We intercrossed N3tg (8) and p50^−/− (39) mice, both on C57BL/6 background, to generate N3tg/p50^−/− double-mutant mice, that were bred and housed under specific pathogen-free conditions. The p50^−/− mice were obtained from Jackson Laboratories, Bar Harbor, ME, USA. All mice were monitored daily and euthanized upon disease detection (8), as evidenced by enlarged spleen, hunched posture, ruffled fur, reduced mobility, and/or labored breathing. Experimental groups were based on age and genotype of mice. The Foxp3EGFP reporter mice (38), are “knock-in” mice on C57BL/6 background, overexpressing an IRES-EGFP cassette in the 3′ untranslated region of Foxp3 gene. The number of mice used in each experiment was reported in figure legends. Animal studies were approved by the local Animal welfare committee and were carried out in accordance with the recommendations of the Italian national guidelines for experimental animal care and use and of the European Directive 2010/63/EU.

Single-cell suspensions were prepared from thymus, spleen, bone-marrow, or peripheral blood in 1x PBS supplemented with 2% FBS and erythrocytes were lysed with ammonium-chloride-potassium lysing buffer, as previously reported (40). Freshly isolated cells were stained with surface markers for 30 min on ice using the following antibodies: CD4 (RM4-5), CD8a (53–6.7), CD11b (M1/70), Gr-1(RB6-8C5), all from BD Biosciences. For Treg detection cells were stained with the following surface antibodies: CD4 (RM4-5), CD8a (53-6.7), CD25 (PC61) (all from BD Biosciences) and then with the intracellular Foxp3 antibody (FJK-16s) (eBioscience), by using Foxp3/Transcription Factor Staining Buffer Set (eBioscience), following manufacturer's instructions. Samples above were run on a FacsCalibur (BD Biosciences) and analyzed with the CellQuest Pro software (BD Biosciences).

To evaluate early myeloid progenitors distribution, bone-marrow cells were stained with the APC mouse lineage Ab cocktail (BD Biosciences), and with the APC-CD4, APC-CD8a, and APC-IL7Ra antibodies (BD Biosciences), to determine the Lineage negative (Lin⁻) subset, and then with the APCH7-cKit (2B8), PerCPCy5.5-Sca-1(D7), FITC-CD34 (RAM34), and PE-FcgRIII/II (2.4G2) antibodies (BD Biosciences). Analysis of apoptosis was performed in gated CD4⁺CD8⁺ DP T subset by staining cells with surface markers V450-CD4 (RM4-5) and PE-CD8a (53–6.7), and then by labeling cells with BUV395-Annexin V and 7-AAD (BD Biosciences), as previously described (41). In order to evaluate apoptosis in Tregs, cells were stained with surface markers PerCPCy5.5-CD4 (RM4-5) and FITC-CD8a (53–6.7), then labeled with Fixable Viability Stain 780 and APC-Annexin V (both from BD Biosciences), just prior to fixation/permeabilization and staining with PE-Foxp3 (FJK-16s), performed as above. Proliferation of CD4⁺CD8⁺ DP T cells was assessed by intracellular staining with the BV510-Ki-67 (B56) antibody (BD Biosciences). For cell cycle analysis 7AAD (Sigma) was used at 25 mg/ml with RNase (Sigma) 40 mg/ml. All intracellular stainings were performed with Foxp3/Transcription Factor Staining Buffer Set (eBioscience). For the intracellular staining with anti-pSTAT5/pY694 (47/STAT5; BD Bioscience) T splenocytes were isolated by using Pan T cell isolation kit (Miltenyi), following manufacturer's instructions. Before the staining, the cells were either left unstimulated or stimulated with increasing doses of rhIL2 (Peprotech) for 15 min at 37°C and then, were fixed and permeabilized with Transcription Factor Phospho Buffer Set (BD Bioscience) according to the manufacturer's instructions. Samples were run on BD LSRFortessa equipped with DIVA software (BD Biosciences) and data were elaborated using FlowJo software (TreeStar).

