024). These findings underscore the possibility that distinct tumour types differentially reprogramme cardiac adrenergic signalling and highlight the need for additional studies to understand tumour‐specific mechanisms governing cardiac autonomic dysregulation in cancer cachexia.

In patients with heart failure β1‐AR downregulation and desensitization are largely driven by altered NE handling, characterized by increased NE spillover and impaired reuptake. Although NE content may initially remain stable, this sustained elevation in NE release with impaired reuptake leads to progressive depletion of catecholamine stores within sympathetic nerve terminals. Interestingly we observed similar features in our PDAC model: although catecholamine content within the heart remained stable, expression of key genes involved in NE reuptake and metabolism, including the NE transporter (Slc6a2), monoamine oxidase (Maoa) and the α2‐AR (Adra2a), was significantly reduced in stellate ganglia. Collectively these changes would be expected to enhance synaptic NE availability and regulate β1‐AR desensitization. However the partial recovery of the adrenergic receptor desensitization signature with metoprolol suggests that beyond adrenergic signalling alternative mechanisms acting independently of traditional neurohormonal pathways also contribute to β1‐AR desensitization in PDAC. This highlights a distinct biology in which, unlike other forms of cardiomyopathy, sympathetic signalling is not the sole driver of receptor desensitization. Moreover because β1‐AR expression was not reduced in models of Lewis lung carcinoma (Muhlfeld et al., 2011), generalized systemic inflammation may not account for the additional desensitization observed. Instead this additional desensitization may be driven by specific tumour‐derived factors or activation of downstream receptor desensitization pathways that are not blocked by receptor antagonism. Therefore further studies are needed to determine additional mechanisms underlying β1‐AR desensitization.

Despite effective β1‐AR blockade with metoprolol we were unable to reverse cardiac wasting in our PDAC cachexia model. Although adrenergic signalling is implicated in the regulation of body weight and wasting of other peripheral tissues in cancer cachexia (Petruzzelli et al., 2014; Suriben et al., 2020; Diba et al., 2024) our findings suggest that cardiac wasting is mechanistically distinct and occurs through pathways independent of β1‐AR input. Prior studies using the novel β‐blocker S‐pindolol reported improvements in body weight in a phase II clinical trial and preservation of cardiac mass in preclinical models (Stewart Coats et al., 2016; Poetsch et al., 2023), but the cardioprotective effects observed are unlikely to result from β1‐AR antagonism alone. In addition to blocking β1‐AR S‐pindolol acts as a β2‐AR agonist and stimulates appetite via central serotonin receptor (5HT1α) antagonism. Therefore the preservation of cardiac mass observed in those studies may have been influenced by increased food intake, an effect not observed in our metoprolol‐treated animals, or trophic signalling through β2‐AR activation. Indeed β2‐AR signalling is a well‐established regulator of cardiomyocyte size, and previous work using the β2‐AR agonist formoterol demonstrated the preservation of cardiac mass in animal models of cancer cachexia (Busquets et al., 2004; Zaglia et al., 2013).

In this study we also demonstrated that AR desensitization in the PDAC heart resulted in reduced contractile reserve under adrenergic stress. Considering that cancer and cardiovascular disease share many risk factors and that the median age of onset for PDAC is 70 years (Grossberg et al., 2020; Wilcox et al., 2024) it is likely that some PDAC patients present with pre‐existing cardiac conditions that heighten their vulnerability to functional decline. Indeed a recent retrospective study in China found that patients with PDAC showed the highest prevalence of comorbid cardiovascular disease among all cancer sites (He et al., 2024). Additionally the psychological stress of a PDAC diagnosis, along with physiological stressors such as surgery and chemotherapy, may further strain the heart and accelerate clinical deterioration. Because metoprolol partially reversed the adrenergic desensitization gene signature, β‐blockers may offer some cardiovascular therapeutic benefits to PDAC patients, especially those with underlying heart disease.

