SRX14993837 GSM6065441_r1 SRP372178 PRJNA831875 SRS12743170 GSM6065441 GSM6065441 RNA-Seq TRANSCRIPTOMIC cDNA By applying single-cell RNA sequencing, we first compared the genexpression profiles of CAR-engineered NK cells in the presence or absence of supplemental IL-15 signaling in vitro. We subsequently validated our findings in vivo in a non-curative lymphoma model by systematically tracking the evolution of both the tumor and CAR-NK cells collected from liver, spleen, blood, and bone marrow from animals in each of our treatment groups, NT-NK, CAR19 NK and Car19/IL-15 NK cells. Mouse-derived samples were collected serailly at different time point after CAR-NK cell infusion. Cryopreserved tumor and NK cells isolated from mouse tissues were thawed and stained with antibodies against human CD45 (Clone HI30: Biolegend), CD56 (Clone HCD56: Biolegend) and CD20 (Clone 2H7: Biolegend). NK cells (identified as hCD45+CD56+) and tumor cells (identified as hCD45+CD20+) were sorted into separate tubes containing cell culture medium and kept on ice. NK cells and tumor cells were mixed in one tube at 1:1 ratio before generating single cell libraries. Single-cell libraries were prepared using the 10X Genomics Chromium 3' scRNAseq according to manufacturer's protocol. Briefly, single cells were partitioned into Gel beads in Emulsion (GEM) in the 10X Chromium Controller instrument followed by cell lysis, barcoded reverse transcription of RNA, and amplification. Subsequent to 5' adapter ligation, samples were indexed for sequencing. On average, 5000 cells were loaded on each channel, resulting in recovery of an average of 1200 cells/sample. Libraries were sequenced on Illumina (NovaSeq6000, SP flow cell -100 cycle kit per pool), Paired end reads: read 1, 28 bp, Read 2, 91 bp. Illumina NovaSeq 6000 SRX14993838 GSM6065442_r1 SRP372178 PRJNA831875 SRS12743171 GSM6065442 GSM6065442 RNA-Seq TRANSCRIPTOMIC cDNA By applying single-cell RNA sequencing, we first compared the genexpression profiles of CAR-engineered NK cells in the presence or absence of supplemental IL-15 signaling in vitro. We subsequently validated our findings in vivo in a non-curative lymphoma model by systematically tracking the evolution of both the tumor and CAR-NK cells collected from liver, spleen, blood, and bone marrow from animals in each of our treatment groups, NT-NK, CAR19 NK and Car19/IL-15 NK cells. Mouse-derived samples were collected serailly at different time point after CAR-NK cell infusion. Cryopreserved tumor and NK cells isolated from mouse tissues were thawed and stained with antibodies against human CD45 (Clone HI30: Biolegend), CD56 (Clone HCD56: Biolegend) and CD20 (Clone 2H7: Biolegend). NK cells (identified as hCD45+CD56+) and tumor cells (identified as hCD45+CD20+) were sorted into separate tubes containing cell culture medium and kept on ice. NK cells and tumor cells were mixed in one tube at 1:1 ratio before generating single cell libraries. Single-cell libraries were prepared using the 10X Genomics Chromium 3' scRNAseq according to manufacturer's protocol. Briefly, single cells were partitioned into Gel beads in Emulsion (GEM) in the 10X Chromium Controller instrument followed by cell lysis, barcoded reverse transcription of RNA, and amplification. Subsequent to 5' adapter ligation, samples were indexed for sequencing. On average, 5000 cells were loaded on each channel, resulting in recovery of an average of 1200 cells/sample. Libraries were sequenced on Illumina (NovaSeq6000, SP flow cell -100 cycle kit per pool), Paired end reads: read 1, 28 bp, Read 2, 91 bp. Illumina NovaSeq 6000 SRX14993839 GSM6065443_r1 SRP372178 PRJNA831875 SRS12743172 GSM6065443 GSM6065443 RNA-Seq TRANSCRIPTOMIC cDNA By applying single-cell RNA sequencing, we first compared the genexpression profiles of CAR-engineered NK cells in the presence or absence of supplemental IL-15 signaling in vitro. We subsequently validated our findings in vivo in a non-curative lymphoma model by systematically tracking the evolution of both the tumor and CAR-NK cells collected from liver, spleen, blood, and bone marrow from animals in each of our treatment groups, NT-NK, CAR19 NK and Car19/IL-15 NK cells. Mouse-derived samples were collected serailly at different time point after CAR-NK cell infusion. Cryopreserved tumor and NK cells isolated from mouse tissues were thawed and stained with antibodies against human CD45 (Clone HI30: Biolegend), CD56 (Clone HCD56: Biolegend) and CD20 (Clone 2H7: Biolegend). NK cells (identified as hCD45+CD56+) and tumor cells (identified as hCD45+CD20+) were sorted into separate tubes containing cell culture medium and kept on ice. NK cells and tumor cells were mixed in one tube at 1:1 ratio before generating single cell libraries. Single-cell libraries were prepared using the 10X Genomics Chromium 3' scRNAseq according to manufacturer's protocol. Briefly, single cells were partitioned into Gel beads in Emulsion (GEM) in the 10X Chromium Controller instrument followed by cell lysis, barcoded reverse transcription of RNA, and amplification. Subsequent to 5' adapter ligation, samples were indexed for sequencing. On average, 5000 cells were loaded on each channel, resulting in recovery of an average of 1200 cells/sample. Libraries were sequenced on Illumina (NovaSeq6000, SP flow cell -100 cycle kit per pool), Paired end reads: read 1, 28 bp, Read 2, 91 bp. Illumina NovaSeq 6000