Carbohydrates and ginsenosides in shenmai injection jointly improve hematopoietic function during chemotherapy-induced myelosuppression in mice

SMI attenuated 5-FU-induced myelotoxicity in tumor-bearing mice

Firstly, we tried to verify that SMI could be administrated as a supportive reagent for chemotherapy. In this regard, tumor-bearing mice accepting 5-FU were used to mimic cancer patients undergoing chemotherapy. We found that SMI improved survival of tumor-bearing mice treated with 5-FU (p < 0.05) (Fig. 1A) but had no obvious impact on the tumor size (Fig. 1B). In addition, SMI increased BMNC count (Fig. 1C) as well as the colony counts of BFU-E and CFU-GM in mice treated with 5-FU (p < 0.05) (Fig. 1D–F). The results indicated that SMI attenuated 5-FU-induced myelotoxicity without affecting the anti-tumor effect of 5-FU.

Fig. 1figure 1

Effects of SMI on 5-FU-induced myelotoxicity. A Survival curves for the three groups of mice. 5-FU and SMI treatment started on Day 0. B Representative photograph of the largest tumors of the three groups. C BMNC count in each group. Cells were isolated on Day 16. D BFU-E cell count in each group. E CFU-GM cell count in each group. BFU-E burst-forming unit-erythroid, CFU-GM colony-forming units-granulocyte-monocyte. F Representative images of colony formation in each group. The results are expressed as the means ± SEM. #p < 0.05, compared with the Untreated group; *p < 0.05, compared with the 5-FU group

Carbohydrates promoted the hematopoietic function in myelosuppressive mice

Next, we deemed to explore the effects and modes of action of SMI, ginsenosides, and carbohydrates on hematopoietic function per se without the interference from tumor signaling. In this regard, 5-FU-induced acute myelosuppressive mouse model was applied (Fig. 2A). 5-FU modeling significantly reduced the BMNC count (p < 0.01), while treatment with carbohydrates (S), ginsenosides (Rg) or their combination (S + Rg) significantly increased BMNC count (p < 0.05) (Fig. 2B). Figure 2C shows HE staining of the whole femur (1 ×), magnified field of the marrow cavities of both the proximal and distal femurs (40 ×), as well as Masson staining of the distal femur (40 ×). The HE staining revealed that in the Control group, the bone marrow structure appeared normal, all types of hematopoietic cells were evenly distributed in rich hematopoietic zones, and mature erythrocytes were present in the sinusoids. In the Model group, few nucleated cells were in the marrow cavities of both proximal and distal femurs; vascular rupture and serious hemorrhage were observed. The treatment with SMI, S + Rg, S, or Rg resulted in improved structure and morphology of the marrow cavity in the proximal rather than distal femur, along with enhanced angiogenesis, reduced hemorrhage, and increased nucleated cell count. The Masson staining showed that 5-FU modeling led to a large number of reticulin fibers and collagen, which were reduced after SMI, S + Rg, S, or Rg treatment. Notably, the effect on improving bone marrow morphology seemed better when using a combination of carbohydrates and ginsenosides than when using ginsenosides alone. We also noticed that the adipocytes count was greatly increased in the S group compared to that in the Model group (Fig. 2C). Taken together, the results showed that both carbohydrates and ginsenosides in SMI promoted hematopoietic function and improved bone marrow morphology in myelosuppressive mice.

Fig. 2figure 2

Effects of SMI, ginsenosides, and carbohydrates on hematopoiesis in mice with 5-FU-induced acute myelosuppression. A Experimental design. Specific administration is described in “Materials and Methods”. B BMNC count in each group. Cells were isolated on Day 7 (n = 3 for each group). The results are expressed as the means ± SEM. ##p < 0.01, compared with Control group; *p < 0.05, compared with Model group; ns, no significant difference compared with Model group. C Representative HE and Masson staining images of bone marrow cavities. Black arrow refers to the hematopoietic zone. Yellow arrow refers to the sinusoid. Blue arrow refers to the adipocyte. Green arrow refers to the collagen or reticulin fiber

Carbohydrates promoted the proliferation of BMSCs from acute myelosuppressive mice in vitro

We further investigated the effects of carbohydrates and ginsenosides on the growth pattern of bone marrow cells by long-term bone marrow culture. Figure 3 shows a representative microscopy photograph of the cells in each group: the adhered cells can be recognized as BMSCs and the resuspended cell clusters as hematopoietic cells. Compared to the Control group, the BMSC count in the Model group decreased significantly, and the hematopoietic cell count was very low. Massive proliferation of BMSCs and the formation of hematopoietic cell clusters were observed in the SMI, S + Rg, S, and Rg groups. The result indicated that both ginsenosides and carbohydrates promoted the proliferation of BMSCs and hematopoiesis. Moreover, SMI, and the combination of carbohydrates and ginsenosides appeared to have the greatest effect on promoting the proliferation of bone marrow cells.

