Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Mar 29;4(6):e202000935.
doi: 10.26508/lsa.202000935. Print 2021 Jun.

Pancreatic cancer is marked by complement-high blood monocytes and tumor-associated macrophages

Affiliations

Pancreatic cancer is marked by complement-high blood monocytes and tumor-associated macrophages

Samantha B Kemp et al. Life Sci Alliance. .

Abstract

Pancreatic ductal adenocarcinoma (PDA) is accompanied by reprogramming of the local microenvironment, but changes at distal sites are poorly understood. We implanted biomaterial scaffolds, which act as an artificial premetastatic niche, into immunocompetent tumor-bearing and control mice, and identified a unique tumor-specific gene expression signature that includes high expression of C1qa, C1qb, Trem2, and Chil3 Single-cell RNA sequencing mapped these genes to two distinct macrophage populations in the scaffolds, one marked by elevated C1qa, C1qb, and Trem2, the other with high Chil3, Ly6c2 and Plac8 In mice, expression of these genes in the corresponding populations was elevated in tumor-associated macrophages compared with macrophages in the normal pancreas. We then analyzed single-cell RNA sequencing from patient samples, and determined expression of C1QA, C1QB, and TREM2 is elevated in human macrophages in primary tumors and liver metastases. Single-cell sequencing analysis of patient blood revealed a substantial enrichment of the same gene signature in monocytes. Taken together, our study identifies two distinct tumor-associated macrophage and monocyte populations that reflects systemic immune changes in pancreatic ductal adenocarcinoma patients.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no conflict of interest.