Beyond their cardioprotective effects β‐blockers can have broader therapeutic implications in PDAC patients. Recent studies demonstrated that β1‐AR signalling in PDAC contributes to T‐cell exhaustion and higher tumour burden (Globig et al., 2023). Moreover PDAC tumours are highly innervated with sympathetic nerves, which promote tumour growth, and both β‐blockers and denervation studies have demonstrated reduced tumour burden in animal models (Kim‐Fuchs et al., 2014; Renz et al., 2018; Globig et al., 2023; Thiel et al., 2025; Sattler et al., 2026). β‐blocker use, including β1‐AR selective agents, has also been associated with improved survival in PDAC patients (Udumyan et al., 2017; Beg et al., 2018). Although we did not observe a reduction in tumour size in our metoprolol‐treated mice, this could be due to the rapid tumour growth in our model, which may not have allowed sufficient time for sympathetic nerves to modulate tumour growth dynamics. Collectively all these findings suggest that β‐blockers may confer dual benefits in PDAC by slowing tumour progression and supporting cardiovascular health. However considering that cancer cachexia is often associated with profound fatigue and exercise intolerance these potential benefits should be carefully weighed against the possibility that β‐blocker therapy may further exacerbate functional limitation in this patient population.

Although we observed a clear reduction and desensitization of cardiac β1‐AR, alterations in the cardiac conduction system may also contribute to the blunted chronotropic response to adrenergic stimulation. Supporting this possibility we noted a slight prolongation of the PR interval. Further studies are needed to clarify the functional impact of these electrophysiological changes in cardiac conduction and rhythm. Given the increased prevalence of arrhythmias in patients with cancer, including those with PDAC (Anker et al., 2021), additional work is also warranted to determine whether altered adrenergic signalling contributes to arrhythmogenesis in this context.

One limitation of our study is that we did not remove osmotic pumps delivering isoproterenol prior to echocardiographic studies. This was a deliberate choice, as we also wanted to evaluate if the blunted chronotropic response to β‐AR agonists persists during chronic adrenergic stimulation. However the continuous presence of a β‐agonist likely altered baseline haemodynamics, and it is possible that the removal of the pumps before imaging would have unmasked a more pronounced cardiac functional decline in PDAC mice compared to sham animals treated with isoproterenol. Despite this limitation our findings clearly demonstrate β1‐AR desensitization in PDAC cachexia under both acute and sustained adrenergic stimulation.

In summary our findings demonstrate that PDAC cachexia reprogrammes cardiac adrenergic signalling, a maladaptive process mediated, in part, by excess sympathetic signalling. The restoration of β1‐AR expression by metoprolol underscores the potential for β‐blocker therapy to preserve adrenergic responsiveness and improve cardiovascular function in patients with PDAC. Additional research is needed to clarify additional mechanisms underlying this desensitization and determine whether it occurs in other tumour types or in cachexia caused by conditions such as cirrhosis or chronic infection.

Materials and methods

Ethical approval

All mouse studies were conducted in accordance with the ethical guidelines of The Journal of Physiology and the National Institutes of Health Guide for the Care and Use of Laboratory animals and approved by the Institutional Animal Care and Use Committee (IACUC) of Oregon Health & Science University (OHSU IACUC no. 2126).

Mice

Twelve‐week‐old male C57BL/6J wild‐type mice (JAX catalogue number # 000664) were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and maintained in our animal facility. All mice were housed in a dedicated mouse room with a temperature of 25°C with a 12 h light/dark cycle for the entire duration of the study. Mice had ad libitum access to standard chow diet and water unless otherwise mentioned. In pair‐feeding experiments pair‐fed sham mice were provided with food equivalent to the mean daily intake of PDAC mice over the prior 24 h.

Food intake measurement

Animals were single‐housed 5 days prior to the start of the study to allow for acclimation. Daily food intake was measured by measuring the remaining food in the hopper and accounting for food spillage. To measure food spillage cage bedding was carefully sifted, and any spilled food fragments (orts) were collected and weighed. Daily food intake was calculated as the difference between the initial food provided and the sum of remaining food in the hopper and food recovered from the bedding.

Cell line and tumour implantation

The pancreatic cancer cell line used in the studies was generously provided by Dr. Elizabeth Jaffee (Johns Hopkins University). These cells are derived from tumours in C57BL/6 mice expressing a constitutively active mutant Kras^G12D oncogene and a point mutation in the tumour suppressor gene Trp53^R172H, with both mutations targeted to the pancreas via Pdx1‐Cre‐mediated recombination (KPC cells) (Hingorani et al., 2005). KPC cells were maintained in RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (Corning) and 50 U/ml penicillin/streptomycin (Gibco). Cells were cultured at 37°C in a humidified atmosphere with 5% CO₂ and were routinely tested for mycoplasma contamination prior to use.

