The spleen in liver cirrhosis: revisiting an old enemy with novel targets

Spleen sizes can vary between cirrhotic patients by primary disease etiologies, with hepatitis C virus (HCV) infected and non-alcoholic hepatitis patients showing significantly larger organ dimensions compared to alcoholic hepatitis patients [18]. Histologically, chronic portal hypertension-induced splenomegaly features expanded white pulp and marginal zone areas and appears different to congestive splenomegaly, which is characterized by more prominent red pulp and less distinct white pulp regions [19, 20]. Clinically, splenomegaly has been associated with a poor prognosis in liver cirrhosis and utilized during radioactive or acoustic examinations as an index for the non-invasive assessment of esophageal varices and bleeding risks [1, 21, 22]. Splenic stiffness can also increase as splenomegaly advances [23]. Portal congestion is widely considered the initial cause of splenomegaly during liver cirrhosis [5, 24]. The subsequent changes in the enlarged spleen are complex and difficult to elucidate, considering the concurrent involvement of multiple cell populations in different compartments. Recently, Mejias et al. induced splenomegaly in rats using a partial portal vein ligation (PPVL) model of chronic portal hypertension. Interestingly, significantly increased activation of the mTOR signaling pathway was observed within the enlarged spleen. More importantly, mTOR inhibition using rapamycin profoundly ameliorated splenomegaly, causing a 44% decrease in spleen size [20]. Although the PPVL model more closely simulates human idiopathic portal hypertension (IPH), these findings remain suggestive for the study of cirrhosis-associated portal hypertension. In another study, Chen et al. utilized a rat model of portal hypertension induced by a combination of bile duct ligation (BDL) and PPVL. They reported that rapamycin-induced mTOR inhibition significantly decreased splenomegaly through the inhibition of lymphocyte proliferation, angiogenesis, fibrogenesis and tissue inflammation levels, which ultimately led to a decrease in portal pressure [25]. Consistent with Mejias et al., the findings from Chen et al. are insightful as the combination of BDL and PPVL models the augmentation of portal hypertension by biliary cirrhosis, which more closely mimics clinical cirrhosis conditions. Overall, the identification of portal hypertension-induced mTOR signaling alterations may be highly significant due to its central roles in immune cell modulation, angiogenesis and hepatic fibrogenesis [26‐28]. Further investigations utilizing animal models of liver cirrhosis-associated portal hypertension will be required to confirm whether and how the mTOR signaling pathway may contribute to liver cirrhosis-associated splenomegaly.

The incidence of hypersplenism has been reported to range from 11 to 55% in patients with cirrhosis and portal hypertension [29]. Hypersplenism often develops in parallel with splenomegaly. The mechanisms responsible for hypersplenism remain a matter of debate although recent studies iterate and corroborate that dysregulated immune cell responses may contribute to this process. In a study from Nomura et al., spleen samples from 26 patients with HCV-associated cirrhosis and hypersplenism who underwent splenectomy were studied by immunohistochemical staining and flow cytometry. They found that the splenic ratio of CD4⁺:CD8⁺ lymphocytes from these patients were higher compared to the control group (P = 0.06) whilst the ratio of FOXP3⁺:CD4⁺ was lower than the control group, implicating an increase in CD4⁺ T cell immune responses during hypersplenism [30]. A series of studies from our group have shown that splenic macrophages may also be hyperactivated and subsequently facilitate cirrhosis-associated hypersplenism [31‐34]. We demonstrated that NF-kB p65/c-Rel signaling was significantly elevated in hypersplenic macrophages and promoted increased phagocytosis and secretion of both pro-inflammatory and pro-fibrogenic factors such as IL-1β, IFN-γ, TNF-α and TGF-β1 [35]. The source of this macrophage hyperactivation, however, remains unresolved. Interestingly, a recent study by Valdes-Ferrer et al. demonstrated that a major expansion of the inflammatory CD11b⁺Ly-6C^high monocyte pool occurs in the spleens of sepsis-induced mice which survive caecal ligation and puncture (CLP) and subsequently develop significant splenomegaly. Serum high-mobility group box 1 (HMGB1) levels were significantly elevated in surviving sepsis-induced mice for 4–6 weeks, with administration of recombinant HMGB1 to naive mice inducing similar splenomegaly, leukocytosis and splenocyte priming phenotypes as sepsis-induced survivors [36]. The administration of anti-HMGB1 monoclonal antibody also significantly attenuated splenomegaly and splenocyte priming levels in sepsis-induced survivors [36]. HMGB1 belongs to the family of damage-associated molecular patterns (DAMPs) and elevated serum and liver HMGB1 and other DAMPs levels have been widely reported in liver fibrosis and cirrhosis [37‐39]. Consequently, it is possible for DAMPs such as HMGB1 to circulate and induce splenic immune cell dysregulation during these conditions.

