Extended Data Fig. 7: KLF2 plays a key role in entry into quiescence.

a. KLF2 and MYC show negatively correlated changes in expression during the different stages of HIV and T cell life cycle. The values shown are the average expression of KLF2 or MYC in four independent biological replicates (4 differently polarized cells) per stage of HIV life cycle +/- SEM as error bars. The values originate from a bulk RNA-seq study on cells from a healthy donor. For this experiment and panels b and c, p-values were calculated using two-tailed paired Student’s T test. b-e. Expression of MYC and KLF2 show negative correlation in several independently performed high throughput sequencing studies, including HIV infection in SUPT1 cells (b), two primary ex-vivo CD4 + T cell HIV latency model studies (c, d) and a primary CD4 + T cell activation time course study in the absence of HIV (e). Shown values in b reflect average values from three biological replicates +/-SD, while those in c and e correspond to average values from four and two healthy donors, respectively, +/-SD. f. Top 30 genes showing the highest level of differential expression in each direction among all datasets studied in Extended Data Fig. 5g include KLF2 as the most strongly upregulated gene and MYC as one of the most downregulated genes (marked by asterisks). The annotation matrix to the left indicates that many of the shown genes are downstream of MYC, p53 and TNF signaling pathways, the top three enriched pathways among those similarly regulated after HIV infection in all datasets studied in Extended Data Fig. 5g. g. Ex-vivo polarized Th17 cells treated with the quiescence inducing cytokine cocktail (see Fig. 1a) for 12 days as part of the QUECEL model (QUECEL quiescent) express KLF2 at a level significantly greater than recently ex vivo-activated memory cells purified from donors, but less than donor-purified naïve CD4 + T cells and closest to the level observed in donor-purified memory cells, indicating that the KLF2 level in QUECEL models is not outside the physiological range. Data shown reflects average values +/- SD of three independent replicates (two PCR replicates performed with different primer sets, and a third technical replicate made on a different date) performed in cells derived from three independent healthy donors. A linear mixed-effects model was used to account for within-donor correlation by including donor as a random intercept. Log₁₀-transformed KLF2 expression was modeled as a function of experimental condition, with QUECEL quiescent cells specified as the reference group. Compared to QUECEL quiescent cells, memory T cells showed the smallest and least significant difference (β = 0.481, 95% CI [0.321, 0.641], z = 5.90, p = 3.64×10⁻⁹). In contrast, naive T cells exhibited a much larger increase in KLF2 expression (β = 1.011, 95% CI [0.851, 1.171], z = 12.40, p < 1×10⁻³⁵), and restimulated memory T cells showed markedly reduced expression (β = −0.943, 95% CI [ − 1.103, −0.783], z = −11.57, p < 1×10⁻³⁰). These results indicate that among all groups, memory cells were the most similar to QUECEL quiescent cells in KLF2 expression. h. Immunoblotting analysis indicates that transfection with KLF2 targeting siRNAs (siKLF2-1 and siKLF2-2) results in decreased KLF2 protein abundance. Th17 cells were transfected with nontargeting control or the two KLF2-targeting siRNAs one week after TCR activation, then incubated in the quiescence inducing cytokine cocktail and harvested 48 hours later. Molecular weight markers are shown on the left in kilodaltons. i. Densitometry analysis of KLF2 immunoblotting results. Bars represent the mean ± SD of expression levels for each siRNA condition, calculated from two independent healthy donors. The conditions include a non-targeting siRNA and a mixture of two distinct KLF2-targeting siRNAs. For each condition, values from the two donors were first normalized to the corresponding β-actin expression level and then averaged. j. Viability of control and KLF2 knockdown cells, corresponding to those in panel i, assessed using Fixable Viability Dye eFluor™ 450 and flow cytometry performed on the day of cell harvest. k. KLF2 expression is upregulated by more than 20-fold in Th17 cells treated with 10 µM Simvastatin for 24 hours. Values shown correspond to average of two independent technical replicates (two sets of PCR primers) per donor and two biological replicates (two independent healthy donors) +/- SD. A linear mixed-effects model was fit to log10-transformed KLF2 expression levels, with fixed effects for treatment, primer sets used, and their interaction, and a random intercept for donor. Simvastatin treatment significantly increased KLF2 expression (β = 1.36 log10 units, 95% CI [1.18, 1.53], corresponding to an estimated ~22.7-fold increase. Wald z-tests were used for inference. l. Flow cytometry results represented as a histogram demonstrate that at 24 hours post-infection with HIV, cells pre-treated with Simvastatin show strong loss of Ki67 expression, while cells pre-treated with vehicle are still in the early phase of the quiescence transition.

Source data