Fig. 3figure 3

Effects of SMI, ginsenosides, and carbohydrates on the bone marrow cell growth in vitro. Bone marrow cells were extracted from 5-FU-treated mice for long-term bone marrow culture with or without the presence of SMI, S + Rg, S, and Rg respectively. Representative microscopy images of cell growth in each group after 4-week culture are shown. The adhered cells are considered as stromal cells and the resuspended cell clusters are hematopoietic cells

RNA-seq identified transcriptome profiles of carbohydrates and ginsenosides in BMSCs of acute myelosuppressive mice

We used 5-FU to establish an acute myelosuppression model on Day 0 followed by SMI, S + Rg, R, or Rg treatment until Day 10 (Fig. 4A). After confirming the improved hematopoietic function due to the treatment by BMNC count (Additional file 3: Fig. S3), transcriptome profiling analyses of BMSCs were performed to understand the potential mechanisms. We first analyzed DEGs in response to 5-FU-induced myelosuppression with or without treatment. Compared to that in the Control group, 2546 genes were upregulated, and 616 genes were downregulated in the Model group (Fig. 4B and Additional file 5: Table. S2, Model vs Control). Compared to that in the Model group, 776 genes were upregulated, and 1304 genes were downregulated in the S + Rg group; 515 genes were upregulated, and 999 genes were downregulated in the S group; 632 genes were upregulated, and 1543 genes downregulated in the Rg group (Fig. 4B and Additional file 5: Table. S2, S + Rg vs Model, S vs Model, Rg vs Model).

Fig. 4figure 4

RNA-seq based gene expression profiles of BMSCs obtained from 5-FU-treated mice with myelosuppression. A Experimental design. Specific administration is described in “Materials and Methods”. B Number of upregulated and downregulated genes in Model vs Control and each treatment group vs Model. C Venn diagrams of DEGs between Model vs Control and each treatment group vs Model. D Heat maps and Venn diagrams of co-regulated DEGs between Model vs Control and each treatment group vs Model. M.up genes are upregulated in Model vs Control, M.down genes are downregulated in Model vs Control, S + Rg.up genes are upregulated in S + Rg vs Model, S + Rg.down genes are downregulated in S + Rg vs Model, S.up genes are upregulated in S vs Model, S.down genes are downregulated in S vs Model, R.up genes are upregulated in Rg vs Model, R.down genes are downregulated in Rg vs Model

We then analyzed the co-regulated DEGs between each treatment group vs the Model group and the Model group vs the Control group. 1520, 1142, and 1618 DEGs were co-regulated in the S + Rg, S, and Rg groups, respectively (Fig. 4C). Among them, 1104 genes were downregulated, and 248 genes were upregulated in the S + Rg group vs Model group, 865 genes were downregulated, and 184 genes were upregulated in the S group vs Model group, and 1272 genes were downregulated, and 229 genes were upregulated in the Rg group vs Model group (Fig. 4D and Additional file 6: Table. S3). Additionally, as shown in the heatmap in Fig. 4D, the variation trends (upward or downward) appeared to vary similarly between the Model vs Control and Model vs treatment groups. Altogether, the results indicate that ginsenosides, carbohydrates, and their combination can recover the expression of most co-expressed DEGs changed by myelosuppression.

We also analyzed the co-regulated DEGs and their GO functional annotations amongst the treatment groups. As shown in the Venn diagram (Fig. 5A), 51 genes responded to S and S + Rg but not Rg treatment (Additional file 7: Table. S4), and 71 genes responded uniquely to S + Rg treatment (Additional file 8: Table. S5). The GO analysis of the 51 genes indicated 55 GO terms, with the top functional annotation clusters being B cell activation, external side of plasma membrane, regulation of lymphocyte activation, regulation of lymphocyte apoptotic process, regulation of cell–cell adhesion, behavior, skeletal muscle tissue development, regulation of immune effector process, production of molecular mediator of immune response, and cellular calcium ion homeostasis (Fig. 5B and Additional file 9: Table. S6). As for the 71 genes uniquely responding to S + Rg treatment, six GO terms were identified with the top functional annotation clusters being Ras protein signal transduction, leukocyte activation, and organic acid binding (Fig. 5C and Additional file 10: Table. S7). The results support the assumption that carbohydrates may have direct regulatory effect or collaborate with ginsenosides to enable new functions when in combination, and therefore, functionally contribute to the hematopoietic process during chemotherapy-induced myelosuppression.

