Modeling Osteoclast Defect and Altered Hematopoietic Stem Cell Niche in Osteopetrosis with Patient-Derived iPSCs

Patients with osteopetrosis present with defective bone resorption caused by the lack of osteoclast activity and hematopoietic alterations, but their bone marrow hematopoietic stem/progenitor cell and osteoclast contents might be different. Osteoclasts recently have been described as the main regulators of HSCs niche, however, their exact role remains controversial due to the use of different models and conditions. Investigation of their role in hematopoietic stem cell niche formation and maintenance in osteopetrosis patients would provide critical information about the mechanisms of altered hematopoiesis. We used patient-derived induced pluripotent stem cells (iPSCs) to model osteoclast defect and hematopoietic niche compartments in vitro. Ribonucleic acid; Real-Time Quantitative Quantitative real-time SEM: Standard error of the mean; SOX2: embryonic Telomerase reverse UTF1: Undifferentiated cell transcription factor


Abstract Background
Patients with osteopetrosis present with defective bone resorption caused by the lack of osteoclast activity and hematopoietic alterations, but their bone marrow hematopoietic stem/progenitor cell and osteoclast contents might be different. Osteoclasts recently have been described as the main regulators of HSCs niche, however, their exact role remains controversial due to the use of different models and conditions. Investigation of their role in hematopoietic stem cell niche formation and maintenance in osteopetrosis patients would provide critical information about the mechanisms of altered hematopoiesis. We used patient-derived induced pluripotent stem cells (iPSCs) to model osteoclast defect and hematopoietic niche compartments in vitro.
Methods iPSCs were generated from peripheral blood mononuclear cells of patients carrying TCIRG1 mutation.
iPSC lines were differentiated rst into hematopoietic stem cells-(HSCs), and then into myeloid progenitors and osteoclasts using a step-wise protocol. Then, we established different co-culture conditions with bone marrow-derived hMSCs and iHSCs of osteopetrosis patients as an in vitro hematopoietic niche model to evaluate the interactions between osteopetrotic-HSCs and bone marrowderived MSCs as osteogenic progenitor cells.

Results
We rst demonstrated myeloid-skewed hematopoietic differentiation potential of osteopetrotic iPSCderived hematopoietic progenitors and phenotypically normal and functionally defective osteoclast formation. Upon co-culture with healthy iHSCs, the expression of the genes involved in HSC homing and maintenance (Ang-1, Sdf-1, Jagged-1, N-Cadherine, Kit-L, Opn) in osteopetrotic MSCs which revealed impaired osteogeneic differentiation, as well as their attraction ability over HSCs recovered signi cantly.
Similar change in the phenotype of osteopetrotic iHSCs occured when they interacted with healthy MSCs.

Conclusion
Our results establish signi cant alterations in both MSC and HSC compartments of the hematopoietic niche in osteopetrosis patients, which are restored with normal MSC activity supporting the role of defective osteoclasts in all these processes.
Osteopetrotic osteoclasts phenotypically identical to healthy donor osteoclasts, but functionally defective Osteopetrotic-iHSCs have a myeloid-skewed hematopoietic differentiation potential Osteopetrotic MSCs shows reduced osteogenic differentiation and their regulatory role in HSC homing and retension is impaired Restoration of MSC activity reestablishes normal HSC phenotype and function, supporting the impact of dysfunctional osteoclast in defective niche formation in osteopetrosis Background Osteopetrosis is a rare inherited disease characterized by impaired osteoclast activity causing defective bone resorption and a signi cant increase in bone mass. Inheritance can be autosomal recessive, dominant, or X-linked, but most severe cases are almost exclusively, autosomal recessive and named as malignant infantile osteopetrosis (MIOP) or autosomal recessive osteopetrosis (ARO). The lack of functional osteoclasts leads to a bone marrow cavity insu cient to support normal hematopoiesis. The resulting extramedullary hematopoiesis gives rise to massive hepatosplenomegaly and macrocephaly. Bony overgrowth results in cranial nerve dysfunction, choanal stenosis, abnormal dentition and developmental delay. Untreated MIOP has a mortality rate of approximately 70% by six years of age, mostly due to complications related to bone marrow failure. Mutations in several genes have been implicated in the pathogenesis of osteopetrosis, affecting osteoclast development, differentiation (RANK and RANKL; osteoclast-poor form), and function (TCIRG1, CLCN7, OSTM1, CA-II, and PLEKHM1; osteoclast-rich form) [1][2][3]. Different loss-of-function mutations in T cell immune regulator 1 (TCIRG1) gene which encodes vacuolar (V)-ATPase isoform a3, a subunit of the osteoclast