N3tg/p50^−/− double-mutant mice at 3 weeks of age were injected intraperitoneally with 250 μg of InVivoPlus anti-mouse CD8α (2.43 clone), plus 250 μg of InVivoPlus anti-mouse CD4 (GK1.5 clone) or with 500 μg of InvivoPlus RatIgG2b (LTF-2 clone) isotype control (all from BioXCell), resuspended in 200 μl/mouse of PBS 1x, twice a week. After 3 weeks of treatment mice were sacrificed and characterized by FACS analysis, as described above. In particular, to evaluate the distribution of T cell subsets we used the following surface antibodies: CD4 (RM4-4, BD Bioscience) and CD8β (H35-17.2, eBioscience).

For sorting experiments, bone-marrow samples were obtained from wt, N3tg, and N3tg/p50^−/− mice, as above. Mononuclear cells were isolated by ficoll gradient and stained with a “lineage cocktail” containing PE-conjugated monoclonal antibodies against CD11b, Gr-1, Ter119, CD45R, CD3ϵ, CD4, and CD8a. “Lineage negative” (Lin⁻) cells were purified with the FACSAria cell sorter (BD Biosciences; purity ≥ 98%).

In other sets of experiments, thymus or spleen cell suspensions were obtained and stained with anti-CD4 and anti-CD8 surface markers, as above. Then CD4⁺CD8⁺ (DP), CD4⁺CD8⁻, or CD4⁻CD8⁺ T cells from wt, N3tg, and N3tg/p50^−/− mice, or CD4⁺CD8⁻EGFP⁺ Tregs from spleen of Foxp3EGFP reporter mice (see above), were isolated (purity ≥ 98%) using a BD FACSAria III (BD Biosciences), equipped with FACSDiva software (BD Biosciences), as previously described (42).

For myeloid colony forming unit assay, Lin⁻ cells (5 × 10³ cells/well), purified as above, were plated in triplicate in methylcellulose-based semisolid medium (Methocult M3236, Stem Cell Technology) with FCS and the following cytokines: IL-3 (2 ng/ml), IL-6 (2 ng/ml), SCF (50 ng/ml), G-CSF (50 ng/ml), and GM-CSF (10 ng/ml). After 7–10 days colonies were analyzed and counted and cells were harvested, controlled for the expression of myeloid markers by FACS analysis and then replated (10 × 10³/well) to assess survival.

All the cell culture samples were cultured at 37°C and 5% CO₂ in complete medium, that is RPMI-1640 medium (GIBCO), supplemented with 10% FBS, 10 U/ml penicillin and streptomycin, 2 mM glutamine. In particular, total cells obtained from bone-marrow of wt mice (0.25 × 10⁶/well) were co-cultured 1:1 in 96 well plates with total T splenocytes from wt, N3tg, or N3tg/p50^−/− mice. T splenocytes were isolated by using Pan T cell isolation kit (Miltenyi), following manufacturer's instructions. After 24 and 48 h of co-culture, cells were counted and stained with CD11b and Gr-1 antibodies for FACS analysis, as above. In some experiments, the co-culture assay was conducted by using CD4⁺CD8⁺ (DP), CD4⁺CD8⁻, or CD4⁻CD8⁺ T splenocytes sorted from double-mutant mice at purity ≥ 98%, as described previously. To check for the role of cell-to-cell contact mechanisms, we performed the co-culture experiments above in 96 well plates in the presence of transwell inserts (pore diameter, 0.4 μm, Corning). Finally, for testing the possible role of Tregs, total cells obtained from bone-marrow of wt mice (0.25 × 10⁶/well) were co-cultured 1:1 in 96 well plates with total T splenocytes from N3tg/p50^−/− mice, without or with different numbers of CD4⁺CD8⁻EGFP⁺ Tregs (1 × 10⁴/3 × 10⁴/6 × 10⁴), purified from spleen of Foxp3EGFP reporter mice (see above).

In another set of experiments, freshly isolated thymocytes (0.5 × 10⁶/well) were cultured in 96 well plates in the following conditions: untreated, upon TCR activation with coated anti-CD3 antibodies (145-2C11) 3 μg/ml, or in the presence of rhIL2 (Peprotech) 50 U/ml or murine IL-15 (Peprotech) 50 ng/ml. Then, apoptosis was assessed at 48 h by FACS analysis, as described above.

