We previously demonstrated that G-MDSCs were enriched within regenerating livers following major Hx in mice (24). Notably, G-MDSC depletion led to a significantly deregulated liver regeneration process, accompanied by a marked increase in post-operative mortality. Among the surviving mice, we observed a substantial decrease in regenerative liver weights and elevated serum phosphate levels, attributed to decreased cellular uptake due to aberrant hepatocyte proliferation (24). Moreover, hepatocytes from MDSC-depleted livers exhibited decreased proliferative capacities on postoperative day 2 (POD2) after resection, along with increased levels of cell necrosis (24). These findings collectively indicate that the role of G-MDSCs in liver regeneration is most pronounced during the initial 48-hour period following liver resection (24).
Genome-wide transcriptional profiling of Reg-G-MDSCs in comparison to G-MDSCs isolated from sham-operated mice revealed that Reg-G-MDSCs exhibited the highest transcriptional activity on postoperative day 1 (POD1) and POD2, featuring thousands of differentially expressed genes in contrast to controls (Fig. 1a). While our previous report focused primarily on the proangiogenic functions of MDSCs, our current transcriptional analysis of Reg-G-MDSCs compared to MDSCs from non-regenerating livers revealed increased expression of key genes associated with the immune suppressive functions of MDSCs. These genes included S100a8, S100a9, Cxcl1/2/3, Ccl2/3, Arg1, Nos2, Il10, Il10rα, and more (Fig. 1b) (24). Notably, gene clusters related to immune suppression, Il6 and Il10 pathways, as well as Toll-like receptor (Tlr) response, displayed significant differential expression within Reg-G-MDSCs during the initial two days following major resections (Fig. 1c) (24).
In light of the data, we sought to investigate the effect of G-MDSC depletion on the gene-expression patterns of regenerating liver hepatocytes through genome-wide transcriptional profiling. We compared the global gene expression pattern of hepatocytes isolated from G-MDSC-depleted mice to those from G-MDSC-sufficient mice during the first two days following major liver resection, and observed a marked transcriptional change within regenerating hepatocytes, particularly on POD2. On POD1, G-MDSC depletion resulted in a relatively mild transcriptional response within hepatocytes, characterized by 587 differentially expressed genes (with over 2-fold change, FDR < 0.05, and a read count > 30), comprising 211 down-regulated and 376 up-regulated genes. However, on POD2, the impact of G-MDSC depletion on hepatocyte gene expression was substantial, with 3178 differentially expressed genes identified (with over 2-fold change, FDR < 0.05, and a read count > 30), consisting of 1245 down-regulated and 1933 up-regulated genes (Fig. 2A-C). In regenerating hepatocytes isolated from G-MDSC-sufficient mice, the transition from POD1 to POD2 was associated with a massive transcriptional response, encompassing 4080 genes that were either up- or down-regulated (with over 2-fold change, FDR < 0.05, and a read count > 30), comprising 2359 down-regulated and 1721 up-regulated genes (Fig. 2a-c). Upon MDSC depletion, this transcriptional response was practically abolished, with only 40 differentially expressed genes (with over 2-fold change, FDR < 0.05, and a read count > 30), consisting of 11 down-regulated and 29 up-regulated genes (Fig. 2a-c). Principal Component Analysis (PCA) further emphasized the impact of G-MDSC depletion on hepatocyte transcriptional progression from POD1 to POD2, as it suggested a clear attenuation of this progression (Fig. 2b).
Pathway analysis revealed that G-MDSC depletion had a significant impact on gene clusters that exhibited differential expression between POD2 and POD1 in MDSC-sufficient mice. Among the most significant gene clusters affected were those related to mitotic spindle organization, cell division, and proliferation, all of which showed decreased expression upon G-MDSC depletion. Additionally, innate immune response gene clusters exhibited notable changes, with a particular focus on genes associated with positive regulation of the Il-6 pathway and innate immune cell activation (Fig. 3a). This set of genes included Il-6 itself, Tnfα, Tlr1/2/3/4/6/7/8/9, Ccr5, Mmp8, Il1, Nod1, Il16, Cd36, Aif1, Nos2, and more. While the expression of these genes decreased on POD2 compared to POD1 in control livers, their expression remained significantly elevated upon G-MDSC depletion (Fig. 3a-c).
Considering the immune suppression related transcriptional response in Reg-G-MDSCs and the observed deregulation of this pattern in G-MDSC-depleted regenerative-liver derived hepatocytes, we further sought to elucidate the impact of G-MDSCs on other immune cell populations during liver regeneration. To achieve this, we employed CyTOF to compare the immune cell milieu in regenerating livers between G-MDSC-depleted and G-MDSC-sufficient conditions on both POD1 and POD2 following major Hx. CyTOF confirmed almost complete depletion of G-MDSCs (CD44, Ly6Clow, GR1hi) both on POD1 and POD2 (Fig. 4a-c). Depletion of G-MDSC was associated with overall increase in myeloid cells within the regeneration liver (Fig. 4c). Overall monocytes were enriched both on POD1 and POD2 while neutrophils (Ly6Chi, CD49b) were enriched on POD1 only. NK cells (CD44, CD49b, CD69, NKp46hi, CD62L) as well as activated T cells (CD3e, TCRb, CD44, CD69) were enriched on POD2 (Fig. 4a-c) upon G-MDSC depletion. No significant change was demonstrated in memory T cells (CD3e, TCRb, CD44, CD62L) or macrophages (CD44, PD-L1, F4-80) upon depletion of G-MDSCs (Fig. 4a-c).