vacuolar proton pump, account for almost 50% of all children with infantile osteopetrosis, impair osteoclast ru ed-border formation and bone resorption by disturbing secretory lysosome tra cking [4]. Allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative treatment for most children with osteopetrosis, which needs to be done as early as possible before the development of irreversible neurological complications. Transplantation of hematopoietic progenitor cells is intended to reverse the process by providing monocyte-derived donor source of functional osteoclasts, but bone remodeling and establishment of normal hematopoiesis require time. Despite the improvement of transplant strategies and better long-term survival after transplantation, morbidity and mortality rates within the rst year after transplantation are still high, with graft failure and early transplant-related complications accounting for most deaths. There is evidence suggesting that the di culty in achieving engraftment in osteopetrosis is associated with abnormalities in the bone marrow niche and hematopoietic stem cell (HSC) homing [1,5]. All patients with osteopetrosis present with defective bone resorption and hematopoietic alterations, but their bone marrow HSC and osteoclast contents might be different. In patients with osteoclast-rich osteopetrosis there is a signi cant reduction in HSC pool and brosis in bone marrow, whereas the presence of some HSC pool has been reported in patients with osteoclast-poor osteopetrosis [6,7]. There is a de nitive need to develop new transplant strategies that will support hematopoiesis to improve engraftment besides providing functional osteoclasts by healthy allografts. Investigation of the mechanism of defective osteoclast function and its role in hematopoietic stem cell niche formation and maintenance would be critical to understanding altered hematopoiesis in osteopetrosis.
Hematopoietic stem cells are maintained in the specialized bone marrow niches in which the fate of HSCs with regard to quiescence, proliferation, differentiation and migration is regulated [8,9]. HSCs reside in the endosteal region, at the interface between the bone and BM, which is the region of active bone remodeling. Enchondral ossi cation which is essential for the initial formation of hematopoietic stem cell niches in bone marrow is tightly controlled by bone modeling/remodeling involving bone-forming osteoblasts and bone-resorbing osteoclasts [10,11]. While the role of osteoblasts located in the endosteal region of bone is well known in the establishment of BM niches, osteoclasts have been mostly studied in HSC mobilization in response to stress or pharmacological stimulants after the establishment of the niche [9,[12][13][14][15] . Recently, osteoclasts have been described as the potential regulators of HSCs niche, besides their bone-resorbing function which provides space for hematopoiesis, since they also provide signals which affect cellular and molecular components of the hematopoietic niche [16]. Osteoclasts regulate the differentiation and function of osteoblasts as well as phenotype of mesenchymal cells, regarded as osteoblast progenitors. Mansour A., et al. demonstrated that osteoclasts are important for the initial niche formation and its colonization by Lin neg Sca1 + cKit + (LSK) hematopoietic stem/progenitor cells in osteopetrotic mice (oc/oc mice) through their impact on the mesenchymal compartment. It seems that the role of osteoclasts is much more complex than just providing space for HSC niche. The lack of osteoclast activity in osteopetrosis results in reduced BM hematopoiesis, extramedullary hematopoiesis in the liver and spleen and BM failure. The development of extramedullary hematopoiesis in osteopetrosis also suggests a strong association between HSC niche formation and functional osteoclasts. However, the exact role of osteoclasts remains obscure and controversial due to the use of different models and conditions [10,17,18]. For this reason, studying osteoclast biology and its impact on hematopoietic stem cell compartment in bone marrow by using patient-derived osteopetrotic iPSCs have unique opportunities, primarily being an unlimited source of pluripotent stem cells carrying patientspeci c disease-causing mutations. This will help to dissect disease pathogenesis and investigate new therapeutic targets through in vitro disease modeling in the laboratory, while paving the way for the development of new transplant strategies to support engraftment in the clinic. The aim of this study is to model the osteoclast defect and to investigate alterations in mesenchymal and HSC compartments of the hematopoietic niche using patient-derived osteopetrotic iPSCs. To our knowledge, modeling osteoclast defect and characterization of HSC compartment and hematopoietic niche using human osteopetrotic iPSCs-derived HSCs and bone marrow-derived mesenchymal stromal cells (BM-MSCs) has not been reported yet.

