Altered trophoblasts, macrophages, and immunologic microenvironment in the placenta: Implications of poorly controlled type 2 diabetes mellitus revealed by single-cell transcriptomics

To the Editor: Different from gestational diabetes mellitus, pregnant women with poorly controlled type 2 diabetes mellitus (T2DM) experience hyperglycemia throughout their whole pregnancies,^[1] potentially interfering with normal placenta formation during the first trimester. Poorly controlled T2DM is more likely to cause serious maternal and fetal complications, such as major malformation, unexplained fetal demise, macrosomia, and small for gestational age.^[1,2] The placenta, which sits between the fetus and mother, is an important mediator between hyperglycemic exposure and maternal and fetal complications. Here, we utilized the 10× Genomics platform to generate the single-cell transcriptomic profiles of the full-term placental villi obtained from pregnant women with poorly controlled T2DM and normal pregnant women. In T2DM placenta, increased macrophages_S100A8/9^high (Macro_S100A8/9^high) at the villous site and nutrient transport defects in syncytiotrophoblasts (STBs) were discovered. We identified elevated proinflammatory regulatory signals from macrophages to placental trophoblasts in T2DM villi, including S100A8/9-TLR4 and IL1B-IL1R1/IL1RAP, which were found to interfere with the normal functions of trophoblasts.

This study was approved by the Ethics Committee of Peking University First Hospital (No. 2021[488]). Informed consent was obtained from all the participants who donated their samples. Placentas of four healthy full-term singleton pregnancies and four poorly controlled T2DM full-term singleton pregnancies were recruited [Supplementary Table 1, http://links.lww.com/CM9/C645]. All four women in the disease group were diagnosed with T2DM before this pregnancy event and received insulin treatment during their pregnancy. We defined hemoglobin A1c (HbA1c) more than 6.0% and glycated albumin (GA) more than 16.0% as poorly controlled glycemia.^[3] Glycemic levels in four women with T2DM met the specified criteria across early, middle, and late gestation [Supplementary Table 2, http://links.lww.com/CM9/C645]. We excluded patients with severe pregnancy complications, including preeclampsia, fetal growth restriction, rheumatological disease, cardiovascular illness, and infectious diseases.

After standard 10× Genomics sequencing procedures, a total of 122,687 single cells were obtained. Qualified cells, defined as mitochondrial gene percentages lower than 10% and genes per cell between 200 and 7500, were retained. Doublet removal was subsequently conducted. Altogether, a total of 45,678 cells from the T2DM group and 51,578 cells from the control (CTRL) group were collected for further analysis.

To investigate the cellular heterogeneity in T2DM placenta, we collected placenta from four pregnant women with poorly controlled T2DM and four women with normal pregnancies, and single-cell RNA sequencing (scRNA-seq) analysis was performed with a non-marker-selected strategy [Supplementary Figures 1A and B, http://links.lww.com/CM9/C645]. Ten cell subtypes in the T2DM and CTRL villi samples presented a uniform manifold approximation and projection (UMAP) distribution [Supplementary Figure 1C, http://links.lww.com/CM9/C645]. As presented in Supplementary Figure 1D, http://links.lww.com/CM9/C645, the cells with high expression of KRT7 were trophoblasts, and the CD14-positive and CD68-positive cluster was macrophage (Macro). The trophoblasts were further divided into cytotrophoblasts (CTBs), proliferative CTBs (CTB_pros), extravillous trophoblasts (EVTs), and STBs based on specific markers such as PEG10, DNA TOP2A, HLA-G, and CYP19A1, respectively. The cells with high levels of COL1A1 and COL1A2 were fibroblasts (FIBs), and those with high expression of CD34 were endothelial cells (Endos). The cluster expressing HBB was erythrocyte (EB). B cells and natural killer/T cells were defined based on markers such as CD79A, CD7, and NKG7 [Supplementary Figure 1D, http://links.lww.com/CM9/C645]. We further explored the cellular proportions of each cell type in the T2DM and CTRL groups. The T2DM placenta showed increased Macros, NK/T cells, FIBs, and B cells, and decreased CTBs and CTB_pros compared to the CTRL placenta, as shown in Supplementary Figure 1E, http://links.lww.com/CM9/C645.

To gain an in-depth understanding of the heterogeneity of trophoblasts in T2DM placenta compared with the normal placenta, we reclustered the abovementioned trophoblasts [Supplementary Figure 2A, http://links.lww.com/CM9/C645]. Further, the CTB_fusion subtype expressed ERVFRD-1, HSD11B2, and ERVV-1 specifically, and the STBs highly expressed PSG, KISS1, and CSH2. EVT_JAM2⁺ showed high expression of JAM2 and IL1R2. EVT_DIO2⁺ specifically expressed DIO2, TNNI2, and HAPLN3. CTB_pros exhibited high levels of proliferation-related genes, such as TOP2A, TYMS, and MKI67 [Supplementary Figure 2B, http://links.lww.com/CM9/C645]. STBs from T2DM placental villi showed significant nutrient transportation impairment, including glucose, amino acid, and lipid transport, which may be linked to the occurrence of adverse complications [Figure 1A].

