Lysosomal Acid Lipase Deficiency Controls T- and B-Regulatory Cell Homeostasis in the Lymph Nodes of Mice with Human Cancer Xenotransplants

Abstract

Utilization of proper preclinical models accelerates development of immunotherapeutics and the study of the interplay between human malignant cells and immune cells. Lysosomal acid lipase (LAL) is a critical lipid hydrolase that generates free fatty acids and cholesterol. Ablation of LAL suppresses immune rejection and allows growth of human lung cancer cells in lal^-/- mice. In the lal^-/- lymph nodes, the percentages of both T- and B-regulatory cells (Tregs and Bregs, respectively) are increased, with elevated expression of programmed death-ligand 1 and IL-10, and decreased expression of interferon-γ. Levels of enzymes in the glucose and glutamine metabolic pathways are elevated in Tregs and Bregs of the lal^-/- lymph nodes. Pharmacologic inhibitor of pyruvate dehydrogenase, which controls the transition from glycolysis to the citric acid cycle, effectively reduces Treg and Breg elevation in the lal^-/- lymph nodes. Blocking the mammalian target of rapamycin or reactivating peroxisome proliferator-activated receptor γ, an LAL downstream effector, reduces lal^-/- Treg and Breg elevation and PD-L1 expression in lal^-/- Tregs and Bregs, and improves human cancer cell rejection. Treatment with PD-L1 antibody also reduces Treg and Breg elevation in the lal^-/- lymph nodes and improves human cancer cell rejection. These observations conclude that LAL-regulated lipid metabolism is essential to maintain antitumor immunity.

Figure 1

Growth of human cancer cells and lymph node (LN) cells in lal^−/− mice. A: The s.c. flank transplantation of human lung A549 cancer cells (1 × 10⁶) into wild-type (WT) or lal^−/− [knockout (KO)] FVB/N mice. Tumor burden was measured by the maximal length and width of tumor: (length × width²)/2. The experiments were independently repeated 12 times. B: The cytotoxicity of lymph node cells from WT or lal^−/− (KO) mice with or without lung A549 cancer cell transplantation. The lymph node cells attaching to lung A549 cancer cells were estimated by fluorescent tracking. The CMTPX-labeled lymph node cells (red) were added to the carboxyfluorescein succinimidyl ester–labeled A549 cells (green) in a 2:1 ratio. A representative image and the statistical analysis of the inclusion rate from five repeated experiments are shown. In each experiment, five randomly selected fields under microscope are chosen for analysis. Arrows show the lymph node cells that were attached and penetrated into lung A549 cancer cells. C: The viability of lymph node cells from WT and lal^−/− (KO) mice with or without lung A549 cancer cell transplantation. The lymph node cells were isolated and cultured overnight in vitro. The percentage of cell survival was determined by flow cytometry using 7-aminoactinomycin D/annexin V staining. The experiments were independently repeated five times. D: The colony morphology of WT and lal^−/− (KO) lymph node cells stimulated by anti-CD3/anti-CD28 antibodies for 2 days. E: Immunohistochemical staining of the T, B, and myeloid populations in lymph nodes from WT and lal^−/− (KO) mice with or without lung A549 cancer cell transplantation. The experiments were independently repeated four times. F: Secretion of interferon (IFN)-γ and IL-10 and intracellular expression of granzyme B (GZB) in CD8⁺ lymph node cells from WT and lal^−/− (KO) mice with or without lung A549 cancer cell transplantation after stimulation of the lung A549 cancer cell lysate for 2 days. IFN-γ and IL-10 in the culture supernatant were determined by enzyme-linked immunosorbent assay. The experiments were independently repeated three to five times. GZB expression in CD8⁺ lymph node cells was analyzed by flow cytometry. The experiments were independently repeated five times. Data are shown as means ± SD (A, C, F). ∗P < 0.05, ∗∗P < 0.01. Scale bars: 50 μm (B and D); 200 μm (E). Original magnification, ×40 (D, top panels); ×200 (D, bottom panels); ×100 (E).

