Zhi-Qiu Hu , Yi Lu , Di Cui , Chn-Yang Ma , Su Shao , Ping Chn , Ran Tao b , ,Jian-Jun Wang , ?

a Department of Surgery, Minhang Hospital, Fudan University, Shanghai 201199, China

b Department of Hepatobiliary-Pancreatic & Minimally Invasive Surgery, Affiliated Zhejiang Provincial People’s Hospital, Hangzhou Medical School, Hangzhou 310014, China

c Center for Clinical Medical Research, Affiliated Zhejiang Provincial People’s Hospital, Hangzhou Medical School, Hangzhou 310014, China

d Department of General Surgery, Chun’an 1st People’s Hospital, Hangzhou 311700, China

e Department of Obstetrics and Gynecology, Shaoxing 2nd Hospital, Shaoxing 3120 0 0, China

Keywords:Hepatectomy Liver transplantation microRNA Long non-coding RNA Ischemia-reperfusion injury Rejection

A B S T R A C T Background: Hepatectomy and liver transplantation (LT) are the two most commonly performed surgical procedures for various hepatic lesions. microRNA (miRNA) and long non-coding RNA (lncRNA) have been gradually unveiled their roles as either biomarkers for early diagnosis or potentially therapeutic tools to manipulate gene expression in many disease entities. This review aimed to discuss the effects of miRNA or lncRNA in the hepatectomy and LT fields.Data sources: We did a literature search from 1990 through January 2018 to summarize the currently available evidence with respect to the effects of miRNA and lncRNA in liver regeneration after partial hepatectomy, as well as their involvement in several key issues related to LT, including ischemia-reperfusion injury, allograft rejection, tolerance, recurrence of original hepatic malignancies, etc.Results: Certain miRNAs and lncRNAs are actively involved in the regulation of various aspects of liver resection and transplantation. During the process of liver regeneration after hepatectomy, the expression of miRNAs and lncRNAs shows dynamic changes.Conclusions: It is now clear that miRNAs and lncRNAs orchestrate in various aspects of the pathophysiological process of LT and hepatectomy. Better understanding of the underlying mechanism and future clinical trials may strengthen their positions as either biomarkers or potential therapeutic targets in the management of complications after liver surgery.

Introduction

Hepatectomy and liver transplantation (LT) are the two most commonly performed surgical procedures for a variety of hepatic lesions. The most important indications for hepatectomy are benign and malignant tumors, while LT has gradually gained popularity as the optimal treatment for end-stage liver diseases (ESLDs),fulminant hepatic failure and early-stage hepatocellular carcinoma(HCC) over the past several decades.

