For virus production, Phoenix E ecotropic packaging cells (a kind gift from G. Nolan, Stanford, CA) were transiently transfected with the retroviral construct MSCV-STOP-NPM-ALK-IRES-EGFP (MSNAIE), Mig^NPM-ALK, pBABE-puroR^Nipa, and viral supernatants were collected as described previously (14, 17, 18, 37). Retroviral titers were determined by the transduction of NIH/3T3 cells (DSMZ) as described previously (17).

The assays were performed as previously described (36). In brief, retrovirally infected Ba/F3 or Karpas299 cells express NPM-ALK and miR^ctrl [as previously described (38)] or miR^NIPA (5’-TGC TGT TGA CAG TGA GCG CTC CAT TGG AAT CCA CAA GCA ATA GTG AAG CCA CAG ATG TAT TGC TTG TGG ATT CCA ATG GAA TGC CTA CTG CCT CGG A-3’) were plated on 96-well plates in triplicates (5,000 cells in 100 µl of RPMI). To assess cell proliferation, MTS reagent (Promega, Madison, WI, USA) was added to the cells at the indicated time points and incubated for 2 h at 37°C. Extinction at 492 nm was measured using a microplate reader (Tecan, Männedorf, Switzerland).

For soft agar proliferation assays, we prepared primary Nipa^+/+ or Nipa^−/− MEFs from embryos (E13.5) and cultured them in DMEM (PAA Laboratories) supplemented with 15% FCS under low oxygen conditions. Early passages only were used for the indicated experiments. Nipa^ko/ko MEFs were retrovirally infected with vectors containing NPM-ALK and Flag-NIPA wt. The assay was performed as previously described (39). A total of 25,000 and 100,000 cells were plated in soft agar in 6-well plates. Colonies were counted between days 15 and 20 after plating.

For cell cycle measurements, EdU along with FxCycleViolet, was used according to the instructions of the manufacturer.

Immunoblotting was performed as described previously (33, 34). Antibodies against βACTIN (A5316) and NIPA (ZC3HC1, HPA024023) were purchased from Sigma. ALK (cs-3333) was purchased from Cell Signaling, GAPDH (OSG-00033G) from Osenses. Quantification of immunoblots was performed using LabImage 1D L340 software (Intas Science Imaging, Goettingen, Germany).

Recombinant murine Interleukin-3 (IL-3), IL-6, and SCF were purchased from R&D Systems (Minneapolis, MN, USA). Fetal calf serum, 5-Flouorouracil, and Polybrene were purchased from Sigma-Aldrich (St. Louis, MO, USA). DMEM (Dulbecco’s Modified Eagle Medium) and ES cell FBS (fetal bovine serum) were purchased from Thermo Scientific (Waltham, USA). Lipofectamine 2000 was purchased from Invitrogen (Carlsbad, CA, USA). TAE-684 was purchased from Axon Medchem (Groningen, NL).

The Lck-Cre mouse line (B6.Cg-Tg(Lck-cre)548Jxm/J) was obtained from the Jackson Laboratory (Bar Harbor, Maine, USA). Nipa ^flox/flox mice were generated using a conditional knockout strategy (34). To achieve tissue-specific Nipa deletion, we crossed Nipa ^flox/flox mice with Lck-Cre transgenic mice. Littermates or age- and sex-matched mice were used as controls. All mice were backcrossed to a C57BL/6 background for more than ten generations. The animals were housed in a special cage system with autoclaved food and acidified water at the University of Freiburg. All procedures were performed in accordance with national and institutional guidelines for animal care and experiments.

Murine bone marrow was collected from Lck-Cre wildtype Nipa ^ko/ko and Nipa ^wt/wt or Lck transgenic Nipa ^flox/flox and Nipa ^wt/wt mice and infected as described previously (14, 38, 40). Briefly, 12-week-old male donor mice were treated once with 5-Fluorouracil (150 mg/kg) on day - 4 and BM cells were harvested from the tibia and femur. After preincubation overnight in BM media (DMEM, 30% FBS, 10 ng/ml mIL-3, 10 ng/ml mIL-6, and 50 ng/ml mSCF), BM cells were infected with retroviral supernatant as described previously (41). The infection efficiency was determined by flow cytometric analysis of EGFP expression. Female C57BL/6 wild-type recipient mice were irradiated with 850 rad and transplanted with the indicated cells via the tail vein injection. Peripheral blood was taken at the indicated time points and WBCs were measured using an automated counter (ABC scil vet). Transplanted mice were monitored for signs of disease and sacrificed and analyzed based on clinical signs.

