2. Results and Discussion
The observation that CD34+
cells harbor signs of an immature state, illustrated by low demarcation membrane system amplification and ploidy, raised the hypothesis of MK differentiation inhibition at a transcriptional level through SR1-modulated AHR. A number of hematopoietic TFs have been identified to promote or favor MK progenitor specification, MK maturation and platelet formation, including GATA-1, Friend of GATA-1 (FOG-1, ZFPM1), Fli-1, RUNX-1 and NF-E2 [7
]. These TFs have also been described to promote the expression of megakaryocytic receptors [3
]. In contrast, IKAROS (IKZF1),a member of the zing-finger protein family important for early lymphocyte development, has been shown to exert a repressor effect on MK differentiation and to restrain terminal MK maturation [8
]. Indeed, some effector pathways downstream from IKAROS are involved in MKP such as CTNND1 (catenin delta 1/p120 catenin), belonging to the Wnt, Rho family GTP-ase, CDKN1A (cyclin dependent kinase inhibitor 1A). These target genes are crucial actors of MK maturation and platelet production. This effect is thought to occur via its interaction with a network of TFs and TF-associated proteins including GATA and NF-E2. Interestingly, it was recently observed that IKAROS negatively regulates the immune response of the gut by interacting with AHR. These findings therefore suggest the hypothesis that IKZF1 could represent a molecular link in CD34+
cells expansion in response to the AHR antagonist SR1 [10
To test this hypothesis, we first refined the phenotype of CD34+
cells by evaluating whether the expression of tetraspanin CD9, a marker of mature MKs [11
], could represent another differentiating criteria. Flow cytometry analysis of CD9 expression within the CD34+
population allowed delineating two subpopulations: CD34+
cells (Figure 1
Ai), representing 41.4 ± 0.6% and 58.3 ± 0.7% of the CD34+
cells, respectively (Figure 1
Aii). We then evaluated the capacities of these two subpopulations to produce mature MKs and platelets following cell sorting and culture in the presence of TPO and SR1 (Figure 1
Bi), as described previously by our group [6
]. Whereas less than 10% of CD34+
-derived MKs exhibited proplatelets, this proportion rose to more than 90% when MKs were derived from CD34+
cells (Figure 1
Bii). This was accompanied by a dramatic difference in platelet production between the two subpopulations (0.6 ± 0.1 × 106
/well vs. 1.85 ± 0.3 × 106
/well, respectively) (Figure 1
Biii). This represented a slightly higher efficiency than with the whole CD34+
population (1.39 ± 0.24 × 106
/well), suggesting a negative effect of the CD9+
subpopulation on the potential of the CD9−
subpopulation to support platelet production. Using this CD34+
phenotype to define the population of interest (POI) we then investigated the role of IKAROS in its expansion. We compared the POI to CD34−
cells, previously shown to represent more mature MKs, based on their ploidy and DMS development (mature MK, MMK) but unable to produce proplatelets when passed in TPO-containing media [6
]. qRT-PCR revealed a reduced level of expression of a number of genes representative of MK maturation (CD9, PF4, GP1BA, GATA1, NFE2) supporting the hypothesis of a negative role of IKFZ1 on the MK maturation in response to SR1 [9
] (Figure S1
We then analysed IKAROS expression in the POI by flow cytometry and observed a 4.5-fold higher expression compared to MMK (134.3 ± 5.9 vs. 29.3 ± 1.2 MFI, respectively) (Figure 1
Ci, ii). This pattern of IKAROS expression suggested its functional involvement in POI specification and expansion upon SR1 treatment. This was evaluated by combining gene silencing and pharmacological approaches. IKZF1 was knocked down using a specific shRNA lentivirally transduced in CD34+
cells. A 75% decrease in IKZF1 expression was reached at day 10 in shIKZF1-transduced cells compared to cells transduced with a scramble shRNA (17.4 ± 2.2 % vs. 68.5 ± 20.2 % expression, respectively) (Figure 2
A), which resulted in a 3-fold decrease in POI generation (12.5 ± 0.7% vs. 39.1 ± 2.1%, respectively) (Figure 2
B). In the second approach CD34+
cells were treated with Lenalidomide (LenaL), a chemical compound inducing rapid and effective degradation of IKAROS (Figure S2
). Exposure to increasing concentrations of LenaL in the presence of SR1 led to a gradual decrease in POI expansion, reaching a factor of 2 at 10 µM compared to control DMSO (16.1 ± 3.9 % vs. 32.9 ± 2.5 %, respectively) (Figure 2
Ci). Together, these results provided evidence for a functional implication of IKAROS in the expansion of the POI and ensuing increased platelet production in response to SR1 (9.44 ± 2.1 × 105
/well in SR1 vs. 4.1 ± 1.5 × 105
/well with LenaL 1 µM) (Figure 2
Cii). To explore a possible molecular link between IKAROS and AHR in POI expansion we then performed a proximity ligation assay (PLA) [12
], a fluorescence-based technique allowing the detection of protein-protein interaction based on physical proximity, using specific antibodies against AHR and IKAROS [12
]. As shown in Figure 2
Di, numerous positive signals were detected in the POI in response to SR1 whereas scarce positive signals were observed in control condition with DMSO (7.8 ± 1.0 dots per POI with SR1 compared to 1.9 ± 0.2 dots in the control DMSO) (Figure 2
Dii), indicating induction of nuclear AHR:IKAROS interaction in response to SR1.