Hepatectomy triggers the release of cytokines, chemokines,complements and the proliferation of quiescent hepatocytes. Hepatocyte proliferation leads to the process of liver regeneration (LR),which is a continuous pathophysiological process and regulated by various cytokines and growth factors. On the other hand, despite that great progress has been made in the field of LT by means of better surgical techniques and more effective immunosuppression,ischemia-reperfusion injury (IRI) and acute rejection (AR) [1] remain major obstacles to achieve better outcomes. The development of these complications not only compromise the short- and longterm patient and graft survivals, but also substantially increase the socioeconomic burdens. Given the scarcity of donor organs and the ever-increasing waiting list, it is mandatory to achieve maximal clinical outcome for both the grafts and patients. Traditionally, histological examination plays a critical role in the diagnosis and management of liver allograft rejection. However, liver biopsy is an invasive procedure and can sometimes result in complications like massive hemorrhage, biliary fistula or even patient death.It also correlates with limited diagnostic accuracy caused by inevitable sampling error and interpretation bias [2] . In addition,biopsy is usually performed when liver graft function has deteriorated and graft damage has long been present, which may prevent prompt anti-rejection therapy [3] . Therefore, much attention has been attracted to develop sensitive biomarkers for early detection of graft rejection. IRI is an inevitable consequence of any transplant procedures. Despite tremendous research activities to understand the pathophysiological mechanism of IRI, there is still unmet need to find novel non-invasive diagnostic tools and develop new therapeutic strategies to tackle IRI and prevent early postoperative graft dysfunction [4] . Disease recurrence, especially recurrence of HCC, is a frequent issue after LT, even in early-stage HCC within the Milan criteria. Besides various imaging studies and serumα-fetoprotein (AFP) assessment, few other measurements can be assisted in the prediction or early diagnosis of tumor recurrence. In addition, there is almost no way to prevent the recurrence of hepatic malignancies after LT [5] . Finally, clinical LT tolerance is the “holy grail”of transplantation medicine. The unique immunological properties of the liver make it an ideal organ to study clinical transplant tolerance. As we have proposed earlier, novel biomarkers as well as therapeutic strategies need to be developed to make clinical LT tolerance a constant success [6] . microRNAs(miRNAs) are endogenous 18-25 nucleotide single chain small noncoding RNAs, which bind to the complementary sequence of target protein-coding RNA and function in the regulation of gene expression through inducing messenger RNA (mRNA) degradation or translational repression [7] . Approximately 60% of genes in the human genome are known to be regulated by miRNAs. Deciphering the role of individual miRNA in the epigenetic regulation of gene expression not only facilitates better understanding of the molecular mechanism of various biological processes, but also provides novel biomarkers and potential therapeutic targets in the diagnosis and treatment of many disease entities like cancer, inflammation, transplantation, etc. [8] . In addition, recent evidence has indicated that miRNAs are involved in regulating hepatocyte proliferation during LR [9] . In the setting of LT, miRNA biomarkers can be exploited to evaluate the immune status of transplant recipients [3] . Farid et al. first revealed that the serum levels of miR-122 and miR-148a were markedly elevated in patients with AR after LT [10] . Since then, RNAs have emerged as potential candidates for noninvasive early detection of allograft rejection and other complications post-transplantation due to their remarkable stability in the plasma or serum and easy detection with appropriate techniques [11] .

Long non-coding RNAs (lncRNAs) represent another class of noncoding transcripts with longer than 200 nucleotides [12] . Although not studied as extensively as miRNAs, we now know that lncRNAs are actively involved in gene transcription, posttranscriptional regulation and many other cellular processes. They also participate in many pathological settings such as cancer,Alzheimer’s disease, psoriasis and cardiovascular diseases [13] .lncRNAs have been reported to modulate hepatocyte proliferation after hepatectomy [14] . Moreover, several groups have attempted to identify the differential lncRNA expression in hepatic grafts in order to provide novel biomarkers and therapeutic targets in the transplant setting [13 , 15] .

Given the relative scarcity of data regarding the role of noncoding RNAs in liver surgery, we herein attempt to summarize the up-to-date knowledge regarding the diagnostic and therapeutic values of miRNAs and lncRNAs in liver resection and transplantation, with special focus on LR, transplant immunology (rejection,tolerance), IRI, etc.

miRNAs and lncRNAs in LR

LR is a tightly-regulated natural process after hepatic resection and partial LT (such as living-donor, split, reduced-size and auxiliary LTs). Inadequate regeneration can lead to untoward outcomes like small-for-size syndrome, liver failure and patient death. There remains a lack of reliable biomarkers to accurately predict the regenerative capacity of the remnant liver mass or partial hepatic allograft. miRNAs have been shown to be involved in cell cycle regulation and differentiation. Therefore, understanding the regulatory mechanism of non-coding RNAs in the fine-tune of hepatocyte proliferation and regeneration may help to develop novel therapeutic means to facilitate LR.

Recent studies indicate coordinated changes can be found in miRNA expression that drive hepatocyte proliferation, innate immunity and angiogenesis during the regenerative process, while failed regeneration was associated with a distinct panel of miRNAs enforcing cell cycle inhibition and DNA methylation. Using a well-established partial hepatectomy-induced LR model in either rat or mouse, a panel of miRNAs were identified to be related to the priming phase of LR, and the potential target genes included those coding for cell apoptosis, survival, cell cycle, inflammation,metabolism, etc. [16 , 17] . The change of miRNA profile (up or down)was most significant during the peak of DNA replication at 24 h after hepatic resection [18] . Genome-wide microarray studies revealed that around 40% of the miRNAs tested were up-regulated at earlier stage after rat partial hepatectomy, while up to 70% of miRNAs were down-regulated at 24 h post-hepatectomy [19] . A similar study found that early major rearrangements of the noncoding transcriptome (including expression of miRNAs and lncRNAs) preceded most changes of the coding genes during LR [20] . Although the detailed mechanism needs to be further explored, these data suggested selective up-regulation of miRNAs in the early phase after hepatectomy can be involved in the priming and commitment to LR, whereas the subsequent down-regulation of the majority of miRNAs is required for restoration of the liver mass.

