Volume 23, Issue 9, Pages 1147-1156 (September 2016) Lysosomal Lipases PLRP2 and LPLA2 Process Mycobacterial Multi-acylated Lipids and Generate T Cell Stimulatory Antigens  Martine Gilleron, Marco Lepore, Emilie Layre, Diane Cala-De Paepe, Naila Mebarek, James A. Shayman, Stéphane Canaan, Lucia Mori, Frédéric Carrière, Germain Puzo, Gennaro De Libero  Cell Chemical Biology  Volume 23, Issue 9, Pages 1147-1156 (September 2016) DOI: 10.1016/j.chembiol.2016.07.021 Copyright © 2016 Elsevier Ltd Terms and Conditions

Cell Chemical Biology 2016 23, 1147-1156DOI: (10. 1016/j. chembiol Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 1 Only Tetra-Acylated PIM Stimulate CD1b-Restricted and PIM-Specific T Cells among the Naturally Occurring PIM Species (A) Structure of PIM2 (left) and PIM6 (right). The acylation sites are indicated by R1–R4. (B) Acyl chains present at R1, R2, R3, and R4 positions in naturally and enzymatically generated PIM. (C) Stimulation of PIM-specific and CD1b-restricted T cells by Mo-DCs loaded with purified acyl forms of PIM6 (left panel), of PIM2 (right panel), and with enzymatically generated forms of PIM2 (Enz1-PIM2 and Enz2-PIM2) (right panel). T cell activation is expressed as GM-CSF released in culture supernatants (mean ± SD of duplicate cultures; **p ≤ 0.01, ***p ≤ 0.001, by two-tailed Student's t test). Graphs are representative of three independent experiments. Purity of individual PIM species was assessed by negative-ion mode MALDI-TOF MS analysis (Figure S1). Cell Chemical Biology 2016 23, 1147-1156DOI: (10.1016/j.chembiol.2016.07.021) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 2 In-Vitro-Generated Di-Acylated PIM Stimulate CD1b-Restricted and PIM-Specific T Cells (A) TLC analysis of purified tetra-acylated PIM2 (I, Ac2PIM2) treated or not with recombinant Fusarium solani cutinase. Products II and III were observed after cutinase treatment. The star (*) corresponds to deoxycholate (DOC). (B and D) Negative-ion mode MALDI-TOF MS spectra of Ac2PIM2 and Ac2PIM6 species treated with Fusarium solani cutinase. Ac2PIM2 (I) was digested in Ac1PIM2 (II) by the loss of one C19, and in CutPIM2 (III) by the additional loss of one C16 (B). Similar digestion products were observed for Ac2PIM6 (D). (C and E) 1D 1H NMR analysis of purified di-acylated PIM2 (III in A and B; CutPIM2) and PIM6 (CutPIM6) generated by Fusarium solani cutinase. The insets show the signals of anomeric protons from mannosyl units (H1, Man) and of myo-inositol (H3, Ins). (F and G) Stimulation of PIM-specific and CD1b-restricted T cells by purified di-acylated CutPIM2 and CutPIM6 but not by natural di-acylated PIM2 and PIM6. T cell activation is expressed as GM-CSF released in culture supernatants (mean ± SD of duplicate cultures; **p ≤ 0.01, ***p ≤ 0.001, by two-tailed Student's t test). Graphs are representative of three independent experiments. Purity of CutPIM2, CutPIM6, PIM2, and PIM6 was assessed by negative-ion mode MALDI-TOF MS analysis (Figure S2). Cell Chemical Biology 2016 23, 1147-1156DOI: (10.1016/j.chembiol.2016.07.021) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 3 Inhibitors of Intracellular Lipolytic Enzymes Impair the Capacity of APCs to Present Tetra-Acylated PIM2 to Specific T Cells (A–D) Mo-DCs pre-treated with inhibitors of lipase/PLA1 (orlistat, MmPPOX) and PLA2 (MJ33, amiodarone) were used to stimulate a CD1b-restricted and PIM-specific T cell clone (PIM-specific) and a control iNKT cell clone (iNKT) in the presence of Ac2PIM2 (10 μg/mL) or α-GalCer (50 ng/mL). (E–G) THP-1-CD1b cells pre-incubated or not (no Drug) with orlistat (5 μM), MJ33 (5 μM), amiodarone (5 μM), or MmPPOX (25 μM) were loaded with Ac2PIM2 (10 μg/mL), α-GalCer (50 ng/mL), or Ac2SGL (10 μg/mL) and used to stimulate CD1b-restricted and PIM-specific T cells (PIM-specific), iNKT cells (iNKT), and CD1b-restricted Ac2SGL-specific T cells (Ac2SGL-specific). Ac2SGL is the abbreviation for di-acylated sulfoglycolipids, corresponding to 2-palmitoyl or 