For orthotopic tumour implantations mice were anaesthetized with isoflurane, and a small laparotomy was performed in the upper left quadrant of the abdomen to expose the pancreas. The tail of the pancreas was injected with 1×10⁶ KPC cells suspended in 30 µl phosphate‐buffered saline (PBS) or an equivalent volume of cell‐free PBS for sham controls. After implantation the abdominal wall and muscle layers were sutured, and the skin was closed using surgical wound clips. Mice were monitored daily for signs of distress, surgical complications and tumour progression.

Quantitative real‐time PCR

Before tissue harvest animals were anaesthetized with isoflurane and killed by exsanguination. The heart and stellate ganglia were quickly dissected, snap‐frozen in liquid nitrogen and stored at −80°C until analysis. Total RNA was extracted using the RNeasy mini kit (Qiagen, Germantown, MD, 74106) according to the manufacturer's instructions. RNA was reverse‐transcribed to cDNA using the High‐Capacity cDNA Reverse Transcription Kit (Applied biosystems, Foster City, CA, 4368813). PCR reactions were run on an ABI 7300 (Applied Biosystems) using TaqMan universal PCR master mix (Applied Biosystems, 4318157) with the following TaqMan mouse gene expression probes: Adra1a (Mm00442668_m1), Adra1b (Mm00431685_m1), Adra1d (Mm_01328600_m1), Adra2a (Mm_00845383_s1), Adrb1 (Mm00431701_s1), Adrb2 (Mm02524224_s1), Adrb3 (Mm_02601819_g1), Grk2 (Mm00804778_m1), Grk5 (Mm00517039_m1), Arrb1 (Mm.PT.58.17272889), Arrb2 (Mm.PT.58.29778119), Chat (Mm01221880_m1), Slc6a2 (Mm00436661_m1), Maoa (Mm00558004_m1), Slc18a1 (Mm00461868_m1), Bnip3 (Mm01275600_g1), Ctsl (Mm00515597_m1), Igf1 (Mm00439560_m1), Ngf (Mm00443039_m1) and Sema3a (Mm00436469_m1). The eukaryotic 18s rRNA (Applied Biosystems, 4352655) was used as the endogenous control. Relative expression was calculated using 2^∆∆ct method, with values normalized to sham controls. Statistical analysis was performed on the normally distributed ∆ct values.

Bulk RNA‐sequencing

RNA was extracted using the RNeasy mini kit (Qiagen, Germantown, MD) according to the manufacturer's instructions. Quality control and sequencing were performed by Novogene Co. Ltd. (Sacramento, CA, USA). Briefly RNA integrity and concentration were assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA) and nanodrop. Only samples with RNA integrity number (RIN) > 7 were used for library preparation. mRNA was enriched using poly‐T oligo‐attached magnetic beads, fragmented, and reverse‐transcribed using random hexamer primers. RNA‐sequencing libraries were prepared using the NEBNext Ultra II RNA Library Prep kit for Illumina (New England Biolabs), following the manufacturer's protocol. Library quality was checked using Qubit and real‐time PCR for quantification and bioanalyser for size distribution detection. Sequencing was performed on the Illumina NovaSeq 6000 platform, yielding paired‐end reads with an average depth of approximately 25 million reads per sample.

RNA‐sequencing data processing and analysis

Raw FASTQ files were processed using Fastp (version 0.20) to trim adapter sequences and remove low‐quality bases. Read quality was assessed using FastQC (version 0.11.7). High‐quality reads were aligned to the mouse reference genome (mm10) using STAR (version 2.7.1a). HTSeq (version 0.9.1) was used to generate gene‐level count matrices. The resulting count matrix and associated sample metadata were imported into DESeq2 (version 1.38.3) using R (version 4.2.3) for differential gene expression analysis. Genes with an adjusted P‐value < 0.05 were considered significantly differentially expressed and were used for downstream analyses.