Finally, T splenocytes from wt, N3tg, or N3tg/p50^−/− mice (0.5 × 10⁶/well), purified as above, were cultured in 96 well plates and stimulated or not with coated anti-CD3 antibodies (145-2C11) 3 μg/ml plus rhIL2 (Peprotech) 50 U/ml, in triplicates. Then, at 24 and 48 h, cells were counted and stained to assess Treg numbers by flow cytometry, as described above.

CD4⁺CD8⁺ thymocytes purified as above were processed to extract total RNA with TRIzol reagent (Invitrogen), and reverse transcription was performed with High-Capacity cDNA Reverse Transcription Kit (ThermoFisher), according to the manufacturer's protocol. TaqMan quantitative real-time PCR (qPCR) was performed on cDNA using the StepOnePlus™ Real-Time PCRSystem (ThermoFisher), following instructions of manufacturer. Taqman Gene Expression Master Mix and Taqman Gene Expression Assays on demand for Bcl2 (Mm00477631_m1), A1 (Mm03646861_mH), and Hprt (Mm01545399_m1) were purchased from ThermoFisher. Relative quantification was carried out using the comparative ΔΔCt method. Hprt expression was used to normalize the expression levels of mRNAs.

Total protein extracts were prepared from sorted DP T thymocyte samples and Western blotting analysis was conducted, as previously described (43), using anti-p21 antibody (C-19, sc-397, Santa Cruz), and anti-Caspase-3 or anti-cleaved Caspase-3 antibody (#9665 and #9661, respectively, both from Cell signaling). Anti-β-actin antibody (Sigma-Aldrich), was used to normalize protein expression levels. Densitometric analysis was performed using ImageStudio software (LI-COR Biosciences).

Results were expressed as means ± SD. We performed unpaired two-tailed Student's test. Differences were considered significant when ^*P ≤ 0.05, ^**P ≤ 0.01, ^***P ≤ 0.001, and ^****P ≤ 0.0001. Kaplan-Meier survival analysis was performed comparing kinetics of disease development in N3tg/p50^−/−, N3tg, wt, and p50^−/− mice. P-value was calculated by Log rank (Mantel-Cox) test. Statistical analysis was performed using Prism, GraphPad software.

In order to analyze the effects of canonical NF-κB pathway modulation on Notch3-dependent T-ALL we generated N3tg/p50^−/− double-mutant mice, by intercrossing NF-κB1/p50^−/− mice (39), with N3tg animals (8). Surprisingly, the follow-up of N3tg/p50^−/−, N3tg, and relative control mice (wt and p50^−/−), revealed that N3tg/p50^−/− double-mutant mice had a median survival of 65.5 days; in contrast, N3tg mice showed a median life span of 109.5 days (Figure 1A). Notably, N3tg/p50^−/− mice presented clinical signs that were different from typical features of T-ALL, routinely observed in N3tg animals (8). At the end point, N3tg/p50^−/− mice appeared indeed smaller in size with respect to wt or single mutant controls (not shown). Moreover, disease of N3tg/p50^−/− mice at 8–9 weeks of age was accompanied by splenomegaly, though less pronounced than that observed in N3tg mice (Figure 1B and not shown). Finally, the thymus of double-mutant mice was dramatically reduced in size (Figure 1C and not shown), starting at 4–5 weeks of age.

To clarify the nature of double-mutant mice pathology we performed immunophenotypic analysis of hematopoietic cell subsets in different organs from N3tg/p50^−/−, N3tg, p50^−/−, and wt mice, by FACS analysis. Regarding the T cell compartment, we focused on immature CD4⁺CD8⁺ (DP) T-cell population. These cells are normally confined to the thymus, while their presence in the periphery represents a reliable marker to follow T-ALL progression (44–46). CD4⁺CD8⁺ (DP) T cells were highly decreased in percentages and numbers in both spleen (SPL; Figures 2A,B) and bone-marrow (BM; Figures 2C,D) of N3tg/p50^−/− vs. N3tg mice at 8–9 weeks of age, whereas they were virtually absent in organs from p50^−/− and wt controls (not shown). Conversely, the analysis of myeloid cell distributions revealed marked expansion of CD11b⁺Gr-1⁺cells in the spleen (Figures 3A,B), as well as in the BM (Figure 3C, upper panel) and peripheral blood (Figure 3C, lower panel) of N3tg/p50^−/− mice at 8–9 weeks of age, when compared with N3tg, p50^−/−, and wt counterparts. Collectively, our results indicate that the deletion of NF-κB1 in N3tg mice induces a delay of T-ALL progression on one hand, and promotes myeloproliferation on the other hand, thus affecting the composition of T-ALL immune-environment.

