1. Volume 44, Issue 6, Pages 1312-1324 (June 2016)
Memory CD8+ T Cells Require Increased Concentrations of Acetate Induced by Stress for Optimal Function 
Maria L. Balmer, Eric H. Ma, Glenn R. Bantug, Jasmin Grählert, Simona Pfister, Timo Glatter, Annaïse Jauch, Sarah Dimeloe, Emma Slack, Philippe Dehio, Magdalena A. Krzyzaniak, Carolyn G. King, Anne-Valérie Burgener, Marco Fischer, Leyla Develioglu, Réka Belle, Mike Recher, Weldy V. Bonilla, Andrew J. Macpherson, Siegfried Hapfelmeier, Russell G. Jones, Christoph Hess 
Immunity 
Volume 44, Issue 6, Pages (June 2016)
DOI: /j.immuni
Copyright © 2016 Elsevier Inc. Terms and Conditions

2. Immunity 2016 44, 1312-1324DOI: (10.1016/j.immuni.2016.03.016)
Copyright © 2016 Elsevier Inc. Terms and Conditions

3. Figure 1 Acetate Is a Systemic Acute Phase Metabolite
(A) Serum acetate concentrations in germ-free mice (GF), germ-free mice colonized for 28 days with E. coli, S. xylosus, and E. faecalis (Triple), mice harboring a low-complexity microbiota (LCM), conventional SPF-mice (SPF), healthy human blood donors (Hum), and acetate concentrations in standard cell culture medium (RPMI 1640, DMEM, and unbuffered RPMI 1640, Med) with and without 10% fetal calf serum (FCS). Dashed line represents lower limit of detection.
(B) Serum acetate concentrations following i.v. infection with 5,000 CFU Listeria monocytogenes.
(C) Serum acetate concentrations in SPF mice 4 hr (blue dots) and 24 hr (red triangles) after Listeria monocytogenes-infection (105 CFU), normalized to uninfected controls (black circles).
(D) Serum acetate concentrations in germ-free mice 48 hr after oral infection with 102 CFU Salmonella typhimurium UK1 (red triangles), normalized to the mean of uninfected controls (black circles).
(E) Serum acetate concentrations in germ-free mice 3 hr (blue dots) and 6 hr (red triangles) after i.v. infection with 107 E. coli JM83, normalized to baseline levels before infection (black circles).
Each dot represents one mouse or human, bars indicate means of pooled data. One-way ANOVA (B and C), unpaired t test (D), and Friedman test (E) were used to compare groups. ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < See also Figure S1.
Immunity  , DOI: ( /j.immuni )
Copyright © 2016 Elsevier Inc. Terms and Conditions

4. Figure 2
Acetate Modulates the Metabolic Signature of CD8+ T Cells and Boosts Rapid Recall Responses
(A) Experimental set-up of in vitro memory differentiation. OVA-specific OT-I memory T cells were activated with 10−9 M OVA peptide in the presence of splenocytes for 3 days, followed by 3–4 day culture in IL-15 (10 ng/mL) in absence of OVA peptide. Acetate was added to the cell culture medium as indicated.
(B and C) Percentage of IFN-γ-positive OT-I memory T cells 4 hr after re-stimulation with medium control (Ctr), or 10 μM of the altered peptide ligands (APLs) G4 and R7, or OVA peptide, determined by intracellular cytokine staining (ICS). Memory cells were generated in vitro without sodium acetate (filled symbols), or in the presence of 5 mM sodium acetate (open symbols). (B) shows memory induction without sodium acetate (black dots) or 7 day acetate exposure (exposure throughout the in vitro memory induction; open circles), and (C) shows acetate exposure for 24 hr (orange circles), or 3 days (blue triangles) prior to re-stimulation.
(D) Abundance of IFN-γ protein in corresponding cell culture supernatants (left panel), and Ifng mRNA (right panel) of the cells shown in (C). IFN-γ was measured by CBA and RT PCR, respectively.
(E) Percentage of IFN-γ positive OT-I memory T cells 4 hr after re-stimulation with medium control (Ctr), 10 μM of APLs G4 and R7, or OVA peptide, determined by ICS. Memory cells were generated in vitro in the presence or absence of the indicated concentrations of sodium acetate for 3 days.
(F) MACS-purified OT-I memory T cells differentiated without acetate (CD8), and freshly isolated wild-type splenocytes (APCs) were each incubated for 24 hr in control (Ctr) or 5 mM acetate medium (Acetate). APCs and OT-I memory T cells were then co-cultured in a ratio of 1:1 prior to re-stimulation with OVA peptide and APLs as described in (B). IFN-γ production was assessed by ICS.
(G) Representative histograms as well as mean + SD. MFI, or percentage of positivity, of surface marker expression of control (black, filled gray) and 3 day acetate-exposed (blue) OT-I memory T cells assessed by flow cytometry. The dashed line represents the isotype control.
(H) Left panel: Representative mitochondrial perturbation assay of control (black dots) and 3 day acetate-exposed (open circles) OT-I memory T cells. Extracellular acidification rate (ECAR) was measured “in Seahorse” after injection of oligomycin (Oligo), FCCP, and Rotenone (Rot). Shown are means and SEM of quadruplicates. Right panel: Glycolytic reserve (mpH/min) was determined by calculating the ratio between the maximal ECAR and the baseline ECAR of control (black dots), and 24 hr (orange circles) or 3 day (blue triangles) acetate-exposed OT-I memory T cells.
(I) Left panel: Representative ECAR-measurements of control (filled symbols) and 3 day acetate-exposed (open-symbols) OT-I memory T cells, after injection of medium control (black) or 10 μM OVA peptide (green) directly into metabolic flux analyzer (dashed line). Right panel: Glycolytic switch (mpH/min) was calculated by subtracting maximal ECAR from baseline ECAR-measurements in control (black dots) and 3 day acetate-exposed (blue triangles) OT-I memory T cells.
(J) Percentage of IFN-γ positive OT-I memory T cells 4 hr after re-stimulation with medium control (Ctr), or 10 μM of the APLs G4 and R7, or OVA peptide, determined by ICS. Cells were cultured in control medium (black dots) or acetate medium (5 mM) (filled blue triangles) for 3 days, and then switched back to control medium (open blue triangles) for 24 hr.
Each dot represents data obtained from cells isolated from one mouse, and bars indicate means of pooled data. Two-way ANOVA (B), (C), (D) left panel, (E), (F), and (J), one-way ANOVA (D) right panel, (H), and unpaired t test (I) were used to compare groups. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < See also Figure S2.
Immunity  , DOI: ( /j.immuni )
Copyright © 2016 Elsevier Inc. Terms and Conditions

