Introduction
GC is a deadly disease accounting for ~768,793 deaths worldwide in 2020, making it the 4th leading cause of cancer-related death [
1]. GC therapy comprises surgery with additional adjuvant or neo-adjuvant chemo- and radio-therapies and the number of relapses remains high. Surgical resection which remains the only curative therapeutic option is possible in case of early tumours but GC is commonly a late-diagnosed disease with 80% of patients detected at metastatic state and with a 5-year survival rate of <5% [
2,
3]. It is thus crucial to find appropriate diagnostic tools for earlier detection of GC as well as strategies for better-targeted therapy of disseminated cancers.
The most aggressive component of heterogeneous gastric tumours is the rare population of cancer stem cells (CSC) which carry specific self-renewal and chemo-/radio-resistant properties [4,5,6,7], allowing them to initiate tumours and cause recurrence. In the metastatic process, part of these CSC corresponding to metastasis-initiating CSC (MIC), is able to acquire invasive characteristics, evade the primary tumour, disseminate, and colonise distant organs to initiate metastases [
8,
9].
We have recently identified CD44v3+ cells as MIC in gastric tumours, associated with bad prognosis in GC [8] and which could constitute an interesting therapeutical target for GC metastatic disease. CD44v3+ cells are a subpopulation of gastric CD44+ CSC and our previous work has demonstrated an enrichment of the Hippo pathway oncogenic members Yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ) in GC and especially in gastric CSC [
10,
11,
12]. YAP and TAZ are well-described for their role in invasion and metastasis and their targeting could affect GC MIC.
We have demonstrated that activating the Hippo kinase core with leukaemia inhibitory factor (LIF) decreases tumorigenic properties of gastric CSC [
4]. Nevertheless, LIF/LIF receptor (LIFR) signalling was never tested in the gastric metastatic context despite its promising anti-metastatic effects in hepatocellular carcinoma [
13], clear cell renal carcinoma [
14] and breast cancer [
15].
LIF–LIF receptor (LIFR) canonical pathway is mediated by JAK/STAT activation, causing STAT3 phosphorylation, dimerisation, and nuclear translocation, resulting in the transcription of target genes involved in cell survival and tumorigenesis [
16]. Other cell signalisation processes have been identified downstream of LIF-LIFR, which could explain its pleiotropy [
16,
17,
18,
19,
20,
21,
22,
23]. Nonetheless, LIF-LIFR-dependent anti-metastatic properties in breast cancer implicated the Hippo pathway activation [
15].
The aim of this study was to decipher LIF’s potential as anti-metastatic therapy in GC and the underlying signalling mechanisms. We have shown that LIF decreased invasive properties of GC cell lines and patient-derived xenograft (PDX) cells and more especially gastric CSC in vitro through activation of Hippo kinases and repression of YAP/TAZ oncogenic signalling. LIF’s anti-invasive function involves the repression of CD44v3 expression and of the epithelial-to-mesenchymal transition (EMT) programme in CSC, important in the GC metastatic process. LIF-treated cells were also less keen to form metastasis compared to non-treated cells when injected into mice’ blood circuits. Finally, LIFR’s high expression in gastric tumours seems to be protective since it is correlated to better prognosis of GC patients despite them having CD44v3+ MIC-rich or mesenchymal ZEB1+ tumours.