Figure 2

Lymphocyte populations in the wild-type (WT) and lal^−/− lymph nodes (LNs). Lymph node cells were isolated from wild-type and lal^−/− mice with or without lung A549 cancer cell transplantation and analyzed by flow cytometry using antibodies against cell surface markers. A: The total numbers of lymph node cells. B: The total numbers of CD4⁺, CD8⁺, and B220⁺ lymphocytes. C: The total numbers of activated (CD69⁺) CD4⁺, CD8⁺, and B220⁺ lymphocytes. D: The percentages of proliferated (Ki-67⁺) CD4⁺ and CD8⁺ lymphocytes. E: The percentages of CD4⁺CD25⁺ and CD4⁺FOXP3⁺ T-regulatory cells. The gating strategies for B, C, D, and E are shown in Supplementary Figure S2A, S2B, S2C, and S2D, respectively. The experiments were independently repeated four to six times. Data are presented as means ± SD (A–E). ∗P < 0.05, ∗∗P < 0.01. KO, knockout.

Figure 3

Increase of B-regulatory cells in the lal^−/− lymph nodes (LNs). Lymph node cells were isolated from wild-type (WT) and lal^−/− mice with or without lung A549 cancer cell transplantation and analyzed by flow cytometry using antibodies against CD23, B220, and IgM. A: Profiling and gating strategy of B-cell subset follicle (Fo) and transitional two-marginal zone precursor (T2-MZP) cells in lymph nodes from WT and lal^−/− [knockout (KO)] mice. B: The percentages of Fo⁺ and T2-MZP⁺ cells in lymph node cells and the T2-MZP/Fo ratio. The experiments were independently repeated six times. C: The mean fluorescent intensity (MFI) of IL-10 and IL-35 in B220⁺, Fo⁺, and T2-MZP⁺ cells. The gating strategy is shown in Supplemental Figure S3A. The experiments were independently repeated six times for IL-10 and four times for IL-35. Data are given as means ± SD (B and C). ∗P < 0.05, ∗∗P < 0.01.

Figure 4

Increased expression of glucose and amino acid metabolic enzymes in T- and B-regulatory cells of the lal^−/− lymph nodes. Lymph node cells were isolated from wild-type (WT) and lal^−/− mice and analyzed by flow cytometry. A: The mean fluorescent intensity (MFI) of glucose-6-phosphate dehydrogenase (G6PD), pyruvate dehydrogenase (PDH), lactate dehydrogenase A (LDHA), and glutamate dehydrogenase (GLUD) in CD4⁺, CD8⁺, CD4⁺CD25⁺, and CD4⁺FOXP3⁺ cells. B: The MFI of G6PD, PDH, LDHA, and GLUD in B220⁺, follicle (Fo)⁺, and transitional two-marginal zone precursor (T2-MZP)⁺ cells. The gating strategy is shown in Supplemental Figure S4A. The experiments were independently repeated four times. Data are given as means ± SD (A and B). ∗P < 0.05, ∗∗P < 0.01. KO, knockout.

Figure 5

Blocking pyruvate dehydrogenase or mammalian target of rapamycin (mTOR) pathway reduced T- and B-regulatory cells (Tregs and Bregs, respectively) of the lymph nodes (LNs). Lymph node cells were isolated from CPI-613 or rapamycin (Rapa) treated or untreated wild-type (WT) and lal^−/− mice and analyzed by flow cytometry. A: The percentages of follicle (Fo) and transitional two-marginal zone precursor (T2-MZP) Bregs and the T2-MZP/Fo ratio in the lymph node cells of wild-type and lal^−/− mice that were treated with CPI-613 (5 mg/kg). B: The percentages of Tregs in the lymph node cells of wild-type and lal^−/− mice that were treated with CPI-613 (5 mg/kg). C: Expression of mTOR downstream pS6 in Bregs and Tregs. The gating strategy is shown in Supplemental Figure S5. D: The percentages of Fo and T2-MZP Bregs and the T2-MZP/Fo ratio in the lymph node cells of rapamycin-treated or untreated wild-type and lal^−/− mice. E: The percentages of Tregs in the lymph node cells of rapamycin-treated or untreated wild-type and lal^−/− mice. The experiments were independently repeated six times (A and B) or four times (C–E). Data are expressed as means ± SD (A–E). ∗P < 0.05, ∗∗P < 0.01. -, untreated; C, solvent control; KO, knockout; MFI, mean fluorescent intensity; S, solvent.