In a previous work, we demonstrated that co-culture of MKs with MSC also promoted the appearance of a CD34+
population through an AHR-dependent pathway, as shown by repression of CYP1B1
]. Thus, we investigated whether this effect was, as for SR1, also dependent on AHR:IKAROS interaction. Beforehand, we established that CD34+
populations were also delineated under MSC co-culture conditions in proportions similar to those observed in SR1 condition, (52.8 ± 6.1% of CD34+
cells vs. 44.7 ± 7.0% of CD34+
cells) (Figure 3
A). Next, we observed that IKAROS was clearly better expressed in the POI (78.2 ± 4.6% cells) than in MMKs (45.9 ± 7.1% cells) (Figure 3
). Additionally, IKZF1 shRNA knock down resulted in a markedly reduced POI amplification (8.4 ± 0.6% in shIKZF1-derived POI vs. 36.2 ± 2.5% in shControl -derived POI) (Figure 3
Ci–iii). Finally, as observed upon SR1 treatment, positive signals were observed using PLA indicating an AHR:IKAROS interaction in MSC-derived POI (1.9 ± 0.2 dots/nucleus in control vs. 3.7 ± 0.4 dots/nucleus in MSC condition n = 81) (Figure 3
D). Altogether, these results support the hypothesis that MSC, similarly to SR1, promotes POI amplification by acting on an AHR pathway that also involves a AHR:IKAROS interaction, thus distinct from the canonical pathway.
Our previous work highlighted the existence and expansion of a megakaryocyte progenitor with high platelet production capacity in response to an AHR antagonist [6
]. Repression of the target gene, CYP1B1
in response to SR1 and also after co-culture of CD34+
cells with MSC initially raised the possibility that this could occur via the canonical pathway. The present study revealed an alternative mechanism via the interaction between AHR and IKAROS, supported by the following observations of (i) increased IKAROS expression in the POI, (ii) prevention of POI generation following down-regulation or degradation of IKAROS, (iii) demonstration of a physical interaction between AHR and IKAROS, and (iv) negative regulation of megakaryocytic TFs.
In addition to highlighting a new mechanism that regulates megakaryopoiesis, our work nicely connects previous reports on the role of IKAROS in MK specification and maturation [8
] and AHR:IKAROS interaction in ILC3 cells [10
]. This work opens the possibility of genetic or pharmacological intervention on this alternative pathway to improve our capacity to produce cultured platelets at more acceptable costs.
3. Material and Methods
3.1. Extraction of CD34+ Cells
-enriched cells were obtained from leukofilters (TACSI, Terumo, Tokyo, Japan by magnetic-activated cell sorting (CD34 MicroBead Kit UltraPure, Miltenyi Biotec, Bergisch Gladbach, Germany) as described [6
] The mean percentage of viable CD34+
cells was 97.9 ± 0.3%, n = 13.
3.2. Megakaryocyte Differentiation in Culture
In the first phase, CD34+ cells were seeded in 24-well plates at a density of 4 × 104/mL in StemSpan Serum-Free Expansion Medium (SFEM) supplemented with 20 µg/mL human LDL, a cocktail of cytokines containing SCF, TPO, IL-6 and IL-9 (StemSpan™ Megakaryocyte Expansion Supplement) and 1 µM SR1 (all from Stemcell Technologies, Vancouver, Canada). On day 7, the cells were harvested, washed and seeded at a density of 5 × 105/mL in StemSpan SFEM containing 1 µM SR1, 50 ng/mL TPO and 20 µg/mL human LDL and cultured for an additional 6 days.
and MSC co-culture, cells were seeded on a confluent layer of MSCs at day 0 and day 7 [13
The inducer of IKAROS degradation lenalidomide (gift from Chan S. lab) was added in the culture at concentrations ranging from 0.1 µM to 10 µMat at days 0 and 7.