Figures

Figure S1.
Figure S1.. Biomaterial scaffolds harbor an immune-dense microenvironment in response to an orthotopic model of pancreatic ductal adenocarcinoma.
(A) Manual gating of CyTOF results for total immune cells (CD45+), myeloid cells (CD45+ CD11b+), myeloid-derived suppressor cells (MDSCs) (Ly-6C+ Ly-6G+), total macrophages (CD11b+ F4/80+), and macrophage subset (F4/80+ CD206+) in control liver (n = 4) compared with tumor-bearing (TB) liver (n = 5) from an orthotopic mouse model of pancreatic cancer. Results are plotted as percent of total live singlets, except for CD206+ macrophages, which are plotted as percent of total myeloid cells. Statistical significance was determined using two-tailed t tests. Data presented as means ± standard error (SEM) and P < 0.05 was considered statistically significant. (B) Manual gating of CyTOF results for statistically significant changes in immune cell populations in PBMCS from control (n = 5) and TB (n = 5) orthotopic mice. Results are plotted as percent of total live singlets, except for MDSCs, which are plotted as percent of total myeloid cells. Statistical significance was determined using two-tailed t tests. Data presented as means ± standard error (SEM) and P < 0.05 was considered statistically significant. (C) Representation of scaffold size. (D) Hierarchical clustering and heat map of inflammatory genes in control scaffolds (black) compared with TB scaffolds (yellow) over time. w, weeks. Red is high expression; blue is low expression. (E) Manual gating of CyTOF results for total myeloid cells (CD45+ CD11b+), MDSCs (Ly-6G+ Ly-6C+), total macrophages (CD11b+ F4/80+), NK cells (CD45+ NK1.1+), total T cells (CD45+ CD3+), and CD4 T cells (CD3+ CD4+) in control scaffolds (n = 10) compared with TB scaffolds (n = 9–10). Results are plotted as percent of total live singlets. Statistical significance was determined using two-tailed t tests. Data presented as means ± standard error (SEM) and P < 0.05 was considered statistically significant. Source data are available for this figure.
Figure 1.
Figure 1.. Biomaterial scaffolds harbor an immune-dense microenvironment in response to an orthotopic model of PDA.
(A) Experimental scheme. Scaffolds were subcutaneously implanted as described in the Materials and Methods section. 7940b (BL/6) cells derived from the LSL-KrasG12D/+; LSL-Trp53R172H/+; Pdx1-Cre (KPC) were orthotopically implanted into the pancreas. Scaffolds were harvested 3 wk post tumor cell inoculation. (B) Co-immunofluorescence of scaffolds from animals who underwent mock-surgery (left) compared with tumor-bearing (TB) mice (right). Tumor cells are marked by CK19 (green), macrophages by F4/80 (red), fibroblasts by αSMA (pink), and nuclei by DAPI (blue). Scale bars, 50 μm. (C) Representative t-SNE plots for the scaffold infiltrate from control and TB scaffolds. Identified populations include, myeloid-derived suppressor cells (blue), Ly-6C+ macrophages (orange), PD-L1+ macrophages (green), CD206+ macrophages (red), CD8+ T cells (purple), CD4+ T cells (brown), NK cells (pink), endothelial cells (grey), and fibroblasts (light green). (D) Manual gating of CyTOF results for macrophage subsets (F4/80+ CD206+; F4/80+ PD-L1+) in control scaffold (n = 8) compared with TB scaffold (n = 7–8). Results are plotted as percent of total myeloid cells (%CD11b+). Statistical significance was determined using two-tailed t tests. Data presented as means ± standard error (SEM) and P < 0.05 was considered statistically significant. (E) Manual gating of CyTOF results for endothelial cells (CD45 PECAM1+), fibroblasts (CD45 PDGFRα+) and CD8+ T cells (CD3+ CD8+) in control scaffold (n = 10) compared with TB scaffold (n = 10). Results are plotted as percent of total live singlets. Statistical significance was determined using two-tailed t tests. Data presented as means ± standard error (SEM) and P < 0.05 was considered statistically significant. Source data are available for this figure.
Figure 2.
Figure 2.. Identification of a pancreatic cancer–specific gene signature.
(A) Plot of Bagged Tree/singular value decomposition prediction produced from inflammatory gene OpenArray. Plot highlights the divergence of tumor-bearing (TB) scaffolds (red) from healthy control (HC) scaffolds (black). n = 6 for control and n = 6 for TB scaffolds. Each dot represents a single mouse. Black line indicates 99.5% confidence intervals. Filled ovals denote the mean for control (black) and TB (red) scaffolds for pooled control or TB scaffolds. (B) Hierarchical clustering and heat map of 21 inflammatory genes of interest in control (n = 6) scaffolds compared with TB scaffolds (n = 6). (C) UMAP visualization of control scaffold (n = 1) and TB (n = 1) scaffolds from an orthotopic mouse model of pancreatic cancer. (D) Dot plot shows average expression of scaffold signature in merged control and TB scaffold infiltrate. Size of dot represents percent expressed. Color of dot represents average expression.
Figure S2.
Figure S2.. Identification of a pancreatic cancer-specific gene signature.
(A) Hierarchical clustering and heat map analysis of all gene expression values from the OpenArray analysis. Black boxes indicate control scaffolds (n = 6) and red boxes indicate tumor-bearing (TB) scaffolds (n = 6). (B) Volcano plot analysis from all OpenArray gene expression data. Black lines indicate fold change of ± 2. Red line indicates P = 0.05 determined by two-tailed t tests. (C) Dot plot of select lineage markers used to define captured cell populations. Color of the dot represents average expression. Size of the dot represents expression frequency. (D) Representative immunohistochemical analysis of Ym1 in control liver and TB liver from mice with orthotopically transplanted pancreatic ductal adenocarcinoma tumors. Scale bars, 100 μm. (E) Quantitation of %Ym1-positive cells in a 20× field in control (n = 3) and TB livers (n = 3). Statistical significance was determined using two-tailed t tests. Source data are available for this figure.