Generation of Induced Pluripotent Stem Cells from Peripheral Blood Mononuclear Cells
Frozen peripheral blood mononuclear cells which were isolated by density gradient separation from the peripheral blood samples of one healthy donor and three patients with three different TCIRG1 mutations were used in this study (Supplementary Table 1). Immunophenotype of isolated mononuclear cells was evaluated with ow cytometry (CD45 + and CD34 neg , CD36 neg , CD71 neg , CD235a neg ). Erythroid progenitor cells were expanded using Erythroid Expansion Media (EEM) (Stem Cell Technologies), and enrichment of erythroid progenitor cells was demonstrated by ow cytometry on day 10. The data was analyzed using Beckman Coulter Navios EX Flow Cytometer (Kaluza Analysis 2.1 Software). All antibodies, cytokines, and primer sequences were shown in Table 1.
Erythroid progenitor cells were transduced with Sendai viral vector (SeV) at MOI of 5:5:3 in 1 ml fresh EEM (1x10 5 cell)/well). After spinning at 2250 rpm for 90 minutes at 25°C ,1ml EEM was added to well and incubated at 37°C. On day 3, cells were transferred onto Matrigel-coated 10 cm plate in ReproTSR media (Stem Cell Technologies). Cells were fed with ReproTSR (StemCell Technologies) between day 5 and day 11, then maintained in TeSR TM -E8 TM (Stem Cell Technologies) medium starting from day 12. iPSC colonies that started to appear around day 12 were harvested with manual microdissection method, transferred into Matrigel-coated dishes, and were expanded with clump passaging method with EDTA [19].
Characterization of iPSC lines iPSC lines were characterized by colony morphology, immuno uorescence staining, ow cytometry and qRT-PCR. Detection of residual SeV sequences was evaluated in reprogrammed IPS cells using conventional PCR and veri cation of the mutations was assessed by sequencing of the genomic locus. Embryoid body formation was performed to show three lineage differentiation potential in vitro. Three iPSC lines were characterized for each sample.
Immuno uorescence staining was performed using PSC 4-Marker Immunocytochemistry Kit (Life Technologies). In brief, after xation and permeabilization, iPSC colonies were rst incubated with primary antibodies (OCT4, SSEA4, SOX2 and TRA1-60) for overnight, and then with conjugated secondary antibodies for 1 h at RT. Following staining with DAPI, samples were examined under uorescence microscope (Olympus-IX73).
Immunophenotype of iPSCs was evaluated with ow cytometry (BD Accuri C6 Plus). iPSC colonies were harvested with accutase and single cell suspensions were stained using antibodies against pluripotency markers SSEA4 and OCT4, and erythroid progenitor cell markers CD36 and CD235a.
The mRNA expression levels of the pluripotency-associated genes (Endo-OCT4, Endo-SOX2, NANOG, c-MYC, KLF-4, UTF-1, DNMT3b, TERT-1, REX-1, N-Cad,) were analyzed by quantitative real-time PCR (qRT-PCR). Total RNA isolation was conducted by Promega, ReliaPrep™ RNA Cell Miniprep System. qPCR studies were performed using ThermoFisher Maxima SYBER Green qPCR master mix on mic qPCR cycler (Bio Molecular System, v2.6.5). Gene expression was normalized to the expressions of β-actin. The losses of Sendai virus genome (SeV genome sequence targeted) was assessed by end-point PCR using Promega GoTaq® DNA polymerase. Agarose gel electrophoresis results of end-point PCR analysis were presented ( Figure S4). The mutation veri cation after reprogramming was done by sequencing of the target genomic loci in the investigated iPSC lines.
In vitro trilineage differentiation potential of iPSC lines was evaluated with embryoid body (EB) assay. iPSC colonies were harvested using Accutase, single-cell suspensions were prepared at passage 20-24 and cells were seeded onto AggreWell TM 800 plates (StemCell Technologies) in EB formation media (StemCell Technologies) containing 10 µM Rock inhibitor (StemCell Technologies). Following a 24-hour suspension culture, EBs were harvested and transferred to ultra-low attachment plates in STEMdiff TM APEL TM 2 Medium (StemCell Technologies) and kept in culture for spontaneous differentiation. Differentiated EBs on day 7, day 14, and day 21 were characterized by IF, IHC, and RT-PCR. For immuno uorescence staining, cell aggregates were xed for 30 minutes with 4% paraformaldehyde, permeabilized with 1.5% Triton X-100 for 1 hour, re-xed in 4% paraformaldehyde for 15 minutes, then blocked with 2% goat serum for 3 h, and incubated with primary antibodies overnight at 4°C (α-SMA, ab7817; MAP2, ab11267; SOX17 , ab192453, Abcam). Aggregates were then washed three times PBS and incubated with secondary antibody (Goat anti-Mouse IgG H&L (Alexa Fluor 488, abcam, ab150113). Next, aggregates were washed in PBS three times, stained with DAPI for 5 minutes, washed again, and imagined using confocal microscopy (CARLL ZEIS LSM 880).