Immunofluorescence

Hearts were fixed in 4% paraformaldehyde for 2 h at room temperature and cryoprotected by overnight incubation in 30% sucrose at 4°C. Hearts were embedded in optimal cutting temperature (OCT) compound, frozen and stored at −80°C until analysis. Transverse 20 µm cryosections were obtained using a CryoStar NX50 cryostat (Thermofisher, Waltham, MA). To minimize background fluorescence sections were rehydrated in PBS and incubated thrice for 10 min in 10 mg/ml sodium borohydride (Sigma, St. Louis, MO, 452882). Permeabilization was performed using 0.3% Triton X‐100 for 5 min at room temperature, followed by blocking in 2% bovine serum albumin (BSA) prepared in 0.3% Triton X‐100 for 1 h. Sections were incubated overnight at 4°C with rabbit anti‐tyrosine hydroxylase (TH) antibody (1:500; Millipore, Burlington, MA, AB152). The following day tissues were incubated with a donkey anti‐rabbit Alexa Fluor 555 secondary antibody (1:500; Invitrogen, Carlsbad, CA, a31572) for 2 h at room temperature. Autofluorescence was quenched using the TrueVIEW Autofluorescence Quenching Kit (Vector Laboratories, SP‐8400), and nuclei were counterstained using DAPI (Molecular Probes, D1306). Sections were mounted using ProLong Gold Antifade Mounting Medium (Invitrogen, P36930).

For analysis of stellate ganglia tissues were similarly fixed in 4% paraformaldehyde for 2 h. To reduce autofluorescence tissues were incubated in 0.25% Sudan Black (Sigma, 199664) solution for 2 h at room temperature. The working solution was prepared by diluting a 1% Sudan Black stock solution (prepared in 100% ethanol) 1:4 with 70% ethanol. After five PBS washes ganglia were cryoprotected in 30% sucrose overnight, embedded in OCT and frozen. Serial 20 µm sections were collected and rehydrated with PBS. Permeabilization was performed using 0.1% Triton X‐100, followed by blocking with 5% BSA in 0.1% Triton X‐100 for 1 h at room temperature. Samples were incubated overnight at 4°C with rabbit anti‐tyrosine hydroxylase (1:500; Milipore, AB152) and goat anti‐choline acetyltransferase antibody (Millipore, AB144P). The next day tissues were washed and incubated with donkey anti‐rabbit Alexa Fluor 488 (Thermofisher, A21206) and donkey anti‐goat Alexa Fluor 555 (Thermofisher, A21432) secondary antibodies (both at 1:500 dilution) for 2 h at room temperature. Coverslips were applied using ProLong Gold mounting medium.

Image acquisition and analysis

All stellate ganglia images were acquired using a Nikon confocal microscope with a 40× objective. For cardiac innervation analysis four sections per heart were evaluated. Whole‐section tile images were acquired using a ZEISS ApoTome2 system, followed by postprocessing and stitching using Zen microscopy software. Nerve density was quantified by applying consistent intensity thresholds to each image and measuring the area of TH+ nerves, DAPI+ nuclei and total tissue area. TH+ nerve area was normalized to DAPI+ area to account for tissue size and cellularity.

Western blot

Heart tissue and stellate ganglia were harvested, snap‐frozen in liquid nitrogen and stored at −80 °C until use. For total protein extraction from heart samples tissues were homogenized using a Bead Mill 24 homogenizer in ice cold lysis buffer containing: 1× RIPA buffer (Cell Signaling, Danvers, MA, 98065), 1mM phenylmethanesulfonyl fluoride solution (Sigma, 93482), protease inhibitor cocktail EASYpack (Roche, Indianapolis, IN, 05892970001), phosphatase inhibitor cocktail 2 (Sigma, P5726) and additional protease inhibitor cocktail (Sigma, P8340). Homogenates were briefly sonicated and centrifuged at 16,000 g for 15 min at 4°C. The resulting supernatant was collected, and total protein concentration was determined using a bicinchoninic acid (BCA) assay (ThermoFisher, 23227) according to the manufacturer's instructions. For western blot analysis 35 µg of total protein was loaded per ventricular sample and 30 µg of total protein per atrial sample.

Stellate ganglia were homogenized in 120 µl of the same lysis buffer by brief sonication, followed by centrifugation at 16,000 g for 15 min at 4°C. The supernatant was collected, and total protein concentration was measured using the same BCA assay. A total of 5 µg of protein was used per sample for western blotting.

Proteins were separated on Nu PAGE 4%–12% bis‐Tris polyacrylamide gradient gels (Invitrogen, Carlsbad, CA, NP0323BOX) for heart samples or 8%–16% Tris‐Glycine plus Midi Gel (WXP81620BOX) for stellate ganglia and transferred to PVDF membranes (Millipore, IPFL00010). Membranes were blocked in 5% BSA for 1 h at room temperature, then incubated overnight at 4°C with primary antibodies under gentle agitation. After three washes in TBST membranes were incubated with secondary antibodies for 1 h at room temperature. Protein bands were visualized using the Li‐Cor Odyssey imaging system. A complete list of antibodies can be found in Table 1.

