Kuznetsova et al. · Blood Advances · 2025
Chronic inflammation deters natural killer cell fitness and cytotoxicity in myeloid leukemia
Kuznetsova Valeriya, Krishnan Vaidehi, Costa Amanda, Ren Xi, Ricketts Tiffany D., Patel Sweta B., Connelly Ashley N., Goel Paran, Knapp Joshua P., Franceski Alana M., Luca Francesca, Lobo de Figueiredo-Pontes Lorena, Bhatia Ravi, Prabhakar Shyam, Ong S. Tiong, Welner Robert S.
The study
What was asked, and what was found
Kuznetsova et al., Blood Advances, 2025 asked what a leukemic environment does to natural killer cells, since the number and maturity of those cells at the moment a tyrosine kinase inhibitor is withdrawn is the strongest immune predictor of lasting remission in chronic myeloid leukemia. In a chimeric BCR::ABL1 positive mouse model the host natural killer cells were fewer, held at an immature stage, poor at degranulating against targets, and carrying more inhibitory receptors. Single cell RNA sequencing of sorted natural killer cells put an inflammatory cytokine response at the centre of the problem, with tumour necrosis factor alpha as the predicted driver and the cytokine checkpoint gene Cish among only 13 genes raised consistently across tissues and maturation stages.
To test whether Cish was the brake, the group silenced it with AUMsilence sdASO from AUM BioTech. Natural killer cells sorted from healthy mice were treated with 2 μM of the oligonucleotide for 24 hours against a scrambled control, with no transfection reagent named in the methods, then exposed to serum from leukemic or control mice and put in front of Yac-1 target cells. Degranulation, which leukemic serum had suppressed, was partly restored in the silenced cells. Blocking tumour necrosis factor alpha with etanercept gave a similar partial rescue and also stopped leukemic serum from inducing Cish, which ties the checkpoint to the cytokine.
The same design was then run on human material. Natural killer cells from healthy donors were silenced for CISH with the same reagent and exposed to plasma from newly diagnosed patients who later failed to reach a major molecular response. The silenced cells degranulated better than the scrambled controls. Reading the patient data alongside this, the inflammatory signature found in the mouse cells was enriched specifically in newly diagnosed patients who went on to become resistant to imatinib, and CISH was one of the genes carrying it.
Key findings
- Silencing Cish with AUMsilence sdASO partly gave back the target killing response that leukemic serum had suppressed in primary mouse natural killer cells.(Results, CML suppression of NK cell cytotoxicity is partially restored by targeting cytokine signaling; Figure 4C)
- The same result held in human material: AUMsilence sdASO against CISH improved degranulation in healthy donor natural killer cells that had been exposed to patient plasma.(Results, Human NK cells display similar sensitivity to the CML microenvironment; Figure 5I)
- AUMsilence sdASO reached sorted primary natural killer cells from both species at 2 μM in 24 hours, with no transfection reagent named in the methods.(Methods, section headed “AUMsilence mRNA silencing”)
- The knockdown tested a target the study had picked out itself. Cish was one of only 13 genes raised in leukemia-exposed natural killer cells in both tissues and in both the mature and the immature cluster, and the rise was confirmed by qPCR in each.(Results, and Figure 3F to 3G)
For research use only. Not for use in diagnostic or therapeutic procedures.