Fig. 5figure 5

Analysis of DEGs regulated in S and S + Rg groups or uniquely in S + Rg group. A Venn diagram of co-regulated callback DEGs between three treatment group and Model group. B GO analysis of common DEGs in S and S + Rg groups. C GO analysis of DEGs uniquely regulated in S + Rg group

Hematopoiesis-related molecular network revealed genes and pathways regulated by carbohydrates and ginsenosides

To further understand if and how carbohydrates contribute to the hematopoietic function of SMI, we built the hematopoiesis-related molecular networks of deregulated genes that were significantly altered in 5-FU-induced myelosuppressive mice. The DEGs caused by myelosuppression were applied to KEGG pathway enrichment (Additional file 11: Table. S8) and interaction network generation (Fig. 6 for up-regulated genes, Model vs Control, Fig. 7 for down-regulated genes, Model vs Control, the main network in gray).

Fig. 6figure 6

Hematopoiesis-related molecular networks based on genes upregulated in Model vs Control. A Model group. B S + Rg group. C S group. D Rg group. Nodes are representative of the genes upregulated in Model vs Control. Lines represent the interaction between the genes. Colors of gene nodes indicate differentially enriched pathways

Fig. 7figure 7

Hematopoiesis-related molecular networks based on genes downregulated in Model vs Control. A Model group. B S + Rg group. C S group. D Rg group. Nodes are representative of the genes downregulated in Model vs Control. Lines represent the interaction between the genes. Colors of gene nodes indicate differentially enriched pathways

Referring to the enriched activated KEGG pathways, the DEGs related to top ranked pathways, including Axon guidance, Hippo signaling pathway, ECM-receptor interaction, Rap1 signaling pathway, PI3K-Akt pathway, TGF-beta signaling pathway, Cytokine-cytokine receptor interaction, Focal adhesion, Wnt signaling pathway, and Glutathione metabolism were annotated with different colors and highlighted in the network (Fig. 6A, Additional file 11: Table. S8, Model vs Control up). As illustrated in Fig. 6B–D, in which the co-regulated DEGs based on individual treatments are highlighted, the expression of the relevant genes appears to be effectively regulated by individual treatments. We specifically examined the expression patterns of collagen-related genes including Col1a1, Col1a2, and Col6a2 in ECM-receptor interaction pathway, as well as genes encoding bone morphogenetic proteins including Bmp4, Bmp5, and Bmp6. These genes were identified as DEGs upregulated due to myelosuppression, which could be regulated by carbohydrates, ginsenosides, and their combination (Additional file 11: Table. S8, S + Rg, S, Rg vs Model down). This result was in accordance with our observation in the histopathological analysis of bone marrow that Masson staining showed that 5-FU modeling led to large numbers of reticulin fibers and collagen, which were reduced after SMI, S + Rg, S, or Rg treatment. We also noticed that amongst the genes in the Glutathione metabolism pathway, Gstt1, Gstp2, Gsta4, and Oplah were identified as DEGs regulated by carbohydrates but not ginsenosides (Fig. 6C, D, Additional file 11: Table. S8, S + Rg, S, Rg vs Model down). In accordance with the RNA-seq result, the immunohistochemical staining of bone marrow indicated that the expression of GSTT1 was increased in the Model group compared with the Control group, which could be downregulated by carbohydrates but not ginsenosides (Additional file 12: Fig. S4).

Referring to the enriched inhibitory KEGG pathways, the DEGs related to top ranked pathways, including Hematopoietic cell lineage, Cell adhesion molecules (CAMs), Primary immunodeficiency, Osteoclast differentiation, B cell receptor signaling pathway, Cytokine-cytokine receptor interaction, Protein digestion and absorption, NF-kappa B signaling pathway, Phagosome, and Complement and coagulation cascades were annotated with different colors and highlighted in the network (Fig. 7A, Additional file 11: Table. S8, Model vs Control down). As illustrated in Fig. 7B, D, in which the co-regulated DEGs based on individual treatments are highlighted, the expression of the relevant genes also appears to be effectively regulated by individual treatments. Nevertheless, alternative regulatory effects were observed: NF-kappa B signaling pathway could be regulated by ginsenosides but not carbohydrates; genes involved in B cell receptor signaling pathway including Cd19, Cd79a, and Cd79b could be identified as DEGs regulated by carbohydrates but not ginsenosides. (Fig. 7C, D, Additional file 11: Table. S8, S + Rg, S, Rg vs Model up).

Taken together, these molecular networks demonstrate the effects of carbohydrates, ginsenosides and their combination at the molecular level on hematopoiesis. Carbohydrates in SMI per se may display unique effects in inhibiting the Glutathione metabolism pathway and stimulating the B cell receptor signaling pathway during myelosuppression.

留言 (0)

沒有登入
gif