The molecular mechanisms of certain miRNAs to regulate LR have also been studied in details ( Fig. 1 and Table 1 ). The most extensively studied miRNA in LR is miR-21, whose transcription can be boosted by multiple factors like nuclear factor-kappa B (NF-κB), activation protein-1 (AP-1), nuclear factor 1 B-type(NF-1B), and signal transducer and activator of transcription 3(STAT3) [21] , as well as ursodeoxycholic acid (UDCA) [22] . miR-21 can either positively or negatively control hepatocyte proliferation and LR through modulating the expression of multiple target genes. Over-expression of miR-21 inhibited NF-κB signaling by targeting Pellino-1 (Peli1), thus negatively regulating the proliferative phase of LR [21] . On the other hand, miR-21 promoted hepatocyte proliferation via inhibition of ras homolog gene family member B (Rhob) [23] or PTEN [24] expressions, thus relieving the Akt/mTOR complex 1 (mTORC1) signaling and eIF-4Fmediated translation of cyclin D1. miR-21 might also promote hepatocyte proliferation by reducing the Fas ligand expression [25] .The prototype miRNA let-7 also seemed to play biphasic roles in regulating LR. Let-7 deletion enhanced while modest let-7 over-expression blocked LR. However, chronically high let-7 overexpression paradoxically caused liver damage, degeneration and tumorigenesis [26] . Besides miR-21 and let-7, the liver regenerative process may be orchestrated by other miRNAs. miR-382 promoted hepatocyte proliferation and cell growth via inhibiting PTEN and thus enhancing the Akt/mTOR pathway [27] . miR-221 also showed pro-proliferative property by targeting p27, p57 and aryl hydrocarbon nuclear translocator (ARNT) [28] . The ubiquitin editing enzyme A20 enhanced hepatic IL-6/STAT3 pro-proliferative signals by down-regulating suppressors of cytokine signaling 3 (SOCS3)in a miR-203-dependent manner [29] . miR-26a repressed the expression of cell cycle proteins CCND2 and CCNE2, thus worked as a negative controller of the proliferative phase of the remnant liver [30] . Hypoxia and hypoxia-induced factor-1 alpha (HIF-1α) inhibited expression of miR-150, therefore led to increased vascular endothelial growth factor (VEGF) expression which stimulated proliferation of both sinusoidal endothelial cells and hepatocytes during LR [31] . In addition to these mouse studies, one recent report using a partial auxiliary liver allograft in human similarly identified coordinated expression of a distinct miRNA profile, which regulated a series of cardinal genes governing the entry of hepatocyte into proliferation. Inhibition of miR-150, miR-663 and miR-503 could be exploited as potentially therapeutic approaches to promote LR after partial LT [32] . Finally, to explore the possibility of using miRNA as a potential biomarker to assess hepatic regeneration, Yan et al. detected enriched serum extracellular miRNA as biomarkers for ongoing LR in mice. In particular, they found miR-1A and miR-181 were most significantly altered miRNAs in both serum and hepatic tissues [33] .

Fig. 1. The molecular mechanisms of certain miRNAs and lncRNA to regulate liver regeneration. UDCA: ursodeoxycholic acid; NF- κB: nuclear factor-kappa B; NF-1B: nuclear factor 1 B-type; STAT3: signal transducer and activator of transcription 3; Peli1:Pellino-1; HIF-1 α: hypoxia-induced factor-1 alpha; AP-1: activation protein-1; VEGF: vascular endothelial growth factor; HGF: hepatocyte growth factor; Rhob: ras homolog gene family member B; mTORC1: mTOR complex 1; SOCS3: suppressors of cytokine signaling 3; STAT3: signal transducer and activator of transcription 3; ARNT: aryl hydrocarbon nuclear translocator.

