Integrin-deficient T-ALL accumulates in the CNS compared to WT T-ALL.

(A-D) WT mCherry+ T-ALL and DKO GFP+ T-ALL cells were mixed 1:1 and transferred intravenously (i.v.) into sub-lethally irradiated recipients. (A) Experiment design. (B) Left: Representative flow cytometry plot distinguishing WT T-ALL and DKO T-ALL from the CNS. Right: Representative surface stain of β1 and αL integrins on T-ALL. Black: WT, red: DKO, grey: isotype. Numbers (C) and ratios (D) of DKO (red) and WT (black) T-ALL cells found in the indicated tissues at a humane endpoint. Data compiled from 3 experiments; n = 15. (E-F) DKO T-ALL cells were transduced with either vectors encoding Itgal and Itgb1 (“rescue”) or empty vector (“empty”) and transferred 1:1 i.v. into sub-lethally irradiated recipients. (E) Experiment design. (F) Ratio of DKO-empty to DKO-rescue T-ALL cells in the indicated tissues at a humane endpoint. Data compiled from two experiments; n = 10. (G-H) WT T-ALL cells were nucleofected with sgRNAs targeting Itgal or non-targeting gRNAs along with recombinant Cas9 protein. The process was repeated for Itgb1 to generate “CRISPR-DKO” and “CRISPR-WT T-ALL.” The cells were mixed 1:1 and transferred i.v. into sub-lethally irradiated recipients. (G) Experiment design. (H) Ratio of CRISPR-DKO to CRISPR-WT T-ALL cells found in the indicated tissues at a humane endpoint. Data compiled from five experiments; n = 15.

Integrin-deficient T-ALL is unlikely to accumulate in the CNS due to enhanced entry, limited exit, or superior immune escape.

(A, B) Sub-lethally irradiated WT mice were injected i.v. with 1:1 DKO:WT T-ALL and tissues were isolated from 3 mice every 3-4 days. (A) Total number of WT (black) and DKO (red) T-ALL cells in the CNS. (B) Ratio of DKO:WT T-ALL in the CNS. Data compiled from two experiments; n = 36. (C) Left, CRISPR-TKO T-ALL generation. Right, representative surface staining of integrin β7 on T-ALL cells (blue: TKO, black: WT, red: DKO). (D) Ratio of DKO:WT T-ALL (black) or TKO:WT T-ALL (blue) in indicated tissues at a humane endpoint. Data compiled from two experiments; n = 5. (E, F) WT mice were infected with AAV encoding a VEGF-trap or control AAV. 5 weeks later, mice were sub-lethally irradiated and injected i.v. with 1:1 DKO:WT T-ALL. At a humane endpoint, tissues were isolated and T-ALL was enumerated. (E) Experiment design. (F) Number of WT (black) and DKO (red) T-ALL or ratio of DKO:WT T-ALL in the CNS of mice treated with either control AAV or AAV-VEGF-trap. Data compiled from 2 experiments; n = 6. (G, H) WT mice were surgically implanted with an intra-cerebroventricular cannula. 1 week later, mice were sub-lethally irradiated and injected i.v. with 1:1 DKO:WT T-ALL. At mid-stage disease, a reactive label (Atto647-NHS-ester) was infused through the cannula into the left lateral ventricle. 48h later, tissues were isolated and labelled T-ALL cells were enumerated. (G) Experiment design. (H) Representative flow cytometry plots of Atto647 labelling in the indicated tissues. WT T-ALL, black; DKO T-ALL, red; other leukocytes, blue. Representative of 3 experiments; n = 6. (I) NRG mice or WT controls (B6 or NOD) were sub-lethally irradiated, injected i.v. with 1:1 DKO:WT T-ALL, and analyzed at a humane endpoint. Ratio of DKO:WT T-ALL cells in controls (black; n = 3) or NRG mice (blue; n = 8). Data compiled from 2 experiments. (J) WT mice were fed either diet supplemented with a CSF1R inhibitor (PLX5622) or control diet for 1 week, sub-lethally irradiated and injected i.v. with 1:1 DKO:WT T-ALL, and maintained on either PLX5622 or control diet until a humane endpoint. Ratio of DKO:WT T-ALL in the indicated tissues of mice fed control diet (black; n = 10) or PLX5622 diet (blue; n = 8). Data compiled from 2 experiments. Bars represent geometric mean and geometric SD. Statistics calculated on log-transformed data using unpaired t-test with Welch’s correction. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001

Integrin-deficient T-ALL proliferates more than WT T-ALL in the CNS.