The accumulation of CD11b⁺Gr-1⁺ cells suggested the presence of alterations in the myeloid cell development of N3tg/p50^−/− double-mutant mice. Thus, we analyzed the distribution of myeloid progenitor subsets [as defined by the differential expression of CD34 and FcγRII/III markers inside the Lin⁻ckit⁺Sca1⁻ compartment, see (47) and Figure 4A legend], in the bone-marrow of mice of different genotypes at 4–5 weeks of age. We revealed that double-mutant mice displayed a significant increase in the percentages (Figures 4A,B, left panel), and absolute numbers (Figure 4B, right panel) of “granulocyte/monocyte progenitor” (GMP) subset with respect to both N3tg and wt littermates, whereas the progenitor distribution in the BM of p50^−/− mice was similar to wt controls [(48) and not shown]. To confirm these data, we performed a myeloid “colony forming unit” assay with Lineage negative (Lin⁻) cells, purified from the bone-marrow of N3tg/p50^−/−, N3tg, and wt mice, and plated on semi-solid methylcellulose medium in the presence of factors stimulating the differentiation of progenitors toward a myeloid fate (see the Materials and Methods section). As depicted in Figure 4C, the number of colonies/plate (CFUs) obtained at the P2 and P3 re-plating points, was significantly higher in N3tg/p50^−/− when compared to N3tg and wt mice.

Interestingly, immature CD4⁺CD8⁺ (DP) T cells are still largely represented in the BM of double-mutant mice at 4–5 weeks of age (Figure 4D), thus leaving open the possibility that these immature N3tg/p50^−/−DP T cells may participate in inducing in trans the alteration of myeloid cell development described above.

To test the hypothesis that T cells may influence the myeloproliferation observed in double-mutant mice, we designed experiments of in vivo T-cell depletion, by the combined administration of anti-CD4 and anti-CD8 antibodies to N3tg/p50^−/− mice, starting at 3 weeks of age. After 3 weeks, N3tg/p50^−/− treated mice displayed a significant reduction of percentages (Figures 5A,B, upper panels) and absolute numbers (Figure 5B, lower panels) of all the examined T cell subsets in the BM and spleen, with respect to the N3tg/p50^−/− controls, as expected. Notably, T-cell depleted mice were characterized by a remarkable decrease in the proportion (Figures 5C,D, upper panels) and absolute numbers (Figure 5D, lower panels) of the CD11b⁺Gr-1⁺ subset in the same organs as above, when compared to the controls. Our results strongly suggest that T cells participate in sustaining the myeloid proliferation of double-mutant mice.

To gain more insight into the T/myeloid cell interaction in N3tg/p50^−/− mice, we performed in vitro co-culture experiments of total wt BM cells with T cells purified from the spleen of double-mutant, N3tg or wt mice at 4–5 weeks of age. Strikingly, we observed a significantly higher CD11b⁺Gr-1⁺ cell count ratio in the co-cultures with N3tg/p50^−/− T cells compared to the co-cultures with N3tg or wt T cells, at any time point considered (Figure 5E), indicating that N3tg/p50^−/− T cells may sustain the growth of CD11b⁺Gr-1⁺ cells in vitro. Then, we repeated the co-culture assay by using transwell inserts or by using different subsets of double-mutant T splenocytes, instead of total T splenocytes (Figure 5F). The results indicated that transwell inserts strongly inhibited the effect of N3tg/p50^−/− T cells on CD11b⁺Gr-1⁺ cell growth, thus suggesting that it requires cell-to-cell contact. Furthermore, we revealed that only CD4⁺CD8⁺ (DP) T cells from double-mutant mice are able to drive the expansion of myeloid compartment, whereas CD4⁺CD8⁻ and CD4⁻CD8⁺ T subsets do not influence this process, significantly. Overall, our in vivo and in vitro combined approaches suggest that the myeloproliferative trait of N3tg/p50^−/− mice may rely on a mechanism exerted in trans by N3tg/p50^−/− DP T cells and that requires cell-to-cell contact.