Table 1.

Complete list of western blot antibodies

Primary antibodies
Antibody	Host	Catalogue number
Anti‐β1‐AR	Rabbit	Cell Signaling, 12271S
Anti‐tyrosine hydroxylase	Rabbit	Cell Signaling, 2792S
Anti‐β‐arrestin 1,2	Rabbit	Cell Signaling, 4674S
Anti GRK5	Mouse	Santa Cruz, sc‐518005
Anti‐GAPDH	Mouse	Cell Signaling, 97166S
Anti‐β‐actin	Mouse	Cell Signaling, 3700S
Secondary antibodies
Antibody	Fluorophore	Catalogue number
Anti‐rabbit IgG	DyLight 680	Cell Signaling, 5366S
Anti‐mouse IgG	DyLight 800	Cell Signaling, 5257S
Anti‐mouse IgG	DyLight 680	Cell Signaling, 5470P

In vivo pharmacological studies

To evaluate the effects of chronic adrenergic stress or β1‐adrenergic receptor blockade mice were implanted with Azlet osmotic minipumps (1002, Azlet) delivering 20 mg/kg/day isoproterenol hydrochloride (Sigma, 420355) or 50 mg/kg/day metoprolol tartrate (Sigma, M5391), respectively. Pumps were implanted subcutaneously in the interscapular space under isoflurane anaesthesia according to manufacturer's instructions. Control animals underwent sham surgery involving a small incision without pump placement. At the time of pump placement animals also underwent orthotopic implantations of either KPC cells or sterile saline in the pancreas, as explained in detail above. After pump and tumour implantation mice were monitored daily for signs of distress, surgical complications and tumour progression.

Echocardiography

Echocardiographic imaging was done by the Small Animal Research Imaging Core (SARIC) at OHSU. Transthoracic echocardiography was performed using a Vevo2100 high‐frequency ultrasound system (VisualSonics), with mice under isoflurane anaesthesia. Body temperature was continuously monitored with a rectal probe and maintained at 37°C using a temperature‐controlled heating platform. Heart rate was recorded via the ECG electrodes integrated into the imaging platform and calculated from the R–R interval. Two‐dimensional images were acquired in both the parasternal long‐axis view and the mid‐papillary short‐axis view. LV endocardial cross‐sectional area and major axis length were measured at both end‐diastole and end‐systole (ENDO_area;d, ENDO_area;s, ENDO_major;d and ENDO_major;s, respectively). LVEDV and LVESV were calculated using the formula: LV volume = (5/6) ENDO_major × ENDO_area. Stroke volume was determined by multiplying the LV outflow tract area by the time–velocity integral obtained via pulsed‐wave Doppler. Doppler recordings were acquired from a high parasternal long‐axis view, with consistent heel–toe angulation to ensure a comparable angle of incidence across animals and within the range of automated angle correction. Ejection fraction was calculated using the formula: EF = (SV/LVEDV) × 100.

Heart rate assessment

Mice were anaesthetized under isoflurane anaesthesia. Body temperature was continuously monitored with a rectal probe and maintained at 37°C using a temperature‐controlled heating platform. Heart rate and ECG recordings were obtained using subcutaneous electrodes connected to an ECG platform, with data acquired and analysed using LabChart Software (ADInstruments). After electrode placement a baseline ECG was recorded for 5 min to allow stabilization. Baseline heart rate was defined as the average heart rate over the final 2 min of this stabilization period. In addition ECG parameters, including PR interval, P‐wave duration and QRS complex duration, were measured in LabChart from the stabilized baseline recordings prior to any drug administration.

To assess adrenergic responsiveness mice received 1.5 mg/kg dobutamine (OHSU pharmacy) in 50 µl sterile saline via retro‐orbital injection, and heart rate was continuously recorded. The peak heart rate response to dobutamine was identified, and the average heart rate over a window centred on the peak was calculated. The change in heart rate was expressed as the difference between peak heart rate and baseline heart rate. After a 15 min washout period, during which heart rate returned to baseline levels, mice received 4 mg/kg propranolol hydrochloride (Sigma, P0884) intraperitoneally in 100 µl volume. Heart rate was continuously recorded for 4 min following propranolol administration. The average heart rate during the final 2 min of this period was used to assess the propranolol effect. Changes in heart rate were calculated relative to baseline heart rate measured during the 2 min immediately prior to propranolol injection. One PDAC animal was excluded from the propranolol analysis due to poor injection quality.