Figure 3.
Figure 3.. Identification of two distinct macrophage subsets in scaffold infiltrate.
(A) Average expression heat map for select differentially expressed genes between macrophages from control and tumor-bearing (TB) scaffolds. Low expression is shown in blue and high expression in red. All genes plotted are statistically significant, determined using nonparametric Wilcoxon rank sum test with a P-value cut off of P < 0.05. (B, C) Violin plot of normalized gene expression of select up-regulated and (C) down-regulated genes in macrophages from control (black) and TB (blue) scaffolds. Statistically significant genes were determined using non-parametric Wilcoxon rank sum test with a P-value cutoff of P < 0.05. (D) UMAP visualization of scaffold-associated macrophage (SAM) 1 (navy) and SAM 2 (pink) subsets in control and TB scaffolds. (E, F) Violin plots of normalized expression of C1qa, C1qb, and Trem2 in SAM 1 and (F) Chil3, Plac8, and Ly6c2 in SAM 2.
Figure S3.
Figure S3.. Identification of two distinct macrophage subsets in scaffold infiltrate.
(A) UMAP visualization of scaffold-associated macrophage (SAM) 1 and SAM 2 subsets in merged control (grey) and tumor-bearing scaffolds (orange). (B) Heat map visualization at single-cell resolution of the top 20 genes that unbiasedly define each scaffold macrophage subset. Low expression is in purple, whereas high expression is in yellow. (C) Violin plots of normalized expression for select macrophage markers in scaffold-associated macrophage subsets. (D) Heat map visualization of average expression of macrophage differentially expressed genes between control and tumor-bearing scaffolds in SAM 1 and SAM 2 subsets.
Figure 4.
Figure 4.. Macrophages in mouse pancreatic tumors overexpress TREM2 and complement genes.
(A) UMAP visualization of mouse orthotopic pancreatic tumors (n = 2). (B) Dot Plot of scaffold-associated macrophage signature, Chil3, Plac8, Ly6c2, C1qa, C1qb, and Trem2 in identified cell populations in the orthotopic KPC tumors. Color represents average expression, whereas size of dot represents percent expressed. (C) UMAP visualization of Chil-tumor-associated macrophages (TAMs) (pink) and Cq-TAMs (navy) subsets in mouse orthotopic pancreatic tumors. (D) Violin plots of C1qa, C1qb, Trem2, Chil3, Plac8, and Ly6c2 across Chil-TAMs and Cq-TAMs. (E) UMAP visualization of Chil-TAM (pink), Cq-TAM (navy), and TAM (green) macrophage subsets in normal pancreas (n = 1) and orthotopic tumors (n = 2). (F) Violin plot of normalized gene expression of Chil3, Plac8, Ly6c2, C1qa, C1qb, and Trem2 in macrophages from normal pancreas (grey) and orthotopic tumors (navy). Statistically significant genes were determined using non-parametric Wilcoxon rank sum test with a P-value cut offof P < 0.05. (G) Co-immunofluorescence of normal mouse pancreas (N Panc), KPC tumor, and KPC liver metastasis samples of C1q (green), F4/80 (red), E-Cadherin (pink) and DAPI (blue). Red arrow denotes C1q F4/80+ macrophage in the normal pancreas. Yellow arrows denote C1q+ F4/80+ macrophages in KPC tumor and KPC liver metastasis. Inlets show higher magnification of select macrophages in boxed region. Scale bars, 50 μm.
Figure S4.
Figure S4.. Macrophages in human and mouse pancreatic tumors overexpress TREM2 and complement genes.
(A) Dot plot of select lineage markers used to define captured cell populations in mouse orthotopic KPC tumors (n = 2). Color of the dot represents average expression. Size of the dot represents expression frequency. (B) Heat map visualization at single-cell resolution of the top 10 genes that define Chil-tumor-associated macrophages (TAMs) and Cq-TAMs. Low expression is in purple, whereas high expression is in yellow. (C) Dot plot of Nos2, Tnf, Chil3, Ly6c2, Plac8, C1qa, C1qb, Trem2, Mrc1, and Cd163 in Cq-TAMs and Chil-TAMs from orthotopic KPC tumors. Color of the dot represents average expression. Size of the dot represents expression frequency. (D) Violin plots for Ccr2, Cd74, H2-Eb1, Chil3, Plac8, Ly6c2, C1qa, C1qb, and Trem2 in Chil-TAM (pink), Cq-TAM (navy), and TAM (green) subsets. (E) Average expression heat map for select differentially expressed genes between macrophages from orthotopic scaffolds and mouse orthotopic primary tumors. Low expression is shown in blue and high expression in red. All genes plotted are statistically significant, determined using non-parametric Wilcoxon rank sum test with a P-value cut off of P < 0.05. (F) Violin plots of normalized expression for Arg1, Il1a, Rgs1, Cxcl3, Mif, and Ifitm1 in macrophages from orthotopic scaffolds and mouse orthotopic primary tumors.
Figure 5.
Figure 5.. Cq-tumor–associated macrophage (TAM) and Chil-TAM markers are elevated in the iKras* p53* model of pancreatic cancer.
(A) UMAP visualization of scaffolds from control (n = 1) and iKras* p53* mice (n = 1). (B) Average expression heat map of Trem2, C1qb, Chil3, and C1qa in control and iKras* p53* scaffolds. High expression is in red, whereas low expression is in blue. (C) UMAP visualization of scaffold-associated macrophage (SAM) 1 (navy) and SAM 2 (pink) macrophage subsets in control and iKras* p53* scaffolds. (D) Violin plots of C1qa, C1qb, Trem2, Chil3, Plac8, and Ly6c2 across SAM 1 and SAM 2. (E) UMAP visualization of CCR2-TAM (green), Chil-TAM (pink), and Cq-TAM (navy) macrophage subsets in control, iKras* and iKras* p53* pancreas samples. (F) Violin plots of Chil3, Plac8, Ly6c2 C1qa, C1qb, and Trem2 across CCR2-TAM, Chil-TAM, and Cq-TAM macrophage subsets. (G) Dot plot of C1qa, C1qb, Trem2, Chil3, Plac8, Ly6c2, Ccr2, Cd74, and H2-Eb1 in control, iKras* and iKras* p53* macrophages. Color represents average expression. Size of the dot represents percent expressed.
Figure S5.
Figure S5.. Cq-tumor–associated macrophage (TAM) and Chil-TAM markers are elevated in the iKras* p53* model of pancreatic cancer.