Recent studies also revealed an indispensable role of lncRNA in hepatic regeneration ( Fig. 1 and Table 1 ). Xu et al. reported differential expression of a subset of lncRNAs during mouse LR using genome-wide lncRNA microarray assay. lncRNA-LALR1 was overexpressed during LR. Functional studies revealed it potently accelerated mouse hepatocyte proliferation and cell cycle progression bothinvitroandinvivo. Mechanistically, lncRNA-LALR1 inhibited Axin1 expression mainly by recruiting CTCF to the Axin1 promoter region, thus activated the Wnt/β-catenin signaling and facilitated cyclin D1 expression [34] . Using a similar strategy,more than 400 lncRNAs were found to be significantly changed upon hepatectomy-induced LR. Among them, lncRNA induced by PHx 2 (LncPHx2) was highly upregulated. Depletion of LncPHx2 using antisense oligonucleotides led to increased hepatocyte proliferation and faster regeneration, suggesting that LncPHx2 is a key negative controller of LR [14] . Future studies are warranted to further establish the regulatory role of lncRNAs in LR after resection or partial LT as either biomarkers or therapeutic targets.

Table 1 miRNAs and lncRNAs in liver regeneration.

miRNAs and lncRNAs in IRI after LT

Liver IRI, caused by reperfusion of hepatic allograft following warm or cold ischemia, is an inevitable process and a major cause of morbidity and mortality after LT. The mechanism underlying liver IRI is complex, which includes a cascade of cellular events involving transcription and translation. The physiological and biochemical events of liver IRI lead to ischemia-induced direct cellular damage and later dysfunction mediated by activation of multiple inflammatory pathways. The activation of endothelial and Kupffer cells, as well as release of cytokines and chemokines during IRI promoted hepatocyte necrosis and apoptosis and subsequently caused organ dysfunction [35] . Liver IRI results in approximately 10% of early graft failure and increases the incidence of both acute and chronic rejections [35] . Therefore, developing potential therapeutic targets of liver IRI can help to implement prompt measures to ameliorate graft damage after LT.

Recent studies have demonstrated that miRNAs are involved in the regulation of several key biological processes in IRI ( Fig. 2 and Table 2 ). miRNAs were shown to correlate with angiogenesis and suppression of cardiomyocyte apoptosis in myocardial ischemia [36] , and to protect from IRI in other cells as well [37 , 38] .Meanwhile, miRNAs have been found to exert important effects on liver pathophysiology. Analysis of 34 liver allograft biopsies revealed that many miRNAs which participated in anti-apoptosis,cellular proliferation, and pro-inflammatory processes were differentially expressed at post-reperfusion in comparison to the pre-reperfusion period [10 , 39] . Overexpression of miR-17 upregulated autophagy to aggravate liver IRI by suppressing STAT3 expression [40] . Yu et al. revealed a three-fold increase of hepatic miR-223 expression upon IRI when compared to sham controls,and the level was positively correlated with the severity of ischemic injury [41] . Three predicted downstream targets of miR-233, acyl-CoA synthesase long-chain family member 3 (ACSL3),ephrin A1 (EFNA1), and Rhob were down-regulated in liver samples with IRI. The hepatic expression of miR-146a was significantly decreased during the first 24 h after reperfusion and the nadir was at 6 h after reperfusion. Its expression pattern was reversely correlated with the expression of tumor necrosis factor receptor-associated factor 6 (TRAF6) and interleukin-1 receptorassociated kinase 1 (IRAK1), two predicted targets of miR-146a which orchestrated in the Toll-like receptor (TLR)/IL-1βsignaling pathway [42] . Overexpression of exogenous miR-146a not only directly reduced the expression of TRAF6 and IRAK1, but also decreased proinflammatory cytokine production via attenuating NFκB activation in macrophages. In addition,invivoadministration of Ago-miR-146a, a stable version of miR-146a, remarkably reduced hepatocyte apoptosis and ameliorated liver IRI via negative regulation of the TLR signaling pathway [43] . Likewise, a recent study showed that inhibition of miR-34a by carbon monoxideactivated sirtuin 1 deacetylase led to deacetylation of p53 and reduced apoptosis in mouse liver IRI model, indicating that miR-34a/sirtuin 1 pathway may represent a promising approach for the management of hepatic injury [44] . Furthermore, miR-155 deficiency protected mice from liver IRI via upregulation of suppressors of cytokine signaling 1 (SOCS1), which facilitated M2 macrophage deviation and suppression of IL-17 expression [45] .In line with this finding, miR-155 inhibition in Kupffer cells prolonged liver allograft survival through indirectly reducing T cell proliferation and boosting T cell apoptosis, suggesting that miR-155 balances Th1/Th2 cytokines and regulates the maturation and function of Kupffer cells in mice [46] . These data strongly support the notion that many miRNAs play crucial roles in the process of liver IRI and can become potential therapeutic targets to alleviate graft dysfunction. Ischemic-type biliary lesions (ITBLs) is a serious consequence of IRI that may result in graft failure and necessitate re-transplantation. An interesting study found that release of cholangiocyte-derived miR-30e, miR-222, and miR-296 during graft preservation could predict the development of ITBLs after LT [47] .