Recent studies in the field of extracellular microvesicles have provided us with a new paradigm for understanding the crosstalk between the liver and spleen. In a study from Saunderson et al., splenic marginal zone macrophages were able to capture in vivo or injected B cell-derived exosomes through the binding of exosome surface α2,3-linked sialic acid to cell surface CD169 expressed on marginal zone macrophages. CD169^−/− mice showed an altered distribution of exosomes, with exosomes freely accessing the outer marginal zone rim of SIGN-R1⁺ macrophages and F4/80⁺ red pulp macrophages in contrast to wild type controls. Interestingly, CD169^−/− displayed enhanced CD8⁺ T cell responses to antigen-pulsed exosomes compared to wild type mice, suggesting that circulating exosomes may be selectively recognized and also captured by splenic macrophages for antigen presentation purposes [40]. Considering the extensive involvement of exosomes in the pathogenesis of liver diseases, especially liver fibrosis, a role for liver-derived exosomes in the development of hypersplenism should not be excluded [41‐43]. Further study will be required to elucidate more detailed mechanisms of how the contributions of hepatic injury, inflammation and fibrogenesis may affect splenic homeostasis and contribute to hypersplenism.

During liver fibrosis, hepatic stellate cells (HSCs) and Kupffer cells (KCs) act as the initial effectors of collagen deposition and inflammation modulation with the aid of the pro-fibrogenic cytokine transforming growth factor beta 1 (TGF-β1) [50, 51]. Previous studies have reported splenic TGF-β1 production in the context of liver cirrhosis and hypersplenism and emphasized its critical role in the development of hepatic fibrogenesis. Splenic macrophages have been suggested as one source of TGF-β1. In a study by Akahoshi et al., a rat model of liver cirrhosis was induced by intraperitoneal injection of thioacetamide (TAA) for 24 weeks, and followed by either a splenectomy or sham operation. Splenic red pulp macrophages were suggested as a major source of TGF-β1 in this study, and splenectomy was reported to decrease serum TGF-β1 levels significantly whilst improve liver fibrosis and regeneration parameters [6]. A recent study by Asanoma et al. detected increased TGF-β1 expression from the resected spleens of liver cirrhotic patients by immunofluorescence staining, and reported it to be significantly correlated with the progression of liver cirrhosis. TGF-β1 immunofluorescence further overlapped with the staining of CD68, a pan marker of tissue macrophages [52]. Overall, the studies from Akahoshi et al. and Asanoma et al. both implicate that splenic involvement during hepatic fibrogenesis may be mediated through splenic macrophage production and secretion of TGF-β1. Thus, further studies elucidating the explicit mechanisms driving splenic macrophage TGF-β1 production during liver cirrhosis in the presence or absence of hypersplenism will be of great interest.

As the largest lymphoid organ in the body, the spleen contains multiple immune cell subsets which are differentially distributed within white and red pulp and marginal zone regions. B cells are mainly distributed within splenic follicles whilst T cells predominate in the white pulp regions. By contrast, DCs and macrophages are located predominately in the marginal zone (MZ) and red pulp regions. Immune cell migration between these different regions is necessary for splenic maintenance of immune cell homeostasis, tolerance and pathogen clearance [13, 53‐55]. Interestingly, the red pulp has been revealed to maintain a reservoir of monocytes capable of rapidly migrating into injured tissues and mediating local inflammatory responses [56]. Similarly, T lymphocytes and some innate lymphoid cell (ILC) subsets have also been reported to be capable of splenic extravasation [57, 58].

As the liver and spleen are closely associated via the portal vein system, it is more likely for the spleen to exert its influences on the hepatic immune microenvironment by cell migration or the secretion of splenic soluble factors via portal vein blood flow (Fig. 2). In a recent study by Yada et al., splenectomy significantly increased the hepatic accumulation of Ly-6C^low monocytes or macrophages in a thioacetamide-induced murine model of liver cirrhosis with hypersplenism, implicating a role for the splenic control of hepatic monocyte or macrophage phenotypes [59]. Although no splenic monocyte tracing studies have been conducted in the setting of liver fibrosis or cirrhosis, the findings from Yada et al. are potentially suggestive when considering the directional movement of red pulp monocyte reservoirs towards cardiac lesions [56].