(A) Schematic for iKras* and iKras* p53* experimental design. (B) Dot plot for lineage markers used to define identified cell populations in scaffolds from iKras* p53* mice. Color of the dot represents average expression. Size of the dot represents percent expressed. (C) UMAP visualization of from control/normal pancreas (n = 2), iKras* (n = 1), and iKras* p53* pancreas samples (n = 2). (D) Dot plot for lineage markers used to define identified cell populations in control/normal pancreas, iKras*, and iKras* p53* pancreas samples. Color of the dot represents average expression. Size of the dot represents expression frequency. (E) Violin plots for Ccr2, Cd74, H2-Eb1 in CCR2-TAM, Chil-TAM, and Cq-TAM subsets in mouse pancreas samples. (F) Violin plots for C1qa, C1qb, Trem2, Chil3, Plac8, Ly6c2, Ccr2, Cd74, and H2-Eb1 in control/normal pancreas, iKras*, and iKras* p53* pancreas samples.
Figure 6.
Figure 6.. Macrophages in human pancreatic tumors overexpress TREM2 and complement genes.
(A) UMAP visualization of Adj/Norm (n = 3) and pancreatic ductal adenocarcinoma (PDA) tumors (n = 16). (B) Dot plot of TREM2, C1QB, C1QA, and PLAC8 in human PDA tumor cell populations. Color of the dot represents average expression, whereas the size of the dot represents expression frequency. (C) UMAP visualization of human tumor-associated macrophages (TAMs) (pink) and CQ-TAMs (navy) from adjacent normal pancreas (n = 3) and human PDA tumors (n = 16). (D) Violin plots of C1QA, C1QB, and TREM2 in human TAMs and CQ-TAMs. (E) Violin plots of C1QA, C1QB, and TREM2 in human macrophages from human PDA tumors compared with adjacent normal pancreas. Statistics were determined using non-parametric Wilcoxon rank sum test with a P-value of P < 0.0001.
Figure S6.
Figure S6.. Macrophages in human pancreatic tumors overexpress TREM2 and complement genes.
(A) Heat map visualization at single-cell resolution of the top 20 genes that define human CQ-tumor–associated macrophages (TAMs) and TAMs. Low expression is in purple and high expression is in yellow. (B) UMAP visualization of human TAMs (pink) and CQ-TAMs (navy) from in individual human pancreatic ductal adenocarcinoma tumors (n = 16). The four-digit number represents the de-identified patient ID. (C) Dot plot of C1QA, C1QB, TREM2, PLAC8, VCAN, FABP5, and RETN in human CQ-TAMs and TAMs. Color of the dot represents average expression. Size of the dot represents expression frequency.
Figure 7.
Figure 7.. Macrophages in human liver metastases express high levels of TREM2 and complement genes.
(A) UMAP visualization of human liver metastasis samples (n = 5) from pancreatic ductal adenocarcinoma patients. (B) Violin plots of normalized expression of C1QA, C1QB, and TREM2 in identified cell populations in the liver metastasis lesions from human pancreatic ductal adenocarcinoma patients (n = 5). (C) UMAP visualization of CQ-tumor–associated macrophages (TAMs) (navy) and TAMs (pink) identified in human liver metastasis samples. (D) Violin plots of normalized expression for C1QA, C1QB, and TREM2 in CQ-TAMs and TAMs from liver metastasis samples. (E) Average expression heat map for select differentially expressed genes between macrophages from human liver metastases and human primary tumors. Low expression is shown in blue and high expression in red. All genes plotted are statistically significant, determined using nonparametric Wilcoxon rank sum test with a P-value cutoff of P < 0.05. (F) Violin plots of normalized expression for IL1A, IL1B, PLAC8, RGS1, MRC1, and TREM2 in macrophages from human liver metastasis and primary tumor samples.
Figure S7.
Figure S7.. Macrophages in human liver metastases express high levels of TREM2 and complement genes.
(A) Dot plot of lineage markers used to define populations in human liver metastases from pancreatic ductal adenocarcinoma patients (n = 5). Color of the dot represents average expression, whereas the size of the dot represents expression frequency. (B) Feature plots of C1QA, C1QB, and TREM2 in human liver metastases. Blue is high expression, whereas grey is low expression.
Figure S8.
Figure S8.. Complement-high monocyte markers are elevated in the blood of pancreatic cancer patients.
(A) UMAP visualization of pancreatic ductal adenocarcinoma (PDA) PBMCs (n = 16) and healthy PBMCs (n = 4). (B) Dot plot analysis for lineage markers used to define human monocyte subsets in the blood. Color of the dot represents average expression, whereas the size of the dot represents expression frequency. (C) Heat map visualization at single-cell resolution of the top 10 genes that define human monocyte subsets. Low expression is in purple, whereas high expression is in yellow. (D) Feature plot of C1QA, C1QB, TREM2, and PLAC8 in human monocyte subsets in the blood of healthy donors and PDA patients. Blue is high expression and grey is low expression. (E) Violin plots of C1QA, C1QB, and TREM2 in human circulating monocytes from PDA patients compared with healthy donors. Statistics were determined using non-parametric Wilcoxon rank sum test.
Figure 8.
Figure 8.. Complement-high monocyte markers are elevated in the blood of pancreatic cancer patients.
(A) Dot plot of C1QA, C1QB, and TREM2 in identified populations in human PBMCs. Color of the dot represents average expression, whereas the size of the dot represents expression frequency. (B) UMAP visualization of CQ-monocytes (navy), monocyte 1 (pink), monocyte 2 (green), and monocyte 3 (purple) in human PBMCs in pancreatic ductal adenocarcinoma (n = 16) and healthy (n = 4). (C) Feature plot of C1QA, C1QB, TREM2, and PLAC8 in human monocyte subsets in the blood. Blue is high expression and grey is low expression. (D) Dot plot of C1QA, C1QB, and TREM2 in PBMCs from healthy donors and pancreatic ductal adenocarcinoma patients. High expression is in blue, low expression is in grey. Size of the dot represents expression frequency.