There are only a handful of studies to evaluate the potential role of lncRNAs in liver IRI ( Fig. 2 and Table 2 ). Expression of lncRNA AK139328 was significantly upregulated in mouse liver IRI. Targeting AK139328 using small interference RNA not only reduced plasma aminotransferase activity and liver necrosis, but also inhibited macrophage infiltration and NF-κB activation in mouse livers, suggesting that silencing AK139328 ameliorate liver IRI [15] .lncRNA TUG1 developed protective effects via the suppression of apoptosis in cold-induced liver injury in mice [48] . These preliminary findings pave the road for lncRNA research in LT.

Fig. 2. The potential roles of miRNAs and lncRNAs in liver IRI. ACSL3: acyl-CoA synthesase long-chain family member 3; EFNA1: ephrin A1; TLR: Toll-like receptor; TRAF6:tumor necrosis factor receptor-associated factor 6; IRAK1: interleukin-1 receptor-associated kinase 1; ITBL; Ischemic-type biliary lesion; CO: carbon monoxide; SIRT1: sirtuin 1; SOCS1: suppressors of cytokine signaling 1; IL-10: interleukin-10; TNF- α: tumor necrosis factor- α; MHC-II: major histocompatibility complex-II.

Table 2 miRNAs and lncRNAs in IRI after LT.

In summary, many hepatic graft-specific miRNAs and lncRNAs exert regulatory roles in IRI after liver resection or transplantation and represent promising biomarkers and/or therapeutic targets to protect liver IRI. The gradually increased clinical use of machine perfusion may provide unique opportunity to exploit RNAbased strategies in the quality improvement of marginal donor livers.

miRNAs in AR after LT

Fig. 3. The potential role of miRNAs in pathogenesis of liver allograft rejection and tolerance.

Despite that the liver is a well-recognized immuno-privileged organ with potent regenerative capacity after injury, AR after LT remains a true clinical challenge. AR is mediated by host-derived T lymphocytes which triggers a cascade of immune response and damage of the liver allograft [49] . Without prompt management,it can quickly cause graft loss and threaten patient life. Therefore,early recognition of AR is critical. miRNA has recently been shown to be an important player in the development, differentiation and modulation of the immune cell repertoire. They are intimately involved in T cell [50-52] and dendritic cells (DCs) [53 , 54] development, differentiation, maturation, proliferation, effector function and death. miRNAs can also regulate the allogeneic immune response after LT. Traditionally, the diagnosis of AR after LT depends on liver biopsy and histological findings. Due to the limitations of the biopsy procedure, various noninvasive biomarkers with high sensitivity and specificity have been investigated to diagnose AR after LT [55] . Among them, miRNAs seem to be suitable biomarkers for clinical application ( Fig. 3 and Table 3 ). In a preclinical LT model using allogeneic grafts from Dark Agouti (DA, RT1a) to Lewis (RT1 l ) rats, miRNA array study of the liver allograft at the peak of AR showed significantly increased expression of 12 miRNAs in comparison to syngeneic controls, such as miR20b-3p, miR-363, miR-204, miR-142-3p, miR-500 and miR-222, suggesting this panel of miRNAs play potential roles in the pathogenesis of liver allograft rejection [56] . Further study found increased plasma concentrations of miR-122, miR-192 and miR-146a in AR rats, however, only miR-146a was increased in the liver allografts, indicating that plasma miR-122 and miR-192 can reflect the magnitude of liver injury, whereas miR-146a may be involved in the initiation of AR [57] . In LT recipients, the serum levels of miR-122 and miR-148a were significantly elevated prior to the increase of aminotransferase levels during the AR episodes, with concomitant decreased expression of these 2 miRNAs in the hepatic tissue [10] . Moreover, given the stability of these 2 miRNAs in serum, they can become ideal surrogate markers for AR after human LT. Over-expression of hepatic miR-301a was found in the lethal AR livers through inducing IL-6 synthesis directly in primary hepatocytes [58] . During the process of minimization of recipient immunosuppression, hsa-miR-483-3p and hsamiR-885-5p in the recipient serum could predict and diagnose liver AR prior to the manifestations of clinical allograft dysfunction [59] . Furthermore, through analyzing intragraft miRNA expression, AR could be distinguished from recurrent hepatitis C virus (HCV) in patients who underwent LT for HCVrelated cirrhosis [60] . Notably, some study failed to identify a distinct serum miRNA signature and was unable to prove the value of miRNA in the noninvasive diagnosis of hepatic allograft rejection [61] .

