CRC has been called "Intestinal accumulation", "Accumulation" or "Intestinal mushroom" in TCM. Modern Chinese medicine physicians considered that the occurrence of CRC is due to the deficiency of Vital Qi, and the stagnation and accumulation of Evil Qi such as "Damp heat", Blood stasis" and "Poison stagnation" in the large intestine for long time, which resulted in the main pathogenesis in TCM of CRC is the "Deficiency of Vital Qi and Excess of Evil Qi" and "Combination of Deficiency and Excess"[26, 27]. Previous study on the distribution of clinical TCM syndromes in 760 cases of CRC has showed 565 in DS, 81 in ES, 52 in DES and 62 in no syndrome (NS), among DS is the most common syndrome type in CRC patients25. In this study, we used Smart-seq2 technology to explore the relationship between the heterogeneity of tumor cells and microcirculation and TCM syndromes in CRC through analyzing the classification of tumor cells and the proportion of each subgroup, the characteristic genes, gene co-expression network, the functional interpretation and the evolution of monocle functions.
Tumor heterogeneity is mainly divided into two types including inter-tumor heterogeneity and intra-tumor heterogeneity . Almendro et al  have classified intra-tumoral heterogeneity into cell, gene and functional heterogeneity, and temporal and spatial heterogeneity under the perspective of systems biology. Aaron S et al.  divided it into genetic and, phenotypic heterogeneity, cell signaling and pathway activity heterogeneity, and temporal and spatial heterogeneity. The manifestations of the tumor heterogeneity are complex and changeable. Different forms of heterogeneity may occur throughout the process of tumor formation and development. It needs to pay attention to is the relationship between the changes in molecular mechanisms and different functions in the process of the tumor heterogeneity. In this study, we found that the intra-tumor cells in CRC have 14 different cell subpopulations, 11 known cell lineages, and their genes have different functions and signal transduction pathways, as well as the development of different single cell lineages, showing obvious cells and genes and their phenotypic heterogeneity and cell signaling and pathway activity heterogeneity. Moreover, the high expressions of MUC2 and REG4 in DES were mainly related to brain development, cytogenesis, MAPK and cAMP signaling pathways; the high expressions of COL1A2 and POSTN genes in ES were mainly related to vascular development, skeletal muscle development, and PI3k-Akt Pathway; SDPR and GPX1 genes were highly expressed in DS, mainly related to vomiting, platelet decomposition, and endocytosis. These results suggested the coloration of TCM syndromes classification and the heterogeneity of tumor cell subpopulations, and their characteristic genes, phenotypes and signaling pathways in CRC.
The tumor microenvironment includes all the components of non-cancerous solid tumors. It is a complex ecosystem composed of multiple cell types, which play different roles in tumor development, and is highly heterogeneous . This study found 14 different cell subpopulations and 11 known cell lineages among 662 cells isolated from 11 primary CRC tumor tissues, including CRC cancer cells (CEACAM5+, EPCAM+, KRT 18+ and MUC1+ cells), stem cells, immune cells (myeloid-monomacrophages, macrophages, DC, Goblet cells, T and B cells), fibroblast cells (ETM like and CAFs) and enterocyte. However, the proportion of cell subpopulations among ES, DS and DES in CRC is different. The main distribution of CEACAM5+ cells subpopulation was in DES; myeloid-monomacrophages, KRT 18+ cells, DC and MUC1+ cells subpopulations were the most distributed in DS; and goblet cells and fibroblasts subpopulations were the most distributed in ES. It was suggested that the TCM syndromes classification may be related to tumor cell subpopulations and the heterogeneity of tumor microenvironment in CRC.
Moreover, the pseudo-time analysis of monocle functional evolution showed that, 4-cell subpopulations of cancer cells in CRC with TCM syndromes appeared in different states and at different pseudo-times. Among, that of DES was mainly distributed in the middle part, ES was in the second part, and DS was all over the trajectory of pseudo-times, indicating that the DS run through the whole process of CRC tumor cell function evolution, the middle stage is the DES, and the later stage is mainly the ES. These results might provide scientific evidence to clarify the main pathogenesis of CRC in TCM, the "Deficiency of Vital Qi and Excess of Evil Qi" and "Combination of Deficiency and Excess" from the approach of tumor cell development.
CAFs are a kind of activated fibroblasts and are important stromal cells in the tumor microenvironment. In CRC, they interact with tumor cells to promote tumor occurrence, development and metastasis . This study identified two cell subpopulations of fibroblasts including CAFs and EMT-like fibroblasts from CRC tumor tissues, which is consistent with the results of Li H et al.  in the previous research of tumor single cell in CRC. In this study, an EMT-related gene, fibronectin FN1 was highly expressed, and there was a significant difference between DES and ES, DES and DS or ES and DS in CRC. ZEB1 is a EMT transcription factor, it has been reported that the high expression of ZEB1 correlates with liver metastasis and poor prognosis in CRC [33, 34]. Moreover, TCF4, a transcription factor, participated in the regulation of Wnt/β-catenin signaling pathway in CRC cells [35, 36], and promotes adriamycin resistance and cell stemness by regulating the expression of EMT-related ZEB1 and ZEB2. This study found that the high expressions of TCF4 and ZEB1 genes have significant difference between DES and ES or ES and DS in CRC. It suggested that CAFs and EMT-related genes, the high expressions of FN1, ZEB1 and TCF4 in the tumor microenvironment may be involved in TCM syndromes classification in CRC.
Cytokeratin is the intermediate filament of the cell body, which can be divided into 20 different types according to its molecular weight and isoelectric point . Among, KRT8, 18 and 19 have been reported to correlate with CRC , and their expression changes associate with progression towards neoplasia . Moreover, KRT8 has been indicated epithelial to mesenchymal transition , and loss of K8 phosphorylation was also suggested to promote tumor migration and formation of metastasis . KRT 18 and 19 increase in peripheral blood as soluble fragments in CRC . In this study, Smart-seq2 analysis shows that, KRT8, 18 and 19 genes were enriched in the tumor tissues of CRC, and the high expressions of KRT18 and KRT19 genes were significant differences in DES and DS or ES, ES and DS classification, and the high expressions of KRT8 gene has significant differences between DES and ES or DES and DS in patients with CRC.
ELF3 is defined by their highly conserved ETS DNA binding domain and predominant epithelial-specific expression profile . The ELF3 drives beta-catenin transactivation and associates with poor prognosis in CRC[44, 45]. We found that the high expression of ELF3 gene has a significant difference between DES and ES, DES and DS or ES and DS in CRC. SERPINE1, a clade E member of the serine protease inhibitor gene family and a prominent regulator of the pericellular proteolytic microenvironment . In addition to play important roles in cell adhesion, migration and invasion, it has been reported to induce tumor vascularization and promote cell dissemination and tumor metastasis , and involved in the survival and prognosis of CRC [48, 49]. In this study, we found that, the high expressions of SERPINE1 has significant difference between ES and DS in CRC. The above results suggest that the high expressions of KRT19, KRT 18, KRT 8, ELF3 and SERPINE1 genes are potential candidates for identifying TCM syndrome types in CRC, which need to be confirmed by further researches.