Fig 2: Enhanced uptake of oxidized-desialylated LDL by LAK cells. LAK cells were cultured in serum free X-VIVO 10 media in a V-bottom 96 well plate with IL-2 in the absence or presence of native pHrodo Green Conjugate LDL or oxidized-desialylated pHrodo Green Conjugate LDL at 10 µg/ml for 1, 2, 8, 16, 32, and 72 hours. The percentage of LDL positive cells was measured by flow cytometry. A) Qualitative flow data showing uptake of oxidized-desialylated LDL and native LDL, which shows the differences in uptake between the two forms of LDL. FSCH stands for forward scatter cell signal height which facilitates selection of single cells. B) Time course of native LDL and oxidized-desialylated LDL uptake by LAK cells. * Indicates statically significant differences between native LDL (control) and oxidized-desialylated LDL treated LAK cells (p<0.001). Statistical significance determined using multiple t-tests (one per group) and corrected for multiple comparisons using the Holm-Sidak method. n = 3 per time point. Error bars represent standard deviation. For some points, errors bars are shorter than the symbol, and error bars are not shown.

Fig 3: Oxidized-desialylated LDL inhibits LAK cell cytotoxicity in vitro. A) Activated and expanded LAK cells were cultured in serum free X-VIVO 10 media in a 24 well plate with 0.1 µg/ml IL-2 in the absence or presence of native LDL, oxidized only LDL, desialylated only LDL, or oxidized-desialylated LDL at 50 µg/ml for 72 hours. Then LAK cells were washed three times with X-VIVO 10 serum free media to remove residual external LDL and incubated in a 4-hr killing assay with K562 cells at a 10:1 effector to target ratio. Percent cytotoxicity was determined by flow cytometry. B) Quantification of K562 cell death. **indicates a statically significant difference between native LDL (control) and oxidized-desialylated LDL treated LAK cells, and oxidized only LDL vs oxidized-desialylated LDL treated LAK cells (p<0.0001).*Indicates significant difference between native LDL and desialylated only LDL, and desialylated only LDL vs oxidized-desialylated (p <0.001) treated LAK cells, n = 5 per group. ns indicates not statistically significant. Statistical significance determined using two-way ANOVA with Tukey posthoc test. Error bars represent standard deviation.

Fig 4: Gut microbial functional modules with altered abundance in the ATLL patient group.(A) Comparison of KO abundances between patients with ATLL and healthy controls. X-axis, fold change (log2-transformed) in the mean relative abundance of each bacterium in each group; Y-axis, -log10-transformed P values from pairwise comparisons. The horizontal line represents P = 0.05. P values were calculated by the two-tailed Brunner–Munzel test. The plot size shows the KO prevalence (number of subjects) in the ATLL patient group.(B) Enrichment analysis of KOs that were significantly more abundant in patients with ATLL. P values are transformed by -log10. The vertical line represents P = 0.05. ∗, q < 0.05.(C) SSA synthesis module. Abbreviations: HPC, homoprotocatechuate; CHMS, 5-carboxymethyl-2-hydroxymuconic-semialdehyde; CHM, 5-carboxymethyl-2-hydroxymuconate; OPET, 5-oxo-pent-3-ene-1,2,5-tricarboxylate; HHDD, 2-hydroxyhepta-2,4-diene-1,7-dioate; OHED, 2-oxo-hept-3-ene-1,7-dioate; HHED, 2,4-dihydroxy-hept-2-ene-1,7-dioate; SSA, succinic semialdehyde.(D) Association of KO abundance and genus Klebsiella with the SSA synthesis module. The box plot represents the relative abundance of the KOs (left) and ortholog genes assigned to the KO from Klebsiella (right). The bar plot represents the bacterial composition of the KOs of each subject (right). The top two genera with the highest average relative abundance in at least one KO are displayed.(E) (F) Cell growth rate. (E) IL-2-dependent ILT-Mat cells and (F) IL-2-independent MT-1 cells treated with SSA. The values represent the means ± s.d. ∗P < 0.05; unpaired two-tailed Student's t-test.

Fig 5: T cell effector and memory responses of persistent patients(A) Frequencies of CD4+ T lymphocytes producing IFNγ, TNFα and IL-6 and CD8+ T lymphocytes producing IFNγ, TNFα and Granzyme B (Grz B) after polyclonal in vitro stimulation of PBMCs from NI, NP and P (≤10 DSSO) with anti-CD3/CD28 beads of PBMCs from (n = 12), NP (n = 11) and P (n = 23).(B) Longitudinal immunophenotyping of PBMC from NP and P depicting CD4+CD27+CD45RA+CCR7+ naive T cells (naive CD4), CD4+CD27+CD45RA−CCR7- effector memory T cells (EM1) and CD8+CD27+CD45RA + −CCR7+ naive T cells (naive CD8). Filled dots represent individual samples longitudinally collected until resolution of infection from NP and P patients positive by qRT-PCR for SARS-CoV-2. Empty dots represent samples from convalescent patients coinciding with the first negative qRT-PCR for SARS-CoV-2.(C and D) Immunophenotyping of SARS-CoV-2 reactive CD4 and CD8 T cells after in vitro stimulation of PBMC from NP (n = 11) and P (n = 9) after viral clearance with peptides spanning the Spike protein of the alpha variant of SARS-CoV-2 in the presence of IL-2. Percentage of (C) CD4+ and CD8+ cells expressing Ki67, (D) CD4+ cells expressing TNFα, CD8+ cells expressing IFNγ and Granzyme B (Grz B), and CD4+CD25+ cells expressing IL-10. Statistical significance was calculated using Kruskal-Wallis analysis followed by Dunn post-test, and indicated by ∗p ≤ 0.05; ∗∗p ≤ 0.01. DSSO, Days since symptom onset.