miRNAs and LT tolerance

Tolerance has been widely recognized as the “holy grail”of organ or tissue transplantation, which represents a state of donorspecific nonresponsiveness in the absence of maintenance immunosuppression. Although tolerance can be induced using various means in rodent animals, it can hardly be achieved in hu-mans and the majority of non-human primates. The liver has long been considered an immune-privileged organ, which can be attributed to the unique anatomy, the “tolerant”phenotype of resident antigen-presenting cells (DC, NK and NKT, etc.), anergy or apoptosis of graft-infiltrating lymphocytes, etc. [6] . Nevertheless,AR still happens in the clinical setting. Dissecting the molecular mechanism underlying “l(fā)iver tolerance”may help to develop new strategies in the induction and maintenance of clinical transplant tolerance.

Fig. 4. miRNAs and lncRNAs in HCC and HCV recurrence after LT. TNF- α: tumor necrosis factor- α; MSC: mesenchymal stem cell; HOXB8: homeobox B8.

The regulatory roles of miRNA or lncRNA in liver allograft tolerance have just come into view ( Fig. 3 ). Vitalone et al. achieved stable allograft tolerance in a fully MHC-mismatched rat LT model by total lymphoid irradiation (TLI) post-transplantation. They found that in the early stage of tolerance induction (D7 after LT), liver allograft showed altered expression of 33 miRNAs, of which 29 were increased when compared with syngeneic controls, including miR-142-5p, miR-184, miR-7a, miR-150, miR-181a, miR-147, etc. While in the established stage of tolerance (D100 after LT), although 42differentially expressed miRNAs were found in the stable liver allografts in comparison to normal donor livers, the miRNA profile in the tolerant graft was resembling that of syngeneic controls, consistent with a state of immune nonresponsiveness. Notably, miR-142-5p and miR-181a were associated with both the induction and maintenance of hepatic allograft tolerance, and graft infiltrating lymphocytes were most likely to be the source of these two miRNAs [56] .

Table 4 Effects of miRNAs and lncRNA in HCC patients with LT.

miRNAs and lncRNAs in HCC and HCV recurrence after LT

To date, HCC is the 6th most prevalent malignant tumor and the 3rd most common cause of cancer-related death worldwide [62] .Liver resection and transplantation are regarded as the only potentially curative treatment for early-stage HCC. Nowadays, transplanting HCC patients within the Milan criteria has resulted in reasonably high survival and relatively low recurrence rates and has become the widely accepted standard for LT in HCC patients [63] .Nevertheless, tumor recurrence is not an uncommon issue after LT,which can be partly attributed to the metastatic nature of HCC as well as the immunosuppressive state of the recipients. Therefore,prevention and early detection of tumor recurrence is crucial.