5. Figure 3
Acetate Is Taken Up by OT-I Memory T Cells and Contributes to the Cellular Acetyl-CoA Pool
(A) Control (black dots) and 3 day acetate-exposed (blue triangles) OT-I memory T cells were incubated with 3H-acetate or medium control for 6 min, and β-radioactivity in cell lysates measured by liquid scintillation counting.
(B) Representative OCR (left panel) and ECAR (right panel) measurements after “in Seahorse” injection of 5 mM acetate (blue) or medium control (black), in control (filled symbols) or 3 day acetate-exposed (open symbols) OT-I memory T cells.
(C) Acetyl-CoA concentration in whole cell lysates of control (black dots), or 24 hr (orange circles) and 3 day (blue triangles) acetate-exposed OT-I memory T cells, as determined using a fluorimetric assay.
(D) Metabolic tracing analysis of OT-I memory T cells after 6 hr exposure to 13C-glucose (open bars) or unlabeled glucose plus 13C-acetate (blue bars). The x axis shows the number of 13C per respective metabolite. Depicted are pooled data from two independent experiments with cells from n = 3 mice each.
(E) Percentage of IFN-γ positive OT-I memory T cells 4 hr after re-stimulation with medium control (Ctr), 10 μM of the APLs G4 and R7, or OVA peptide, as determined by ICS. The ATP-citrate lyase inhibitor SB was added during the 3 day acetate exposure in the indicated concentrations (open symbols).
(F) Heatmap (left panel) and MA-plot (right panel) of the control and 3 day acetate-exposed OT-I memory T cell transcriptome. Triplicates each were used to compare gene-expression profiles between groups.
(A, C, and E): Each dot represents data obtained from cells isolated from one mouse, and bars indicate means of pooled data. Two-way ANOVA (A), (D), and (E), and one-way ANOVA (C) were used to compare groups. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < See also Figure S3.
Immunity  , DOI: ( /j.immuni )
Copyright © 2016 Elsevier Inc. Terms and Conditions