Studies focusing on T cells have also shed light in another direction. In a study of Schistosoma japonicum associated liver fibrosis by Romano et al., splenomegaly was found to be correlated with higher levels of FOXP3⁺ regulatory T cells (Tregs) in the blood and increased liver fibrosis severity. Splenectomy decreased Treg cell and hepatic fibrogenesis levels, implicating a role for the splenic modulation of the liver via alterations in T cell subsets [60]. More interestingly, in a whole genome microarray analysis conducted by Burke et al. [61] following Schistosoma japonicum infection, lymphocyte and monocyte chemokines and cell adhesion molecules were significantly upregulated in the liver whilst concurrently downregulated or unaltered in the spleen, possibly suggesting a recruitment of effector cells from the spleen to the liver. Another study by Tanabe et al. has explicitly pointed that splenic T cells may migrate into the fibrotic liver to promote hepatic fibrogenesis. They reported a decrease in splenic CD4⁺ T cell numbers following carbon tetrachloride (CCl4) or thioacetamide (TAA) induced liver injury in BALB/c mice. Splenectomy shifted the liver helper T cell (Th) Th1/Th2 balance towards Th1 and suppressed the progression of liver fibrogenesis. Adoptive transfer of GFP⁺ splenocytes into the spleens of syngeneic wild type mice following fibrosis induction led to the accumulation of GFP⁺CD4⁺ lymphocytes with a Th2 phenotype in the diseased liver. The transfer of GFP⁺ splenocytes into the portal vein of syngeneic splenectomized mice abolished the suppressive effects of splenectomy on liver fibrosis, highlighting that splenic contributions to liver fibrosis may be directly mediated by the migration of splenic immune cells into a subsequently altered hepatic immune microenvironment [57]. Sophisticated in vivo tracing studies will be needed, however, to confirm the impact of splenic immune cell migration and/or secreted soluble factors on the hepatic immune microenvironment during liver fibrosis and cirrhosis.

Although the liver normally possesses a huge regenerative capacity, this capacity is compromised during liver injury or resection, especially in the presence of severe chronic liver injury with marked fibrosis and tissue architecture aberrations [62, 63]. Liver regeneration is frequently overwhelmed by fibrogenesis in chronic liver diseases [64]. During chronic liver injury, hepatic progenitor cells (HPCs) act as the regenerative source for hepatocyte or cholangiocyte replenishment [65, 66]. Liver resident macrophages, also known as KCs, have also been shown to promote the differentiation of progenitor cells through debris engulfment-induced Wnt3a signaling in chronic liver diseases [67, 68]. Recently, residential KCs and recruited monocyte-derived macrophages have been both found to play central roles in liver regeneration, most likely through a mechanism involving KC–monocyte crosstalk and the consequent activation of HPCs [69]. No study has yet to determine, however, whether splenic monocyte-derived macrophages can directly influence KCs or HPCs to alter liver regeneration capacities during chronic liver injury.

Several studies point towards the potential of splenic contributors to impact liver regeneration capacities. Studies from Murata et al. and Yamada et al. have demonstrated that splenectomy improved liver regeneration in cirrhotic animals and patients respectively [70, 71]. In a study by Ueda et al. of rats undergoing major hepatectomy with or without splenectomy, early stage splenic red pulp TGF-β1 production and secretion into the portal blood was suggested to exert an inhibitory effect on liver regeneration. Splenectomy reversed this inhibition and enhanced the regeneration of hepatocytes [72]. In a separate study by Lee et al., which used rats undergoing a 70% hepatectomy with or without splenectomy, decreased TGF-β1 and increased hepatocyte growth factor (HGF) levels in the portal vein occurred concurrently to increase liver regeneration following splenectomy. Liver TGF-β1 and HGF receptor levels were also correspondingly regulated by splenectomy in this study [73]. Altogether, these studies suggest a role for the spleen in the inhibition of liver regeneration via TGF-β1 upregulation and HGF downregulation and its subsequent disruption of the hepatic pro-versus anti-regenerative signaling balance. However, more detailed mechanistic pathways remain to be resolved.