Various biomarkers have been investigated to predict HCC recurrence in LT patients. Among them, miRNAs have been found to play essential roles in HCC patients with LT ( Fig. 4 and Table 4 ).Analysis of tumor samples from HCC patients after LT showed that down-regulations of five miRNAs (miR-122_st, miR-126_st,miR-15a_st, miR-22_st and miR-30a_st) were associated with HCC recurrence [64] . By examining the miRNA expression profile in paired tumor and non-tumor liver tissues in HCC patients, tumorderived miRNAs were found to better predict early than late recurrence, and vice versa for non-tumor-derived miRNAs [65] . In particular, miR-96 in non-tumorous tissues had the most intimate relationship with HCC recurrence. Another study revealed the expression of miR-718 in the serum exosomes of HCC patients with tumor recurrence after LT was significantly decreased compared with those without recurrence [66] . The reduction of miR-718 led to the overexpression of homeobox B8 (HOXB8)which was involved in the progression and recurrence of HCC.In addition, upregulation of miR-148a and miR-1246 in the early phase after LT was correlated with tumor recurrence and dismal prognosis [67] . Using miRNA analysis of the formalin-fixed paraffin embedded samples, 67 miRNAs were found to effectively distinguish patients with HCC recurrence from those without recurrence after LT [68] . Notably, high expression of miR-155 and miR-18a and low expression of miR-20a, miR-744 and miR-199a-5p were associated with HCC recurrence and worse survival [69 -72] . miR-18a might have a pro-proliferative effect on HCC cells. Another study revealed that liver cancer-associated mesenchymal stem cells(MSCs) promoted HCC metastasis in an LT model through the S100A4-miR-155-SOCS1-STAT3-matrix metalloproteinase 9 (MMP9)axis [73] . Furthermore, a high predictive score for disease-free survival after LT for HCC within the Milan criteria was developed using the combination of special miRNA expression profile (miR-140,miR-214, miR-455 and miR-3187) [74] . The upregulation of miR-424 expression significantly reduced the migration, invasion and proliferation of HCC cells in patients with HCC following LT [75] .

Interestingly, a recent study found that besides the influence of recipient-derived miRNAs, certain donor miRNA might affect the transplant outcome as well. The tumor recurrence rate was much higher in patients with the donor rs11614913 homozygous CC variant type when compared with those with donor rs11614913 homozygous TT wild type [76] .

Little attention has been paid to the effects of lncRNAs on liver cancer recurrence after transplantation. Examination of the HCC samples from diseased liver explants identified increased expression of lncRNA HOTAIR in the tumor compared to adjacent non-tumorous tissues [77] . High expression of HOTAIR was related to poorer recurrence-free survival, and siRNA-mediated suppression of HOTAIR reduced HCC cell invasion. These findings highlighted that HOTAIR could predict tumor recurrence in HCC patients after LT and might also be used as a potential therapeutic target.

Conclusions and future directions

In this concise review, we summarize the recent progress with regard to the roles of miRNA and lncRNA in liver surgery. Based on the currently available data, there is no doubt that certain miRNAs and lncRNAs are actively involved in the regulation of various aspects of liver resection and transplantation. During LR after hepatectomy, the expression of miRNAs shows dynamic changes,some miRNAs decrease while others increase [9] . However, the detailed mechanism of how individual non-coding RNA regulates this highly coordinated and complicated LR process remains largely unknown. Further studies may help to identify a core panel of noncoding RNAs which can be exploited either in the monitoring of the pace or capability of LR, or as therapeutic targets to accelerate the regeneration process. miRNAs and IncRNAs also affect various aspects of LT, including the immunological response against allogeneic hepatic graft, the inevitable IRI which leads to organ damage and functional impairment, recurrence of original diseases including hepatic neoplasms and hepatotropic viruses, etc. However,there are apparent caveats of the published studies. The majority of them only discover associations between non-coding RNA and the pathological state but without any mechanistic studies to understand the functional role of these non-coding RNAs. Most of the clinical studies only contain small patient numbers so the results need to be validated in large cohort of LT recipients. Nevertheless, these data corroborate that non-coding RNAs play irreplaceable roles in the pathogenesis of various complications after LT.