Non‐invasive blood pressure assessment

Blood pressure was assessed using a tail‐cuff system (MC4000 Multichannel Blood Pressure Analysis System; Hatteras Instruments, Grantsboro, NC). Measurements were collected over 10 cycles and averaged for analysis.

High‐performance liquid chromatography

NE was measured as previously described (Li et al., 2003; Parrish et al., 2009). Briefly the LV was rapidly frozen on dry ice and pulverized using a prechilled mortar and pestle. Pulverized tissue was weighed and homogenized in 0.1 M perchloric acid containing 1.0 µM dihydroxybenzylamine (DHBA) as an internal standard to correct for sample recovery. Catecholamines were isolated using alumina extraction. After two washes in deionized water catecholamines were eluted from the alumina with 0.1 M perchloric acid. Samples were analysed by reversed‐phase HPLC using a C18 column (Agilent Microsorb, 150 × 4.6 mm, 5 µm). Separation was achieved with a mobile phase containing 75 mM sodium phosphate buffer (pH = 3.0), 1.7 mM sodium octane sulfonate and 3.0% acetonitrile. Catecholamines were detected using an electrochemical detector (Coulochem III; ESA, Bedford, MA, USA) with the electrode potential set to +180 mV. NE standards (0.5 µM) were processed in parallel with tissue samples to ensure accurate quantification.

Statistical analysis

All data are presented as mean ± SD. Statistical analyses were performed using GraphPad Prism version 10. Comparisons between two groups were made using unpaired Student's t test. One‐way ANOVA was used for comparisons among three groups, and two‐way ANOVA followed by Tukey's multiple‐comparisons test was used for analyses involving factorial studies. A P‐value of < 0.05 was considered statistically significant.

Competing interests

D.L.M. is a consultant for Pfizer, Inc. and Alkermes, Inc. D.L.M. is a consultant, has received grant funding, is the chief medical officer and has equity in Endevica Bio Inc. A.J.G. is a consultant for Endevica Bio, Inc. The other authors declare no competing interest.

Author Contributions

Conceptualization: P.D., D.L.M., A.J.G. Conducting experiments: P.D., P.R.L., S.N., K.M.S.K., P.C.A.W., X.Z., H.M., L.W., A.S., W.R.W. Analysing data: P.D., P.R.L., P.C.A.W., P.P., X.Z., L.W., D.L.M., B.A.H., A.J.G. Supervision: D.L.M., B.A.H., A.J.G. Writing − original draft: P.D. Writing − review and editing: P.D., P.R.L., S.N., K.M.S.K., P.C.A.W., X.Z., N.B., P.P., H.M., L.W., A.S., W.R.W., B.A.H., D.L.M., A.J.G.

Funding

This study was funded by the National Cancer Institute of the National Institute of Health, R01 CA264133 (A.J.G.); National Cancer Institute of the National Institute of Health, R37 CA280692 (A.J.G.); National Heart, Lung, and Blood Institute of the National Institute of Health, R01 HL093056 (B.A.H.); National Heart, Lung, and Blood Institute of the National Institute of Health, R01 HL146833 (B.A.H.); National Heart, Lung, and Blood Institute of the National Institute of Health, F30HL170476 (P.D.); the Oregon Clinical and Translational Research Institute (OCTRI), TL1TR002371 (P.D.); and the National Cancer Institute of National Institute of Health, K99CA286709 (P.C.A.W.).

Generative AI statement

No generative AI tools were used in the preparation of this manuscript.

Supporting information

Peer Review History

Biography

Parham Diba is an MD/PhD student at Oregon Health & Science University (OHSU), currently completing the final 2 years of medical school. He earned his undergraduate degree from the University of Oregon, where he studied genes involved in neural crest development. He then worked as a research assistant studying mechanisms of brain injury in premature infants before matriculating at OHSU. He received his PhD in biomedical sciences in 2026 under the mentorship of Dr. Aaron Grossberg, focusing on how pancreatic cancer cachexia remodels cardiac adrenergic signalling and the cardiac immune microenvironment.

Data availability statement

All data supporting the findings of this study are available within the paper and supplementary materials. Bulk RNA‐sequencing data are deposited in BioProject under accession ID PRJNA1309756. The RNA‐sequencing analysis pipeline is available at https://github.com/maxsonBraunLab/Bulk‐RNA‐seq‐pipeline‐PE. Additional raw data and materials are available from the corresponding author upon reasonable request.