PDA Got SERious Nerves

Abstract

Pancreatic innervation is an important factor in pancreatic cancer etiology and progression. Recent work by Banh et al. has revealed that serine released from the axons of sensory and sympathetic neurons supports pancreatic cancer metabolism during nutrient-deprived conditions. These findings rationalize a therapeutic strategy to combine dietary manipulation and pharmacological denervation to target pancreatic cancer.

Pancreatic ductal adenocarcinoma (PDA) has a complex and unique tumor microenvironment (TME) consisting of extracellular matrix, fibroblasts, immune cells, endothelial cells, and nerves. Although the TME is a crucial mediator of PDA growth, metastasis, and resistance to treatment, the role of neural paracrine signaling in these processes is poorly understood. In recent years nerves in the TME have emerged as an important contributor to malignancies such as prostate [1], gastric [2], and pancreatic [3,4] cancer. Of these cancer types, PDA exhibits the highest prevalence of perineural invasion – the neoplastic invasion of tumor cells into or surrounding the nerves – at a frequency of 80–100% [5]. Although nerves appear to facilitate tumor growth in most contexts, the role of neural input in PDA development is more complex. Whereas sensory and sympathetic nerves promote pancreatic neoplasia through neuropeptide signaling [3], cholinergic signaling through parasympathetic nerves suppresses the growth of PDA cells [4].

In deciphering the mechanism of neuron–tumor interaction, much focus has been placed on paracrine neurotrophic signaling [1–4]. However, given that nutrient deprivation is a hallmark of the pancreatic TME, this raises the question of whether intratumoral innervation might contribute to the metabolic reprogramming of PDA tumor cells. Indeed, in a recent study published in Cell, Banh et al. [6] reported that sensory and sympathetic neurons deliver serine (Ser) through intratumoral axonal projections to support the growth of exogenous Ser-dependent PDA cells during serine/glycine (Ser/Gly) deprivation. Although Ser deprivation limits mRNA translation in PDA cells, it was found to induce the secretion of nerve growth factor (NGF), which then promotes intratumoral innervation to supply PDA cells with axon-derived Ser. As such, pharmacological inhibition of the TRK family of neurotrophin receptors using LOXO-101 suppressed pancreatic tumor growth in mice on a Ser/Gly-free diet (Figure 1A). This study establishes axon–cancer metabolic crosstalk as a therapeutically targetable adaptation that supports PDA growth in nutrition-poor environments.

Figure 1. Neural Support of Pancreatic Ductal Adenocarcinoma (PDA) via Serine Release and Future Areas in Neural Cancer Research.

(A) High levels of extracellular Ser in the PDA microenvironment sustain the growth of exogenous-Ser (ex-Ser)-dependent PDA cells by promoting Ser-tRNA charging, Ser codon translation, and protein synthesis. Ser deprivation leads to selective Ser-codon ribosome pausing allowing enhanced nerve growth factor (NGF) translation. Secreted NGF stimulates tumor innervation to provide PDA cells with axon-derived Ser. Pharmacological blockade of the neurotrophin receptor TRK by LOXO-101 suppresses NGF-induced axonogenesis, leading to arrested tumor growth. (B) Three areas for future neuron–cancer research in PDA include (i) cell type-specific neuron–PDA interplay, (ii) crosstalk with other components in the tumor microenvironment (TME), and (iii) cancer-associated neural remodeling.

Protein synthesis is the most energy-consuming process in the cell and is therefore tightly controlled by amino acid availability. In eukaryotes, amino acid limitation inhibits mTORC1 signaling and activates the kinase GCN2, resulting in an overall decrease in protein synthesis through suppression of translation initiation [7]. This leads to a decrease in the number of elongating ribosomes, which are major consumers of the cytosolic amino acid pool. An alternative response to amino acid limitation is ribosome pausing at a subset of synonymous codons, a phenomenon that is observed in bacteria deprived of auxotrophic amino acids, and more recently in mammalian cells during arginine limitation [8], resulting in abortive termination and decreased translation [9]. Interestingly, exogenous Ser-dependent PDA cells were found to regulate translation through a similar mechanism. Specifically, Ser deprivation in these cells results in ribosome stalling selectively at two of the six Ser codons, TCC and TCT. Owing to differential codon usage bias among different mRNAs, the effects of Ser deprivation on protein synthesis are transcript-specific, and developmental pathways and G protein-coupled receptors (GPCR) signaling are the most strongly affected, whereas transcription and secreted soluble factors are least affected. Notably, the two Ser codons at which ribosome pausing is observed are not rare codons or decoded by low-abundance tRNAs, suggesting that nutrient context is likely crucial for defining the specific codons or tRNAs that are functionally optimal for efficient translation.

These findings considerably advance our understanding of nerve–PDA interactions and raise intriguing questions for future studies. For example, do the axons of parasympathetic nerves, like those of sensory and sympathetic nerves, also release Ser to support the growth of PDA cells? In addition, what are the mechanisms and factors that drive non-neurotransmitter amino acid release from TME sensory neurons (Figure 1B)? It is known that astrocytes take up glucose from the blood vessels to synthesize and export L-Ser, which is then imported by neurons for generating the neurotransmitter D-Ser by serine racemase [10]. It would be interesting to examine whether serum levels of D-Ser, a proxy for serine racemase activity, can serve as a biomarker to identify the patient population responding to the TRK inhibitor LOXO-101, as well as in other settings where Ser is limiting, such as KEAP1-mutant lung tumors.

We are only beginning to understand the bidirectional crosstalk between pancreatic cancer cells and the nervous system. Of the three divisions of the autonomic nervous system, much focus to date has been on the sympathetic and parasympathetic systems. Because pancreatic cancer is also populated by intrapancreatic ganglia from the enteric nervous system (ENS), examination of the role of the ENS in pancreatic tumorigenesis would be of great interest. It is known that hyperinnervation coincides with regions of strongest desmoplastic activity in pancreatic tumors [5]. Despite this, the impact of the peripheral nervous system on the various components of the pancreatic TME, such as cancer-associated fibroblasts, vasculature, and immune cells, remains elusive. Dissecting the interaction between neurons and these stromal compartments may thus shed new light on TME remodeling and potentially on the treatment of pancreatic cancer. Lastly, cancer cell-derived paracrine factors can induce profound nervous system remodeling and dysfunction. Locally, this can result in perineural invasion and chronic pain syndromes. Systemically, cancer-derived paracrine factors can induce aberrant synaptogenesis and interfere with normal brain functions, and may contribute to sleep disorders in cancer patients (Figure 1B). Thus, further understanding how PDA cells structurally and functionally remodel the nervous system should facilitate the development of non-opioid approaches to minimize the debilitating neurological side effects of PDA therapy and alleviate cancer pain during palliative care.