Despite greatly improved outcome of LT in the past 30 years,short and long-term post-transplant complications remain obstacles to achieve better patient and graft survival. There is an urgent need to search for biomarkers with satisfactory sensitivity and specificity to herald the development of serious complications like rejection, chronic allograft dysfunction, etc., so that we can initiate the salvage therapy as early as possible before the condition is getting worse. The development of these complications may be associated with change of a panel of genes, whose expression can be controlled or fine-tuned by several miRNAs or lncRNAs. Therefore,it is critical to find the “signature”non-coding RNA biomarkers to alarm the development of serious complications after LT. Another potential application of non-coding RNA is the surveillance of clinical transplant tolerance. Staged weaning of immunosuppression is currently an integral step to achieve the “near-tolerance”state after clinical LT [6] . Finding reliable miRNA or lncRNA markers can be an ideal strategy to better guide the drug-weaning process and avoid untoward rejection or graft loss. Recent studies have also shown the miRNAs can be released from cells until they are taken up by another cell population either locally or in remote part of the body,thus play a role in cell-cell communication [78] . These circulating miRNAs incorporated into the exosomes are believed to involve in cell signal transduction, while those conjugated with Argonaute-2 are considered to be cell by-product [79] . It is now believed that vesicle miRNAs can be used to assess many pathological entities.Future studies will systemically evaluate the peripheral circulating miRNA profile at different time points after transplantation, therefore set up the basis for exploiting these miRNAs and lncRNAs as important “l(fā)iquid biopsy”technique to early detect untoward complications after LT.

Emerging evidence also indicates that certain miRNAs orchestrate the immunological reaction against allogeneic antigens.However, our understanding of the role of non-coding RNAs in transplant immunology is still in its infancy. Before we study their functional role individually, it is mandatory to first have an integral profile of the miRNAs or lncRNAs during allogeneic immune response. Given the complexity of the allogeneic immune response,it is unlikely to inhibit rejection or induce tolerance by targeting any individual gene. The ability of miRNAs to regulate important cellular process by simultaneously targeting multiple mRNA targets illustrates their potential as a viable therapeutic tool. lncRNAs can even bind to and exert effects on multiple miRNAs. Based on the miRNA profile in allogeneic immune response, we may use replacement therapy like miRNA mimetics to reinforce those miRNAs capable of inhibiting untoward immune response and facilitate the development of transplant tolerance. By contrast, many miRNAs that cause undesirable consequences after transplantation need to be inhibited by way of miRNA sponges, miRNA masks, antagomirs,synthetic anti-sense anti-miRNA oligonucleotide or locked nuclear acid anti-miRNAs, etc. [80] . The recent discovery of small molecule inhibitors of miRNA (SMIRs) represents a very promising means to inhibit the function of miRNAinvivo[81] . Besides directly targeting the function of miRNAs, we may also target miRNA-containing exosome to interrupt the connection between immune cells [82] .However, there is a relative paucity of data regarding the “pharmacokinetics”and safety profile of these RNA agentsinvivo. Therefore, a huge amount of work has to be done to design more efficient delivery system, to improve the precision and efficacy by minimizing the “off-target”effects, and to improve the biostability and “half-life”of these non-coding RNAsinvivo. Furthermore, in order to minimize the potential “toxic”side effects of these RNAbased therapeutics, clinical trials need to be initiated to evaluate the safety and effectiveness of combination therapies using noncoding RNAs in association with chemotherapy or targeted therapy.

In summary, it is now clear that non-coding RNAs such as miRNAs and lncRNAs play critical roles in various aspects of liver surgery. In-depth understanding of their functional significance hopefully will bring novel diagnostic and/or management modes to the clinical care of patients who require liver resection or transplantation in the future.

CRediT authorship contribution statement

Zhi-Qiu Hu:Data curation, Funding acquisition, Writing - original draft.Yi Lu:Data curation, Writing - original draft.Di Cui:Formal analysis.Chen-Yang Ma:Software, Visualization.Su Shao:Investigation.Ping Chen:Data curation.Ran Tao:Supervision, Funding acquisition, Writing - review & editing.Jian-Jun Wang:Conceptualization, Supervision, Writing - review & editing.

Funding

This work was supported by grants from National Health Commission Scientific Research Fund-Major Project of Zhejiang Medical and Health Science and Technology Plan (WKJ-ZJ-1901),National Natural Science Foundation of China ( 81001324 and 8137316 3), Natural Science Foundation of Zhejiang Province grant for “Outstanding Youth”( LR15H10 0 0 01 ), an interim starting Funding from ZJPPH, Commission of Science Technology of Minhang District ( 2019MHZ079 ), and Minhang Scientific Research Found projects grant (2017MHJC02).

Ethical approval

Not needed.

Competing interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.