2-stearoyl-3-hydroxyphthioceranoyl-2′-sulfate-α-α′-D-trehalose. Data are expressed as % of GM-CSF released by T cells stimulated with drug-treated APCs relative to the amount of GM-CSF produced by the same T cells stimulated with untreated APCs. Absolute values of GM-CSF released by T cells stimulated with untreated Ag-pulsed APCs (control groups) expressed as means ± SD were the following: (1) PIM-specific T cells stimulated with Ac2PIM2-pulsed Mo-DCs, 12.8 ± 0.72 ng/mL, or with Ac2PIM2-pulsed THP-1 cells, 12.4 ± 0.75 ng/mL; (2) iNKT cells stimulated with α-GalCer-pulsed Mo-DCs, 15.4 ± 0.68 ng/mL, or with α-GalCer-pulsed THP-1 cells, 8.6 ± 0.06 ng/mL; (3) Ac2SGL-specific T cells stimulated with Ac2SGL-pulsed THP-1 cells, 13.3 ± 0.4 ng/mL. Each graph represents the cumulative results of three independent experiments in which duplicate cultures were analyzed. Bars represent SD (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, by two-tailed Student's t test). Cell Chemical Biology 2016 23, 1147-1156DOI: (10.1016/j.chembiol.2016.07.021) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 4 Recombinant PLRP2 and LPLA2 Digest Tetra-Acylated PIM2 In Vitro (A) Autoradiography of the TLC of 14C-Ac2PIM2 and 14C-Ac1PIM2 containing fractions untreated (−) or treated with the different recombinant enzymes: LPLA2, LAL, gPLRP2, hPLRP2, cutinase. Lanes 2 (TpLPLA2) and 8 (TpCutinase) correspond to the same fraction incubated only with LPLA2 and cutinase buffers, respectively. The apolar products (Rf 0.84, annotated “released fatty acids*”) correspond to free fatty acids hydrolyzed by all the lipolytic enzymes, except by LPLA2. For LPLA2 digestion, it corresponds to 1-O-acylceramide (see Supplemental Experimental Procedures). In the upper panel, bars represent the percentage of total radioactivity associated with released fatty acids*, Ac2PIM2, and Ac1PIM2 spots. (B) Orcinol staining of the same TLC run. DOC, deoxycholate. Cell Chemical Biology 2016 23, 1147-1156DOI: (10.1016/j.chembiol.2016.07.021) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 5 PLA1 Inhibitors Block the rPLRP2 Hydrolytic Activity on Tetra-Acylated PIM2 In Vitro (A) Orcinol staining of the TLC analysis performed on gPLRP2 digestion of Ac2PIM2 (slightly contaminated by Ac1PIM2) treated with different phospholipase inhibitors (MmPPOX, orlistat, MAFP, amiodarone, and MJ33). Lanes 1 (PIM) and 2 (PIM2 + buffer) correspond to the same fraction of PIM untreated or incubated only with gPLRP2 buffer, respectively. The star (*) corresponds to taurodeoxycholate (NaTDC). (B) Negative-ion mode MALDI-TOF MS analysis of the Ac2PIM2 fraction untreated (up) or hydrolyzed (down) by gPLRP2. (C) Negative-ion mode MALDI-TOF MS analysis of the Ac2PIM2 fraction hydrolyzed by gPLRP2 and treated with the different inhibitors. Cell Chemical Biology 2016 23, 1147-1156DOI: (10.1016/j.chembiol.2016.07.021) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 6 PLRP2 or LPLA2 Gene Silencing Abrogates the Ability of APCs to Present Tetra-Acylated PIM2 to T Cells (A) Assessment of PLRP2 and LPLA2 individual gene silencing in THP-1-CD1b cells upon transfection with specific siRNA. The percentage of expression of the two genes in cells transfected with individual siRNA relative to control cells (Ctrl) is shown. Bars represent the SD of triplicate samples. (B and C) THP-1-CD1b cells not transfected (no siRNA) or individually transfected with siRNA targeting PLRP2 (PLRP2 siRNA) or LPLA2 (LPLA2 siRNA) were used to stimulate CD1b-restricted and PIM-specific (PIM-specific, B) or Ac2SGL-specific (Ac2SGL-specific, C) T cells in the presence or absence (no Ag) of respective antigens (Ac2PIM2 or Ac2SGL, respectively) (10 μg/mL). Release of GM-CSF is expressed as the mean ± SD of triplicate cultures (***p ≤ 0.001, two-tailed Student's t test). Data are representative of four independent experiments. Cell Chemical Biology 2016 23, 1147-1156DOI: (10.1016/j.chembiol.2016.07.021) Copyright © 2016 Elsevier Ltd Terms and Conditions