Figure 1.

STBs show nutrient transport defects, and Macro_S100A8/9^high exhibit proinflammatory activation in T2DM placentas. (A) GSVA of STBs in the CTRL and T2DM groups. (B) GSVA of Macro_S100A8/9^high in the CTRL and T2DM groups. CTRL: Control; GOBP: Gene ontology biological process; GSVA: Gene set variation analysis; Macro: Macrophages; Slc: Solute carrier; STB: Syncytiotrophoblasts; T2DM: Type 2 diabetes mellitus.

A significant increase in the number of macrophages was found in T2DM placenta. To further explore the potential immunological microenvironment alterations mediated by placental macrophages in T2DM, we reclustered the macrophages [Supplementary Figure 2C, http://links.lww.com/CM9/C645]. The clusters with high expression of HLA-DRA and LPL were Macro_MHC-II⁺ and Macro_LPL⁺, respectively. Macro_S100A8/9^high highly expressed S100A8, S100A9, and S100A12. Macro_CXCL9⁺ specifically expressed CXCL9, CXCL10, CXCL11, and GBP. The CCL4L2-positive and C3-positive cluster was Macro_CCL4L2⁺. The CCL2-positive and TCEB1-positive cluster was Macro_CCL2⁺. Macro_MT1G⁺ was specifically expressed MT1G. Hofbauer cells expressed FOLR2 specifically, and LYVE1 and CD209 were also expressed abundantly [Supplementary Figure 2D, http://links.lww.com/CM9/C645]. T2DM placenta showed increased Macro_S100A8/9^high and Macro_IL1B⁺ and decreased Hofbauer cells and Macro_CXCL9⁺ [Supplementary Figure 2C, http://links.lww.com/CM9/C645].

Differentially expressed genes (DEGs) in each macrophage subcluster were separated by CTRL and T2DM group. The basic functions of each molecule were also shown based on gene ontology biological process (GOBP) database, as well as previously published papers. Hofbauer cells and Macro_S100A8/9^high in T2DM villi exhibited enhanced proinflammatory characteristics. The T2DM Macro_MT1G⁺ was more related to angiogenesis and induction of chemotaxis than the CTRL group. Macro_CCL2⁺ in T2DM villi was also more related to angiogenesis [Supplementary Figure 2E, http://links.lww.com/CM9/C645]. Considering that the number of Macro_S100A8/9^high significantly increased in T2DM group and played a proinflammatory role, we further analyzed the differences of Macro_S100A8/9^high between the two groups. Gene set variation analysis (GSVA) indicated that T2DM Macro_S100A8/9^high showed greater enrichment in terms such as angiogenesis, migration, and inflammation [Figure 1B]. Meanwhile, CTRL Macro_S100A8/9^high was more related to the induction of positive chemotaxis [Figure 1B].

To further explore the cell–cell interaction differences of Macro_S100A8/9^high and trophoblasts between the normal and T2DM placenta, we performed the CellTalkDB analysis. The number of ligand–receptor pairs for Macro_S100A8/9^high and other cell types was increased in the T2DM group. Enhanced cellular communication of cytokine-related pairs such as S100A8/S100A9–TLR4 between Macro_S100A8/9^high and CTB_pros, as well as IL1B–IL1R1/IL1RAP between Macro_S100A8/9^high and STBs, suggested that Macro_S100A8/9^high could enhance the inflammatory level of trophoblasts [Supplementary Figures 2F–I, http://links.lww.com/CM9/C645].

Placental chronic inflammation has been observed in pregnant women with hyperglycemia.^[4] The changes in placental function may be influenced by inflammatory profiles in the placenta under a hyperglycemic environment.^[4] Here, we have also described Macro_S1008/9^high in placenta, which was involved in chronic inflammatory response and migration. Macro_S1008/9^high was increased in the T2DM placenta obviously. Imbalances in the immunological microenvironment, along with decreased Hofbauer cells and increased Macro_S1008/9^high might lead to chronic inflammation and could interfere with normal placental trophoblast functions in T2DM. A better understanding of how maternal hyperglycemia relates to the placental chronic inflammation may lead to the development of better therapeutic interventions to prevent the development of T2DM during pregnancy.

In summary, our study provides evidence of alterations in the immunological microenvironment in the placentae of women with T2DM, which interfere with the normal functions of trophoblasts and underlie the pathogenesis of T2DM during pregnancy. These alterations may provide possible intervention targets in the future.

Funding

This work was supported by the National Key Research and Development Program of China (Nos. 2021YFC2700700 and 2021YFC2700302), the National Natural Science Foundation of China (Nos. 81830044 and 31900602), and the Scientific Research Seed Fund of Peking University First Hospital (No. 2024SF89).

Conflicts of interest

None.

Supplementary Material

cm9-139-779-s001.docx ^{(1,020KB, docx)}