Recent studies using stem cell therapy in chronic liver diseases have also provided some clues regarding the role of the spleen in hepatic regeneration. In a study by Iwamoto et al., splenectomy enhanced the repopulation of adoptively transferred bone marrow cells in cirrhotic liver and decreased collagen deposition through the upregulation of MMP9 expression in transferred bone marrow cells [74]. In a separate study of liver cirrhosis in rats, Tang et al. reported that splenectomy improved the efficiency of adipose tissue-derived mesenchymal stem cell transplantation into the liver by enhancing liver SCF-1 (stromal cell-derived factor-1) and HGF expressions [75]. Further studies, however, will be required to characterize whether and how spleen-derived cells or soluble factors can mediate direct effects on liver-recipient stem cell expansion or hepatic progenitor cell behavior.

Splenectomy has been traditionally performed to treat liver cirrhosis for the alleviation of portal hypertension. The utility of splenectomy is well accepted and may be especially important for the treatment of fatal complications such as bleeding esophageal or gastric varices. Apart from its effects on portal hypertension and hypersplenism, splenectomy has also been reported by many groups to be an efficient method for improving liver function and the prognosis of esophageal varices [8, 76]. Previous studies have also reported splenectomy to increase the efficacy of liver transplantation and improve the prognosis of hepatocellular carcinoma [77‐79]. A study by Ogawa et al. has also suggested splenectomy as a supplemental treatment for anti-HCV therapy in combination with interferons and other pharmaceuticals, which is consistent with our own finding [80, 81]. Although animal studies suggest that splenectomy can ameliorate collagen deposition, improve liver function and regeneration capacity and enhance the therapeutic efficacy of stem cell transplantations [6, 59, 75], many cirrhosis patients present with contraindications which preclude splenectomy. Thus, it remains critical for us to identify alternative novel and non-surgical methods which can target the spleen for the treatment of liver cirrhosis.

Previous studies suggest that there is potential therapeutic value in targeting splenic mTOR signaling for the amelioration of splenomegaly and hypersplenism. Inhibition of splenic macrophage TGF-β1 production may be another therapeutic target in the treatment of liver cirrhosis progression. Therapies selectively modulating splenic macrophage or T cell activation may also be useful as these immune cell aberrations have been linked to hepatic immune microenvironment dysregulation and the enhancement of hepatic fibrogenesis. One of the biggest obstacles to these approaches has traditionally been the absence of spleen-selective delivery options for therapeutic agents. This problem, however, may be circumvented by the use of nanoparticles for novel and targeted drug delivery. In a study of nanoparticle distribution during LPS-induced systemic inflammation in mice, the spleen showed increased retention of carboxylated polystyrene latex bead nanoparticles of different sizes compared to all other organs studied. Nanoparticle retention localized mostly to the splenic marginal zone and red pulp to a lesser extent, with most nanoparticles found to be ingested by splenic macrophages. The uptake of nanoparticles by splenic macrophage was size dependent, with 20–100 nm particles preferentially engulfed by CD68⁺ marginal zone macrophages, 100–500 nm particles engulfed by CD169⁺ metallophilic marginal zone macrophages and 500 nm particles engulfed by F4/80⁺ red pulp macrophages. This size dependence may be caused by the increased permeability of splenic vessels during inflammatory conditions, and highlights one interesting prospect for the selective targeting of differential splenic regions [82].

An excellent review of the principles of nanoparticle design can be found elsewhere [83]. In a recent study by Hu et al., poly(lactic-co-glycolic acid) nanoparticles were fashioned with membranes (containing both lipid and protein components) isolated from red blood cells in order to prolong nanoparticle retention time in the circulation. This approach may also be applicable for spleen selective drug targeting as splenic macrophages actively phagocytose red blood cells during hypersplenism and liver cirrhosis [84]. Research into the targeting of splenic and liver Leishmaniasis infections using nanoparticles may also be of potential strategic use [85, 86]. Standardized studies in animal models and pre-clinical trials measuring the efficacies of nanoparticle-based therapeutic agents will also be required for the development of novel spleen-selective therapeutic agents. In addition to nanoparticles, recombinant exosomes may also be useful for splenic targeting given the potentially selective capture of exosomes by splenic macrophages [40, 87, 88].

In addition to nanoparticle-based therapeutic agents, the crosstalk between the spleen and nervous system should also be considered in order to develop a comprehensive spleen-modulating approach for treating liver cirrhosis. A dense network of sympathetic noradrenergic fibers are closely associated with the splenic artery and branch into the white pulp and consequently T and B cell regions. Secreted neuropeptides may also directly affect the behaviour of splenic immune cells, and this may also need to be factored when designing spleen modulating therapies [89, 90]. Furthermore, the inhibition of liver injury-induced signaling molecules that perturb spleen homeostasis, such as HMGB1 or other DAMPs, may also be beneficial.