The thymus represents the natural environment in which DP T cell develop under the control of many pathways, including Notch and NF-κB. The reduction of DP T cell subset in the spleen and BM of double-mutant mice, coupled with thymus regression, suggested that such phenomena may rely on alterations of T cell development. This prompted us to analyze T cell compartment in more details. However, to minimize interference by T-ALL or myeloproliferation, we performed our studies in young mice, at 4–5 weeks of age, when N3tg/p50^−/− mice presented no evident symptoms of illness.

Percentages (Figures 6A,B) and absolute numbers (Figure 6C) of CD4⁺CD8⁺ (DP) T cells were remarkably reduced in the thymus of N3tg/p50^−/− mice, when compared with those of N3tg, p50^−/− and wt mice. Concurrently, double-mutant mice presented increased percentages of CD4⁺CD8⁻, CD4⁻CD8⁺, and CD4⁻CD8⁻ thymocytes (Figures 6A,B). However, in terms of absolute numbers, the reduction observed in N3tg/p50^−/− vs. N3tg mice was not limited to DP T cells, but concerned also other subsets, with particular regard to CD4⁺CD8⁻ T cells (Figure 6C). Interestingly, the expansion of all thymocyte subsets that characterize N3tg vs. wt mice [(8) and Figure 6C], was abrogated in double-mutant mice. Moreover, the effects of NF-κB1 deletion on thymocytes emerged specifically in the Notch3 transgenic background, being absent in p50^−/− mice, when compared to wt controls. In summary, our results suggest that deletion of NF-κB1 may revert the consequences of Notch deregulated activation inside the thymic T-cell compartment.

Similar conclusions derived from analysis of splenic DP T cell compartment, that appeared significantly restricted in N3tg/p50^−/− vs. N3tg mice, in both percentages (Figures 7A,B) and absolute numbers (Figure 7C). These results suggest that the delay of T-ALL progression observed in N3tg/p50^−/− vs. N3tg mice at 8–9 weeks of age (see Figure 2), is already effective at an initial stage of the disease. We noted a slight decrease of CD4⁺CD8⁻ T cell percentages and absolute numbers (Figures 7D,E, respectively, left panels), in N3tg mice, with respect to those of p50^−/− and wt controls, though these reductions were recovered in double-mutant animals. Furthermore, percentages and absolute numbers of CD4⁻CD8⁺ T subset (Figures 7D,E, respectively, right panels), were reduced in N3tg/p50^−/− and N3tg mice in a similar way.

Notch signaling deregulation in T-cell compartment correlates with the expansion of Tregs during T-ALL development in N3tg mice (37), depending in part on the activation of canonical NF-κB pathway (38). Besides, canonical NF-κB subunits, such as c-rel and p65, have been described as crucial in Treg development, function, and homeostasis, including in the context of cancer (49–52). Moreover, c-rel has a crucial role in promoting the formation of an enhanceosome specific for the FoxP3 promoter and that also includes p65, whereas p50 does not activate FoxP3 promoter (52). Notably, a major role of NF-κB1 in Treg biology has been excluded (49, 52, 53), though p50 modulation seems to exert detrimental effects on Tregs in particular conditions, such as in p50/c-Rel double knock-out mice (53, 54) or in mice carrying a mutation of p105 (the p50 precursor) that blocks its degradation (55). In this context, N3tg/p50^−/− model allowed us to verify if NF-κB1 may cooperate with Notch3 in regulating Treg behavior in T-ALL environment. To this end, we examined CD4⁺CD8⁻Foxp3⁺ Treg distributions in the thymus and spleen of N3tg/p50^−/− mice, by flow cytometry. In the thymus, percentages of Tregs were normal, relatively to those of controls (Figures 8A,B), while their absolute numbers declined (Figure 8C). This event translated in the spleen of N3tg/p50^−/− mice, where Tregs diminished significantly in percentages (Figures 8D, E), as well as in absolute numbers (Figure 8F). Notably, Treg subset was exclusively reduced when NF-κB1 ablation occurred in the N3tg background, suggesting that Notch and NF-κB1 may co-operate in regulating Treg numbers specifically in the context of T-ALL.