Similar articles

Cited by

References

    1. Afshar-Kharghan V (2017) The role of the complement system in cancer. J Clin Invest 127: 780–789. 10.1172/jci90962 - DOI - PMC - PubMed
    1. Aguado BA, Bushnell GG, Rao SS, Jeruss JS, Shea LD (2017) Engineering the pre-metastatic niche. Nat Biomed Eng 1: 0077. 10.1038/s41551-017-0077 - DOI - PMC - PubMed
    1. Aiello NM, Rhim AD, Stanger BZ (2016) Orthotopic injection of pancreatic cancer cells. Cold Spring Harb Protoc 2016: pdb.prot078360. 10.1101/pdb.prot078360 - DOI - PubMed
    1. Azarin SM, Yi J, Gower RM, Aguado BA, Sullivan ME, Goodman AG, Jiang EJ, Rao SS, Ren Y, Tucker SL, et al. (2015) In vivo capture and label-free detection of early metastatic cells. Nat Commun 6: 8094. 10.1038/ncomms9094 - DOI - PMC - PubMed
    1. Benoit ME, Clarke EV, Morgado P, Fraser DA, Tenner AJ (2012) Complement protein C1q directs macrophage polarization and limits inflammasome activity during the uptake of apoptotic cells. J Immunol 188: 5682–5693. 10.4049/jimmunol.1103760 - DOI - PMC - PubMed

Publication types

MeSH terms