Expression of immune checkpoint ligands is a mechanism that many tumors use to escape attack by host immune cells. PD-L1, the ligand for checkpoint receptor PD-1 on T cells, is often expressed on tumor cells. Engagement of PD-1 on T cells by PD-L1 on tumor cells attenuates T-cell receptor signaling and suppresses anti-tumor response. PD-1 and PD-L1 blocking antibodies have been implemented clinically as treatment for many cancers, but the pattern of PD-L1 expression on AML is not well characterized. To answer this question, we studied how PD-L1 expression on AML is regulated under in vitro conditions that simulate the leukemia-host microenvironment.

We examined surface expression of PD-L1 by flow cytometry on 4 AML lines, THP-1, KG1, KG1a, HL60, and a CML line, K562. Under basal conditions, these lines expressed no or low levels of PDL1. The AML cells were then subjected to conditions that mimic the leukemia-host microenvironment. AML cells were stained with the green fluorescent dye CFSE and co-cultured with Ficoll-separated PMNCs from healthy donors. After a day of co-culture, expression of PD-L1 was analyzed on AML cells and PD-1, CD25 and CD69 activation markers on PMNCs. Only a small increase of PD-L1, up to 2-4 fold, was seen on AML cells under this condition. To simulate the pro-inflammatory milieu in the tumor microenvironment, anti-CD3/CD28 microbeads were then added in culture to activate T cells. We observed a marked up-regulation of PD-L1 on AML cells, up to 5-60 fold, plus prominent expression of PD-1, CD25 and CD69 on T cells. These findings were confirmed by an alternative method of T cell activation in which AML cells were first coated with an anti-CD123 antibody, linked to anti-CD3/CD28 antibodies via a biotin-streptavidin bridge, and then cultured with PMNCs. To test whether pro-inflammatory cytokines were the sole inducers of PD-L1 expression, AML cells were treated with IFN-γ or TNF-α alone. IFN-γ treatment enhanced PD-L1 expression by 2-10 fold, while TNF-α showed a <2-fold increase. These results show that expression of PD-L1 on AML is dynamically regulated through interaction with activated T cells, by multiple mechanisms including cytokine production and cell-cell interaction.

MYC has been shown to regulate PD-L1 expression on T-ALL and solid tumors (Science 2016; 352:227). We asked whether MYC inhibition would suppress PD-L1 on AML. AML and PMNCs were co-cultured in the presence of anti-CD3/CD28 beads, with JQ1, a BET bromodomain inhibitor that blocks MYC expression. JQ1 inhibited PD-L1 expression by >90%. Dose-effect titration showed sigmoidal curves with ED50 of 0.03 to 0.1 μM for the 5 AML lines. Treatment with another MYC inhibitor, CPI-203, yielded similar results. These observations indicate that MYC inhibition can suppress PD-L1 expression on AML induced by activated T cells. TP53 has been shown to regulate PD-L1 expression on non-small cell lung cancer (JNCI 2016; 108:djv303). We asked whether TP53 activation in AML would also affect PD-L1 expression. Since the AML lines we used did not express wild-type TP53, we overexpressed TP53 in these cells by transfecting with a TP53-GFP plasmid. Expression of TP53 in the cells decreased PD-L1 levels by >80%. Treatment of the cells with pifithrin, an inhibitor that blocks trans-activating function of TP53, did not rescue PD-L1 expression, suggesting that the effect of TP53 on PD-L1 expression is independent of its canonical trans-activating pathway. We asked if MYC and TP53 would synergistically affect PD-L1 expression on AML. We transfected AML cells with the TP53-GFP plasmid and co-cultured the cells with PMNCs and anti-CD3-/CD28 beads, in the presence or absence of JQ1. We found that JQ1 treatment of TP53-transfected cells further decreased PD-L1 expression by another 15%, indicating that MYC and TP53 independently and synergistically affect PD-L1 expression on AML.

In summary, PD-L1 expression on AML cells is dynamically up-regulated upon interaction with activated T cells and suppressed by perturbation of the MYC and TP53 pathways. These findings have implications in the use of immune effector cell therapy against AML, since the activated effector cells could up-regulate PD-L1 expression on target cells and attenuate anti-leukemia effects. MYC inhibitors and TP53 activators could potentially be used in combination to suppress PD-L1 up-regulation and abrogate the ability of AML cells to escape host immune elimination.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution