Why do Pancreatic Tumor Cells Love the Liver?
Pancreatic ductal adenocarcinoma (PDAC) and its precursors secrete high levels of tissue inhibitor of metalloproteinases 1 (TIMP1) into the circulation, which activates hepatic stellate cells (HSCs) in liver. These cells help create a niche that promotes metastasis of pancreatic tumors to liver, researchers report in the November issue of Gastroenterology. Strategies to block TIMP1 signaling and HSC activation might therefore be developed to prevent PDAC metastasis.
PDACs metastasize to liver at early stages. Researchers have believed that this is because PDACs secrete factors that create a premetastatic niche, to support homing and proliferation of incoming tumor cells.
TIMP1 is one of the only tumor-derived secreted factors identified that can create such a niche in the liver. Levels of TIMP1 correlate with poor outcomes of patients with PDAC and with premalignant pancreatitis. TIMP1 has been associated with activated HSCs, which make the liver more susceptible to incoming tumor cells by producing cytokines, chemokines, and growth factors and promoting inflammation.
Barbara Grünwald et al investigated whether HSCs respond to TIMP1 secreted from pancreatic lesions to promote PDAC metastasis to liver.
The authors noticed that chronic pancreatitis, pancreatic intra-epithelial neoplasias (PanIN), and PDAC tissues from patients expressed higher levels of TIMP1 protein than normal pancreas. From PanIN1 to PDAC, ductal cell expressed levels of TIMP1 (see figure).
In PanIN3 and PDAC lesions, TIMP1 became detectable in pancreatic stellate cells. Plasma levels of TIMP1 were associated with incidence of liver metastases in patients.
Then, they found that in mice genetically engineered to develop pancreatic cancer (Ptf1a+/Cre;Kras+/LSL-G12D;Trp53loxP/loxP mice), plasma levels of TIMP1 greatly increased during disease progression. In KPC mice, which develop pancreatic cancer via PanIN lesions, levels of TIMP1 in pancreas and blood increased from the pancreatitis stage through progression of PDAC.
Grünwald et al found that livers of mice with PanINs had activated HSCs, and that disruption of the gene encoding TIMP1 reduced activation of HSCs during PDAC development. HSCs were also activated in liver tissue from patients with chronic pancreatitis or PDAC patients, compared with livers of healthy individuals. Markers of HSC activation correlated with TIMP1 expression in matched pancreas and plasma samples.
Transgenic expression of TIMP1 was sufficient to activate HSCs in tumor-free mice, and incubation of cultured HSCs with TIMP increased their expression of markers of activation and promoted their motility.
But how does TIMP1 activate HSCs, and how does this promote tumor metastasis to liver? Grünwald et al found TIMP1 to activate HSCs by inducing PI3K signaling in the cells (rather than via TIMP’s well-known anti-proteolytic activity). Premalignancy-induced activation of HSCs was associated with an increase in numbers of neutrophil numbers in the liver, which depended on TIMP1 and its receptor, CD63.
HSCs from mice expressing high levels of were induced to express C-X-C motif chemokine ligand 12 (CXCL23 or also called SDF1), which attracts neutrophils. Neutrophils have been shown to promote formation of a metastatic niche in several tissues, and specifically in the liver.
So, TIMP1 signals via CD63 to activate HSCs to express SDF1 (in vitro and in vivo), which attracts neutrophils and helps create the premetastatic niche.
Knockout or knockdown of CD63 in mice reduced metastasis of pancreatic tumors to liver, as did pharmacologic inhibition of TIMP1–CD63 signaling via PI3K. This pathway might therefore be blocked to prevent PDAC metastasis in patients.
The authors say that pancreatic tumor aggressiveness is determined by path-paving mechanisms that occur before and during tumor formation, and that their findings provide a clue has to how disseminated premalignant pancreatic cells are attracted to and begin to grow in a distant organ.
Studies are needed to determine whether increased levels of TIMP1 in patients with pancreatitis or pancreatic lesions indicate that they are at high risk for metastatic PDAC—this might provide a window of opportunity for early intervention.