(A) Representative (of n = 12) immunofluorescence image of whole-mounted dural meninges of mice injected i.v. with 1:1 DKO:WT T-ALL. Mice were analyzed at mid-stage disease. Red, WT T-ALL; green, DKO T-ALL; scale bar, 200µm. (B) Representative flow cytometry gating to identify proliferating cells isolated from tissues of mice that received EdU 4 hours before analysis (in two intraperitoneal doses, 100 µg at -4h and 100 µg at -2h). (C, D) Sub-lethally irradiated WT mice received 1:1 DKO:WT T-ALL cells i.v., and were analyzed at mid-stage disease. (C) The percent EdU+ of WT (black) or of DKO (red) T-ALL cells and (D) the ratio of % EdU+ DKO: % EdU+ WT T-ALL cells in the indicated tissues. Data compiled from 3 experiments; n = 9. (E, F) As in (C, D) but for the CRISPR-WT and CRISPR-DKO pair of T-ALL lines. Data compiled from 4 experiments; n = 8. For (C, E), statistics comparing WT and DKO within a tissue were calculated on logit-transformed data using paired t-test. Statistics comparing WT or DKO from CNS to spleen were calculated on logit-transformed data using unpaired t-test with Welch’s correction. For (D, F), statistics calculated using unpaired t-test with Welch’s correction. Bars represent geometric mean and geometric SD. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001

Integrin-deficient T-ALL is more susceptible to chemotherapy targeting dividing cells and integrin blockade synergizes with chemotherapy to limit CNS T-ALL

(A-C) Sub-lethally irradiated WT mice received 1:1 DKO:WT T-ALL i.v., were treated daily i.p. with 40mg/kg 5-fluorouracil (5-FU) for 3 days at mid-stage disease (starting 12-14 days after T-ALL transfer), and were analyzed 24h after the last dose of 5-FU. (A) Experiment design. (B) The ratio of DKO:WT T-ALL cells in vehicle-treated (black; n = 10) or 5FU-treated (blue; n = 10) mice. Data compiled from 2 experiments. (C) As in (B) but for the CRISPR-WT and CRISPR-DKO pair of T-ALL lines. Vehicle-treated (black; n = 8), 5FU-treated (blue; n = 8). Data compiled from 2 experiments. (D-F) Mice were surgically implanted with a cannula in the left lateral ventricle. At least 7 days later, they were sub-lethally irradiated and injected i.v. with WT T-ALL. At mid-stage disease (12-14 days after T-ALL transfer), mice were injected i.p. with 40mg/kg 5-FU daily for three days. On days 1 and 3 of 5-FU treatment, mice received 5 microliter intraventricular infusions of isotype (47 microgram) or anti-LFA-1 and anti-VLA-4 (26 micrograms each) blocking antibodies. 24h after the last dose, tissues were isolated and T-ALL cells were enumerated. (D) Experiment design. (E) Surface expression of α4 and αL integrins on WT T-ALL from the CNS after intra-cerebroventricular injection of LFA-1- and VLA-4- blocking antibodies (blue; representative of n = 7) or isotype (black; representative of n = 7). Representative of 4 experiments. (F) Total number of WT T-ALL in the CNS normalized to the average isotype control per experiment. Data compiled from 4 experiments; n = 7 pairs. Bars represent geometric mean and geometric SD. Statistics calculated on log-transformed data using unpaired t-test with Welch’s correction. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001