Flow cytometry analysis was performed as described (35). The BD LSR Fortessa (BD Biosciences, Heidelberg, Germany) was used for analysis. Antibodies used to stain cell surface proteins were anti-mouse CD4 (GK1.5), CD8a (53-6.7), CD11b (Mac-1, M1/70), CD25 (PC61.5), CD44 (IM7), CD45R/B220 (RA3-6B2), CD45 (30-F11), CD90.2 (THY1.2, 53-2.1), CD117 (c-KIT, 2B8), CD127 (IL-7Ra, A7R34), GR1 (Ly-6G, RB6-8C5), SCA1 (D7), and TER119 (TER119) and the corresponding isotypes, which were obtained from BD Biosciences or eBiosciences (Frankfurt am Main, Germany).

A two-sided Student’s t-test was used for statistical analyses. The mean ± standard deviation was analyzed as indicated. The survival curves were produced using a log-rank (Mantel–Cox) test. P-values were defined as *p <0.05, **p <0.01, ***p <0.001, and ****p <0.0001.

To analyze the impact of NIPA on NPM-ALK mediated transformation and colony formation, we performed in vitro soft agar assays. Nipa ^ko/ko mouse embryonic fibroblasts (MEFs) were retrovirally infected with NPM-ALK and NIPA or empty vector control. As seen in Figure 1A , Nipa-deficient MEFs infected with NPM-ALK showed significantly lower numbers and smaller sizes of colonies in soft agar assays than NIPA-re-expressing controls (28.9 colony forming units (CFUs) vs. 58.8 CFUs; p = 0.008). Western blotting revealed the correct expression of NPM-ALK and NIPA ( Figure 1B ). In the absence of NPM-ALK, no colony growth was observed in either group, suggesting that Nipa deficiency alone has no transformative potential in MEFs in soft agar assays.

Using targeted genetic approaches, an efficient and durable NIPA knockdown of more than 70% was achieved in the murine IL-3 dependent Ba/F3 cell line and the human ALCL cell line Karpas299 ( Figure 1C ). Upon efficient NIPA downregulation, Ba/F3 cells were retrovirally infected with NPM-ALK. Proliferation was assessed by MTS assays under IL-3 withdrawal, where the optical density (OD) reflects the metabolization of MTS reagent and thus the number of viable cells present. NIPA knockdown significantly impaired the proliferation of NPM-ALK-positive Ba/F3 and Karpas299 cells. Within 24 h, the number of viable NPM-ALK-positive Ba/F3 grew 4.2 fold, whereas only 3.1 fold in NIPA knockdown cells ( Figure 1D ). Two-dimensional cell cycle analyses revealed no differences in the cell cycle profile of NPM-ALK-positive cells upon NIPA knockdown, pointing to a cell-cycle-independent function of NIPA in dependence of NPM-ALK ( Supplementary Figures 1A, B ).

To expand our data to human cells, we designed NIPA siRNAs targeting the human NIPA mRNA and downregulated NIPA in the human ALCL cell line Karpas299. Effective NIPA downregulation resulted in a significantly reduced proliferation (75% compared to controls) of ALCL cells measured at numerous time points after seeding ( Figure 1E ). Interestingly, NIPA downregulation in Karpas299 cells showed a significantly higher susceptibility to the ALK inhibitor TAE-684 ( Figure 1F ), suggesting a possible synergistic effect of ALK inhibition and NIPA knockdown.

To analyze the impact of NIPA on NPM-ALK-induced tumorigenesis, we used a retroviral murine BM transplantation model for NPM-ALK-driven malignancies. As previously demonstrated by Miething et al. (17), transplantation of NPM-ALK-positive bone marrow cells (BMCs) in lethally irradiated recipient mice leads to two distinct phenotypes (polyclonal histiocytic malignancy vs. monoclonal B-lymphoid tumors), depending on disease latency. We performed analogous transplantation experiments using Nipa ^ko/ko and Nipa ^wt/wt donor BMCs. For the short latency model, mice were given 300,000 BMCs with 3.0% NPM-ALK (EGFP) positive cells. For the long latency model, 200,000 cells with 0.4% NPM-ALK (EGFP) positive cells were injected. Independent of the model used, mice transplanted with Mig^NPM-ALK Nipa ^{ko / ko} BMCs showed significantly prolonged survival with 28.5 vs. 27 days (p = 0.03) in short latency and 118 vs. 84 days (p = 0.008) in the long latency model, respectively ( Figures 2A, B ).

In the short latency model, the progression of the disease was furthermore assessed by the number of spleen colonies. Animals transplanted with Mig^NPM-ALK Nipa ^ko/ko bone marrow were found to have a significantly lower number of spleen colonies than controls, with 10 colonies per spleen vs. 28 in controls (p <0.001) ( Figure 2C ). However, the disease immunophenotype was not altered by the absence of NIPA in both the long and short latency models (data not shown). Taken together, our results highlight the crucial role of NIPA in NPM-ALK-driven tumorigenesis.