In the attempt to explain the reduction of DP T cells and Tregs in N3tg/p50^−/− mice, we analyzed apoptosis of these populations, by 7AAD/Annexin V assay on freshly isolated cells. CD4⁺CD8⁺ (DP) T subset from N3tg mice presented an accumulation of Annexin V⁺ cells in both the thymus (Figures 9A,C) and spleen (Figures 9B,C), in comparison to wt DP thymocytes, used as a control. Interestingly, similar results were reported in transgenic mice with an lck-driven constitutive activation of Notch1 intracellular domain (56). However, the percentage of apoptotic cells was even more increased in DP T cells from N3tg/p50^−/− mice (Figure 9), suggesting that Notch and NF-κB1 may cooperate in regulating survival of DP T cells. Then, we analyzed if the increase of apoptotic rate described above may depend on a T cell intrinsic mechanism or whether it is a secondary effect of the myeloproliferation on microenvironment. To this end, we performed in vitro experiments to measure the percentages of apoptotic cells in DP thymocytes from N3tg/p50^−/−, N3tg, and wt mice that were cultured in the following conditions: untreated, upon TCR activation via anti-CD3 stimulation or in the presence of survival cytokines such as IL-2 or IL-15. Interestingly, significant differences in the percentages of Annexin V⁺ cells persisted among the three groups of DP thymocytes (N3tg/p50^−/− > N3tg > wt) in all the conditions tested (Figure 9D), in a similar extent to what observed in vivo, thus indicating the cell-autonomous nature of this event.

Inside Treg subset, the proportion of apoptotic cells was increased in CD4⁺CD8⁻FoxP3⁺ cells from both the thymus (Figure 10A, left panels and Figure 10C, upper panel) and spleen (Figure 10B, left panels and Figure 10C, lower panel) of N3tg/p50^−/− mice. However, percentages of Annexin V⁺ Tregs in N3tg mice were comparable to those of wt controls, and this apparently rules out a pro-apoptotic role of Notch signaling activation in Tregs. Importantly, no differences were noted in the apoptotic rate of CD4⁺CD8⁻FoxP3⁻ T cells from mice of different genetic backgrounds (Figures 10A,B, right panels), used as an internal control. We also excluded defects in the proliferation rates of Tregs that increased in a similar way in both N3tg and N3tg/p50^−/− mice, when compared to those of wt counterparts, as measured by Ki67 staining (not shown). In summary, our results suggest that the decrease of DP T and Treg cells in double-mutant mice relies on the enhancement of their apoptotic rate.

The IL-2/IL2R system is extensively involved in regulating many aspects of Treg biology, including the survival. Interestingly, in Tregs from double-mutant mice we did not observe a defective expression on a per cell basis of the CD25 receptor (Figure 11A, left panel), or of FoxP3 (as a specific target of the IL-2 signaling in Tregs), (Figure 11A, right panel), as evaluated by calculating their Mean Fluorescence Intensity (MFI). Nevertheless, we revealed an altered response to IL-2 of N3tg/p50^−/− Tregs in vitro. Indeed, the Treg cell count ratio increased significantly upon IL-2 stimulation in wt samples and even more in N3tg samples, with respect to the untreated controls, whereas no differences were noted in IL-2-treated with respect to untreated double-mutant Tregs (Figure 11B). These data suggest that the reduced number of Tregs in double-mutant mice could be related to the impaired response to IL-2. To corroborate our findings, we examined Tregs for the expression of the activated form of STAT5 (pSTAT5), a critical downstream target of IL-2 signaling in Treg development and function. We stimulated T splenocytes with increasing doses of rhIL-2, and then revealed the expression of pSTAT5, by FACS analysis. Data obtained evidenced that N3tg/p50^−/− mice present a slight but significant decrease in the proportion of pSTAT5⁺ cells inside the CD4⁺CD8⁻Foxp3⁺ Treg compartment, upon activation with high doses of IL-2 (50 or 200 U/ml; Figure 11C), when compared to wt controls. These results suggest the presence of an altered IL-2/STAT5 signaling in double-mutant Tregs.