6. Figure 4 Acetate Assimilation Drives Acetylation of GAPDH
(A) Representative Western blot from whole-protein extracts of control (Ctr, black dots) and 3 day acetate-exposed (Ac, blue triangles) OT-I memory T cells, probed for GAPDH and Actin. GAPDH protein concentration was quantified and normalized to Actin (middle). Gapdh mRNA normalized to 18S and control memory cells, determined by RT PCR, are shown in the right panel.
(B) GAPDH activity measured in cell lysates of control (Ctr, black dots) and 3 day acetate-exposed (Ac, blue triangles) OT-I memory T cells in the presence (open symbols) or absence (filled symbols) of 5 μM garcinol for 3 days.
(C) Western blot analysis of whole-protein extracts of control and 3 day acetate-exposed OT-I memory T cells probed for acetylated-lysine. Left panel: representative blot; right panel: pooled data (n = 9 independent experiments) (Ctr, black dots; Ac, blue triangles). Bands were quantified and normalized to Actin and fold-change over control OT-I memory T cells determined.
(D) GAPDH was immunoprecipitated in n = 3 control and n = 3 OT-I memory T cells exposed to acetate for 3 days, each. The figure represents detection of an acetylated peptide (K217) (green) in GAPDH among acetate-exposed but not control OT-I memory T cells. Yellow indicates coverage.
(E) GAPDH activity in mouse fibroblasts transfected with wild-type (blue triangles) or K→A 217 mutated (black dots) GAPDH for 48–96 hr. To test for overall acetylation-dependency of GAPDH activity, we treated cell lysates with the bacterial de-acetylase CobB (open symbols) or buffer control (filled symbols) for 1 hr.
Each dot represents data obtained from cells isolated from one mouse, bars indicate means of pooled data. One-way ANOVA (B and E), unpaired t test (A, protein), and Wilcoxon matched-pairs signed rank test (A, mRNA) and (C) were used to compare groups. ∗∗p < 0.01, ∗∗∗∗p < , ns = not significant. See also Figure S4.
Immunity  , DOI: ( /j.immuni )
Copyright © 2016 Elsevier Inc. Terms and Conditions

7. Figure 5
Increased Acetylation of GAPDH Promotes Glycolysis-Driven Recall Responses
(A–C) Percentage of IFN-γ positive OT-I memory T cells 4 hr after re-stimulation with medium control, or 10 μM of the APLs G4 and R7, or OVA peptide, as determined by ICS. Memory cells were generated in vitro without sodium acetate (filled dots) or in the presence of 5 mM sodium acetate for 3 days (blue triangles) prior to re-stimulation. Open symbols indicate re-stimulation in glucose-free medium (A), or the presence of the GAPDH-inhibitor heptelidic acid (HA) (B), or the acetyltransferase-inhibitor garcinol (C), for 3 days each. The right panel in (A) and (C) shows IFN-γ protein in supernatants of medium control (Ctr) and OVA-stimulated (OVA) cells corresponding to the left panels.
(D) Representative “in Seahorse” ECAR measurements of control (black) and 3 day acetate-exposed (blue) OT-I memory T cells in presence (open symbols) or absence (filled-symbols) of garcinol for 3 days prior to analysis, after injection of 10 μM OVA peptide directly into the metabolic flux analyzer (dashed line).
(E) Glucose-uptake assessed by fluorescent 2-NBDG in PBS- (black) or acetate-exposed (blue) CD44+ (left panel) and CD44– (right panel) CD8+ T cells after re-activation with anti-CD3/CD28 mAb for the indicated time.
(F) Intracellular IFN-γ staining of CD44+ CD8+ memory T cells from mice treated with PBS (black) or acetate (blue), restimulated with anti-CD3/CD28 mAb for the indicated time.
(G) Frequency of IFN-γ producing CD44+ CD8+ T cells (shown in F), within the CD62L+ (left panel) or CD62L– (right panel) gate.
(H) Intracellular IFN-γ staining of PBS- (black) or acetate-exposed (blue) CD44+ CD8+ memory T cells after re-activation with np118 peptide or anti-CD3/CD28 mAb for the indicated time.
(I) Intracellular IFN-γ staining of FACS-sorted human EM CD8+ T cells cultured in control medium (black) or acetate medium (5 mM) (blue) for 3 days, prior to activation with anti-CD3/CD28 mAb for 4 hr.
(J) Serum IFN-γ concentration 4 hr after infection with 1 × 105 CFU LmOVA in mice adoptively transferred with 5 × 105 in vitro generated control (black dots) or 3 d acetate-exposed (blue triangles) OT-I memory T cells 24 hr prior to infection.
(K) Listeria monocytogenes colony forming units (CFU) in the liver and spleen 24 hr post-LmOVA infection, after adoptive transfer with 105 (liver) or 5 × 105 (spleen) in vitro generated control (black dots), or 3 day acetate-exposed (blue triangles) OT-I memory T cells 24 hr prior to infection with 105 CFU LmOVA. Dashed line shows lower limit of detection.
(A–C) Each dot represents data obtained from cells isolated from one mouse, and bars indicate means of pooled data. (J and K) Each dot represents an individual mouse. Two-way ANOVA (A)–(C), (E)–(I), Mann-Whitney test (J) and unpaired t test (K) were used to compare groups. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < See also Figure S5.
Immunity  , DOI: ( /j.immuni )
Copyright © 2016 Elsevier Inc. Terms and Conditions