Based on these results, we hypothesized that NIPA influences NPM-ALK-driven transformation in an ALCL mouse model resembling the human clinical phenotype. This ALCL-like model is based on a lineage-specific Cre/LoxP-dependent expression of NPM-ALK by the retroviral construct MSCV-STOP-NPM-ALK-IRES-EGFP (MSNAIE) ( Supplementary Figure 2A ). Infection of LckCre^TG/wt BMCs with MSNAIE retrovirus and transplantation into lethally irradiated wild-type recipient mice leads to a systemic CD30-positive ALCL-like T-cell lymphoma (14, 18).

To establish a Nipa-deficient ALCL-like disease without “off-target” effects of Nipa deficiency, we used donor BMCs from LckCre^TG/wtNipa^{flox /flox} mice for MSNAIE infection, thus restricting NPM-ALK expression and Nipa deletion to the identical T cells. Transplanted mice were monitored for clinical signs of disease, such as wasting, tachydyspnea, lymphadenopathy, and changes in the complete blood count. LckCre^TG/wtNipa^flox/flox MSNAIE transplanted mice showed a later onset of disease with significantly prolonged survival. The median survival was significantly shorter (p = 0.002) with 121 days in LckCre^TG/wtNipa^wt/wt MSNAIE transplanted mice compared to 143 days in LckCre^TG/wtNipa^flox/flox ( Figure 3A ). At the final stage of the disease, mice were sacrificed and their organs harvested. Diseased mice presented with enlarged thymi of 470 mg on average, mediastinal and inguinal lymphadenopathy, and moderate splenomegaly independent of NIPA status ( Figure 3B and Supplementary Figure 2B ). Both bone marrow infiltration and leukocytosis were heterogeneous at comparable levels in both Nipa ^wt/wt and Nipa ^flox/flox transplanted mice ( Supplementary Figures 2C, D ). Correct Nipa deletion in lymphoma cells was validated by PCR analysis ( Figure 3C ) and lymphomas of LckCre^TG/wtNipa^{flox /flox} MSNAIE transplanted recipients were further referred to as Nipa ^ko/ko. As has previously been reported for wild-type lymphomas, immunophenotyping of Nipa ^ko/ko lymphoma cells showed a pure T-cell phenotype with negativity for myeloid and B-cell markers in the thymus, spleen, lymph nodes, peripheral blood, and bone marrow ( Figure 3D ) (Kreutmair et al., 2020a; Shoumariyeh et al., 2020). CD4/CD8 subpopulation analyses revealed a heterogeneous CD4/CD8 distribution in both Nipa ^ko/ko and Nipa ^wt/wt lymphomas ( Figure 3E ), which was, despite minor variations, similar in both groups with distinct CD4/CD8-double positive, CD4/CD8-double negative, and CD4- and CD8-single positive populations. Further analysis regarding the DN stages showed no significant difference between Nipa ^wt/wt and Nipa ^ko/ko lymphomas ( Figure 3F ). Our results therefore show that NIPA seems to play a significant role in NPM-ALK-induced lymphomagenesis but does not alter the disease immunophenotype of ALCL.

Given the prolonged survival benefit in three different Nipa-deficient NPM-ALK-positive lymphoma mouse models, we investigated the regular activation of different oncogenic signaling pathways. Interestingly, in the presence of NPM-ALK, STAT3, AKT, and ERK1/2 demonstrated regular phosphorylation and therefore activation independent of NIPA ( Supplementary Figure 3A ). Thus, as it has been shown previously that Nipa deficiency significantly reduces hematopoietic stem cell frequency after replication stress due to impaired DNA damage repair followed by apoptosis (35), we analyzed thymic lymphomas of Lck/Cre^TG/wtNipa^{wt /wt} and Nipa ^flox/flox MSNAIE transplanted mice for stem cell markers. We indeed found a small but distinct lymphoma subpopulation characterized by expression signatures Lineage-negative (CD4⁻CD8⁻CD25⁻CD44⁻), cKIT, and SCA1-positive, which was significantly lower in Nipa^{ko /ko} lymphoma cells, being only one-third of Nipa ^wt/wt cells (0.19% vs. 0.57%, p = 0.04, Figures 4A, C and Supplementary Figure 3B ). Common lymphoid progenitors, measured by positivity for Il7Rα, demonstrated no significant difference in the dependency of NIPA ( Figures 4B, C ).

Taken together, the Nipa-deficient ALCL-like mouse model demonstrates a crucial role of NIPA in primary ALCL-like lymphomas. The Nipa-deleted lymphomas showed a reduced frequency of lymphoma cells expressing stemness markers, which may thereby contribute to the decelerated course of disease.