Recently, it was reported that the lack of Tregs in FoxP3-deficient “scurfy” mice induces indirectly a deregulated myelopoiesis, resembling the myeloproliferative trait of our N3tg/p50^−/− mice (57). Interestingly, N3tg/p50^−/− young mice (at 4–5 weeks of age), present a dramatic reduction of Treg percentages (Figures 11D, E, upper panel), and absolute numbers (Figure 11E, lower panel), in the BM, comparable to what observed in their thymus and spleen (see Figure 8). Thus, we investigated the possible role of the Treg lack inside T cell compartment of N3tg/p50^−/− mice in promoting the CD11b⁺Gr-1⁺ cell growth observed in our co-culture system. To this aim, we performed in vitro co-culture experiments of total wt BM cells with T cells purified from the spleen of double-mutant mice at 4–5 weeks of age, in combination with different numbers of FoxP3⁺ Tregs, purified from the spleen of the Foxp3EGFP “knock-in” reporter mice (38). Interestingly, the presence of Tregs in the co-culture does not influence significantly the positive effect of N3tg/p50^−/− T cells on myeloid cell growth (Figure 11F). Thus, it seems unlikely that the lack of Tregs could represent a crucial event in promoting myeloproliferation of double-mutant mice.

In the attempt to find mechanistic basis for the enhancement of apoptosis in DP T cells from double-mutant mice, we first considered the Bcl-2 protein family, that exerts essential function in cell death regulation. In particular, Bcl-2 and A1 members have antiapoptotic roles and represent important targets of NF-κB in lymphocyte development and hematological malignancies (16, 58). Moreover, the overexpression of Bcl-2 and A1 has been associated to survival of T-lymphoma cells in N3tg mice (8). On this premise, we purified DP thymocytes from N3tg, N3tg/p50^−/−, and wt mice at 4–5 weeks of age and evaluated the expression of Bcl-2 and A1 mRNAs. Surprisingly, the expression of both Bcl-2 and A1 was increased in samples from N3tg DP thymocytes, when compared to wt DP thymocytes, and the deletion of NF-κB1 magnified this effect in double-mutant counterparts (Figure 12A). Then, we explored the possibility that NF-κB1 may affect the expression of “cyclin-dependent kinase inhibitor” p21^Waf1/Cip1. This protein represents a main factor in promoting cell cycle arrest, upon various stimuli. However, p21^Waf1/Cip1 also acts as an inhibitor of apoptosis in many cell types and through different mechanisms [as reviewed in (59)]. Intriguingly, we detected a downregulation of p21^Waf1/Cip1 protein expression levels in samples of DP thymocytes from N3tg/p50^−/− mice (Figure 12B), when compared with both N3tg and wt samples. We also studied the possible effects of p21^Waf1/Cip1 downregulation on both the cell cycle and the proliferation of double-mutant DP thymocytes with respect to N3tg and wt mice counterparts. We did not observe any significant alteration in the distribution of DP T thymocytes of different genotypes in the different phases of the cell cycle, as assessed by 7-AAD staining (not shown), as well as in the percentages of proliferating DP thymocytes, as assessed by the staining with the Ki-67 proliferation marker (Figure 10C, upper panel). However, the expression of Ki-67 on a per cell basis was significantly increased in double-mutant DP thymocytes (Figure 12C, lower panel). Finally, we revealed that the decrease of the p21^Waf1/Cip1 protein in double-mutant DP thymocytes does not translate into differences in the protein expression levels of activated cleaved caspase3 or pro-caspase3 (Figure 12D). Overall, our results suggest that NF-κB1 deletion may promote apoptosis in N3tg DP T cells through mechanism/s that are dependent on p21^Waf1/Cip1 expression.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.