Figure 6

Blocking mammalian target of rapamycin reduced PD-L1 expression in T- and B-regulatory cells (Tregs and Bregs, respectively) of the lymph nodes and decreased human lung A549 cancer cell growth in lal^−/− mice. A: Immunohistochemical staining of PD-L1 in lymph nodes of wild-type (WT) and lal^−/− [knockout (KO)] mice. B and C: PD-L1 expression in follicle (Fo) and transitional two-marginal zone precursor (T2-MZP) Bregs (B) and Tregs (C) of the lymph node cells from rapamycin (Rapa)–treated or untreated WT and lal^−/− mice. D: Human lung A549 cancer cells were transplanted into WT or lal^−/− (KO) mice that were pretreated with rapamycin, as described in Materials and Methods. The tumor volume at 10 days after A549 transplantation is shown. Other time points are shown in Supplemental Figure S6A. The experiments were independently repeated 4 times (B and C) or 12 times (D). Data are expressed as means ± SD (B–D). ∗P < 0.05, ∗∗P < 0.01. Original magnification, ×400 (A). -, untreated; S, solvent.

Figure 7

Treatment of 9-HODE (H) reduces T and B-regulatory cells (Tregs and Bregs, respectively) of the lymph nodes (LNs) and human lung A549 cancer cell growth in lal^−/− mice. Lymph node cells were isolated from 9-HODE treated or untreated wild-type (WT) and lal^−/− [knockout (KO)] mice and analyzed by flow cytometry. A: The percentages of follicle (Fo) and transitional two-marginal zone precursor (T2-MZP) Bregs, the T2-MZP/Fo ratio, and PD-L1 expression in Fo and T2-MZP Bregs. B: The percentages of Tregs and PD-L1 expression in Tregs. C: Human lung A549 cancer cells were transplanted into wild-type or lal^−/− mice that were pretreated with 9-HODE, as described in Materials and Methods. The tumor volume at 10 days after A549 transplantation is shown. Other time points are shown in Supplemental Figure S6B. The experiments were independently repeated 4 times (A and B) or 12 times (C). Data are expressed as means ± SD (A–C). ∗P < 0.05, ∗∗P < 0.01. -, untreated; EtOH, ethanol.

Figure 8

Blocking PD-L1 reduced T- and B-regulatory cells (Tregs and Bregs, respectively) of the lymph nodes (LNs) and decreased human A549 cancer cell growth in lal^−/− mice. Lymph node cells were isolated from PD-L1 antibody-treated wild-type (WT) and lal^−/− [knockout (KO)] mice and analyzed by flow cytometry. A: The percentages of follicle (Fo) and transitional two-marginal zone precursor (T2-MZP) Bregs and the T2-MZP/Fo ratio. B: The percentages of Tregs. C: Human lung A549 cancer cells were transplanted into wild-type or lal^−/− mice that were pretreated with PD-L1 antibody, as described in Materials and Methods. The tumor volume at 10 days after A549 transplantation is shown. Other time points are shown in Supplemental Figure S6C. The experiments were independently repeated 4 times (A and B) or 12 times (C). Data are expressed as means ± SD (A–C). ∗P < 0.05, ∗∗P < 0.01. -, untreated.