Background

Gastric cancer (GC) is now the fifth most common malignant cancer globally, and the number of GC cases in China accounts for >40% of all new cases of GC in the world [1]. Most patients have a definite diagnosis of GC only in the advanced phase and miss the chance to undergo radical surgical treatment [2]. The prognosis of patients with GC remains poor. Therefore, it has become an urgent need to conduct in-depth research on the pathogenesis of GC and to identify effective therapeutic targets.

Mitogen-activated protein kinase (MAPK) is a class of serine/threonine protein kinases that can be activated by various intracellular and extracellular stimuli, including growth factors, hormones, oxidative stress, and endoplasmic reticulum stress [3]. The MAPK signal transduction pathway consists of three types of sequentially activated protein members: MAP kinase kinase kinase (MAPKKK or MEKK), MAP kinase kinase (MAPKK or MEK), and MAPK, which play a role in enhancing the expression of target genes or directly acting on cytoplasmic downstream kinases, regulating cell proliferation, differentiation, stress response and cell apoptosis, and other physiological activities [4]. Each MEK can be activated by at least one MEKK, and each MAPK can be activated by different MEKs, forming a complex regulatory network of MAPK [5, 6]. MAPK consists of three main subgroups: extracellular signal-regulated kinase (ERK), c-Jun amino-terminal kinase (JNK), and p38 [7, 8]. Abnormal expression or overexpression of MAPK members plays an important role in the malignant transformation and evolution of cells.

Histone deacetylases (HDACs) are a hotspot in the field of cancer drug development. Inhibition of histone deacetylation has become a recognised approach for tumour therapy [9,10,11]. HDACs are involved in the regulation of tumour proliferation, invasion, and migration [12, 13]. Until now, 18 HDAC subtypes have been found in the human body, which can be further subdivided into four categories: Class I HDACs (HDAC1–3 and 8) mainly exist in the nucleus, and their main function is the deacetylation of histones. Class II HDACs are further divided into Class IIA (HDAC4, 5, 7, and 9) and Class IIB (HDAC6 and 10). Class IV HDAC11 is only expressed in the brain, kidney, and testes. Class III HDAC (SIRT1–7) is associated with the yeast protein SIR2 [14, 15]. Different types of HDACs have different structures, and their effects are also different. Different subtypes of HDACs also have great differences in their baseline expression levels as well as the mechanism of action in different tumour tissues [16,17,18]. The role of HDACs in GC development has been studied but the mechanisms are inadequately understood [19, 29]. We demonstrate the same results using real clinical data. Kang et al. found that HDAC4 promotes GC progression via p21 repression [30]. We found a different mechanism: HDAC4 facilitates the progression of GC mainly by activating the p38 MAPK pathway. Our results suggest that high expression of HDAC4 may be a poor predictor of GC.

As an important pathway of intracellular protein degradation, the autophagy-lysosomal system plays an important role in both nutrient cycling and scavenging and maintenance of stability [31]. Target proteins degraded by the autophagy-lysosomal system, such as WNT and Keap1, first bind to the key autophagy protein p62/LC3B and are recognised by the receptor proteins [32, 33]. Then they are wrapped by the autophagosome with a bilayer membrane structure, after which they enter the autophagy lysosomes to complete the autophagic degradation of the proteins. We found that HDAC4 knockdown enhances the autophagic degradation of MEKK3 and reduces the expression of MEKK3 in cells, thus inhibiting the activation of the MAPK pathway and the proliferation, migration, and invasion of GC cells.

HDAC4 plays different roles by regulating autophagy. After HDAC4 interacts with autophagy-related microtubule-associated protein 1S (MAP1s), the acetylation level of MAP1s decreases, and it becomes unstable. This inhibits autophagy and promotes the accumulation of MHTT aggregates, causing the occurrence of Huntington’s disease. The polyamine spermidine can improve MAP1s instability induced by HDAC4 and inhibit the occurrence of cirrhosis and hepatocellular carcinoma by promoting autophagy [34]. In diabetic nephropathy, HDAC4 promotes the deacetylation of signal transduction and transcriptional activator 1 (STAT1), and activated STAT1 inhibits podocyte autophagy, thereby inducing podocyte injury [35, 36]. However, during vascular inflammation, the increased expression of HDAC4 can reduce the acetylation of FoxO3a in vascular endothelial cells, and activated FoxO3a can promote the transcription of autophagy-related genes ATG5 and LC3B, thereby inducing the autophagy of vascular endothelial cells [37]. In our study, HDAC4 inhibited the transcription of the autophagy-related gene ATG4B and consequently autophagy in GC.

Traditional HDACs contain the amino acid tyrosine in their enzyme active region; however, for type II HDACs, the tyrosine is replaced by histidine, so that their activity is >1000 times lower than that of type I HDACs. Class II HDACs have a type of protein structure that has a specific amino acid sequence targeting the acetyl modification of lysine and can recruit HDAC3. HDAC3 can perform the deacetylase activity in case of class II HDAC deletion and can continue to bind to the NCoR/SMRT transcription co-inhibitory complex, remove the acetyl groups of histones and non-histone proteins, and inhibit DNA transcription [38]. Non-histone proteins studied in recent years mainly include runt-associated transcription factor 2, hypoxic-inducible factor-1α, and STAT1 [39,40,41]. Class II HDACs can also bind to transcription factors such as MEF2s, thereby inhibiting the transcription of genes regulated by these transcription factors [42]. Our study also confirmed that HDAC4 in GC cells inhibits the expression level of ATG4B by inhibiting the effect of MEF2A on the transcription of ATG4B, thus inhibiting the autophagy of GC cells.

In conclusion, our study confirmed that HDAC4 plays an important role in the development of GC, and high HDAC4 expression can be used as an independent predictor of poor prognosis of GC. High expression of HDAC4 inhibits the transcriptional activity of MEF2A, which in turn inhibits the transcription of ATG4B, thereby inhibiting the autophagy of GC cells, reducing the degradation of MEKK3, activating p38, and promoting the growth and metastasis of GC. Therefore, HDAC4 can be used as a new potential GC therapeutic target.