3.3. Lentiviral Short Hairpin Knockdown
shRNA constructs targeting the human IKZF1 and shRNA scramble control cloned intro psi-LVRU6GP were obtained from GeneCopoeia (Rockville, MD, USA). The plasmids pCG-HΔ24 and pCG-FΔ30 encoding for glycoprotein of measles virus envelope were kindly provided by Verhoeyen E [14
] and p8.91, an encapsidation plasmid lacking all accessory HIV-1 proteins (Vif, Vpr, Vpu and Nef) by A. Dubart-Kupperschmitt.
3.4. Lentivirus Production and CD34+ Cells Trasnduction
Lentiviruses were produced by transient transfection of HEK293T cells in DMEM (Invitrogen, Carlsbad, CA, USA).
Plasmids: pCG-HΔ24 (5 µg), pCG-FΔ30 (5 µg), p.8.91 (16 µg), psi-LVRU6GP (IKZF1 or shControl) (16 µg) were transfected with Lipofectamine 2000 (100 µL) in Optimem medium (Invitrogen) into HEK293T cells. After 18 h of transfection, the medium was replaced by Opti-MEM (Invitrogen). Two days later, viral supernatants were harvested and filtered. They were concentrated by ultracentrifugation and the resulting pellet was suspended in phosphate buffered saline (PBS) and frozen at −80 °C in 25-μL aliquots until use.
Infectious titers (in transduction units [TU]/mL) were determined by flow cytometry analysis of target cells using serial dilutions of the supernatants added to HEK293T.
CD34+ cells transduction. CD34+ cells were grown for 24 h in Stemspan medium with StemSpan™ Megakaryocyte Expansion Supplement (Stemcell Technologies). Then, cells were thoroughly washed and seeded in wells coated with retronectin (Takara Bio, Kusatsu, Japan). Transduction was subsequently performed by adding lentiviral particles at 5 MOI for 72 h. At days 7 and 10, following infection, cells were analyzed by flow cytometry.
3.5. Cell Sorting and Flow Cytometry Analysis
. Cell sorting was performed using an Aria III flow cytometer as previously described by our group [6
]. Antibodies are listed in Table 1
. The sorted CD34+
populations were seeded in Stemspan medium containing TPO 50 ng/mL, LDL 20 µg/mL and SR1 1 µM for 7 day.
Flow cytometry analysis
. Flow cytometry analysis was performed using Fortessa X20 (Becton Dickinson, Franklin Lakes, NJ, USA), as previously described by [6
]. Antibodies are listed in Table 1
For intracellular staining, cells were first labeled with surface antibodies, fixed, permeabilized using the Foxp3/Transcription Factor Staining Buffer Set (Invitrogen) according to the manufacturer’s instructions and then labeled with an anti-IKAROS antibody or mouse IgG2a κ-647 (BioLegend, San Diego, CA, USA) (Table 1
. Culture-derived platelets were released following successive pipetting after the addition of 0.5 µM PGI2 and 0.02 U/mL apyrase in the culture medium and analyzed as previously described [6
All samples were analyzed using a Fortessa X20 flow cytometer (Becton Dickinson).
3.6. Quantitative RT-PCR
RNA extraction and cDNA synthesis were performed using RNeasy Mini Kit (Qiagen, Hilden, Germany) and RT2 First Strand Kit (Qiagen), respectively, according to manufacturer’s instructions. qPCR was performed using the RT² SYBR Green qPCR Mastermix (Qiagen) on a CFX96 Touch Real-Time PCR Detection System (Biorad, Hercules, CA, USA). Relative transcript levels were calculated by the method of ΔCt with TBP like reference gene. Primers set for TBP, IZKF1, ZFMP2, CD9, PF4, GP1bA, NF-E2, GATA1 were purchased from Genecopoeia.
3.7. DuoLink Proximity Ligation Assay (PLA) Analysis
-derived megakaryocytes were harvested and cytospun onto poly-L-lysine coated slides. Immobilized cells were then fixed with PFA 4–0.01% glutaraldehyde and permeabilized with 0.05%Triton X100. PLA was subsequently performed according to manufacturer’s instructions using primary antibodies against IKAROS and AHR described in Table 1
. The interacting tandems AHR/IKAROS were visualized by fluorescence on a SP8 confocal microscope (Leica, Wetzlar, Germany).
3.8. Statistical Analyses
Results were expressed as the mean ± SEM and statistical comparisons were performed using an unpaired, two-tailed Student’s t-test or one-way ANOVA followed by the Bonferroni post-hoc test (Prism, Graph-Pad Software Inc., San Diego, CA, USA, version 5). p values of less than 0.05 were considered to be statistically significant.