For immunohistochemistry staining; cell aggregates were dehydrated through a series of graded ethanols, xylene, and para n before embedding into para n blocks. Five-micron sections, were cut and adhered to charged glass slides. Selected sections were then stained with hematoxylin and eosin. The rest of the sections were depara nized, rehydrated, and blocked with endogenous peroxidase, incubated with primary antibodies for one hour, then secondary antibodies for 30 min. After treatment with streptavidin peroxidase, and DAB, hematoxylin staining and dehydration, series of graded ethanols and xylene processes were performed, they were imagined using Olympus microscopy (Table S1). Change in the gene expression pro le of spontaneously differentiated EBs was assessed on day 7, day 14, and day 21 with RT-PCR using transcripts targetting OCT4 and α-SMA, MAP-2 and SOX17 as lineage-speci c markers.

Directed Differentiation of Osteopetrotic Induced Pluripotent Stem Cells to Hematopoietic Precursors and Osteoclasts
IPS cells at passage 20-24 were differentiated into HSCs using STEMdiff TM Hematopoietic Kit (StemCell Technologies) using a step-wise differentiation protocol which starts with induction of primitive streak/mesoderm, then hematopoietic speci cation and followed by hematopoietic cell maturation, myeloid cell expansion and osteoclast lineage differentiation [20]. iPSCs were passaged using EDTA clump passaging method, and 10-20 iPSC aggregates per cm 2 were plated on matrigel coated-6 well plate in TeSR TM-E8 maintenance medium (StemCell Technologies). The cells were then incubated sequentially in STEMdiff™ differentiation medium containing different combinations of cytokines and hematopoietic growth factors (hVEGF, hbFGF, IL-6, IL-3, IL-11 and hSCF) for about 10 days. Then a fraction of the cells were harvested with accutase, and the expression of hematopoietic cell markers (CD43, CD34, and CD45), and also myeloid cell markers (CD11b, CD14, CD16, CD18, CD38, and CD115) were evaluated to show time-dependent up-regulation of lineage-speci c cell surface markers. Rest of the differentiated cells were used in order to continue myeloid-lineage differentiation and to characterize iHSCs following the selection of CD34+ cells with magnetic-activated cell sorting (Miltenyi, Cat # 130-042-201), i.e. methocult assay, immunophenotyping, co-culture experiments, and migration assays. The hematopoietic precursors were cultured for another 4 days in order induce myeloid differentiation using StemSpan SFEM (StemCell Technologies)-based differentiation medium enriched with 2mM Glutamine, 4x10 -4 M monothyioglycerol, 50 µg/ml L-ascorbic acid, 10 ng/ml hVEGF, 4 U/ml EPO, 50 ng/ml TPO, 100 ng/ml hSCF, 10 ng/ml hIL-6, 5 ng/ml hIL-11, and 40 ng/ml hIL-3 (Supplement Table 1). Then a fraction of cells were harvested and stained with CD11b, CD14, CD16, CD18, CD34, CD38, CD43, CD45, CD115 antibodies to demonstrate the e cacy of myeloid differentiation with ow cytometry. Osteoclast differentiation protocol was applied to the remaining fraction of the cells.
Hematopoietic colony-forming potential of CD34+ iPSC-derived HSCs (iHSC) were assessed with semisolid Methocult cultures (StemCell Technologies, MethoCult TM H4434 Classic). 4x10 4 cells were seeded per 35-mm culture dishes in 1,5 ml of methylcellulose, and cultured for 14 days. Colony types were determined by shape, cell size, and extent of visible erythoid content on Olympous IX73 microscope as BFU-E, CFU-G, CFU-M, CFU-GM, and CFU-GEMM. The colony counts were scored and averaged between duplicates. The myeloid lineage speci cation and expansion of the cells were con rmed with lineagespeci c staining with CD45, CD14, CD16, CD18, CD33, CD36, CD41 by ow cytometry.
At the nal step, myeloid precursors were harvested and 1x10 5 cells/well were seeded onto Corning Osteo Assay Surface Microplate (Corning Cat#3988) in IMDM medium containing 10 ng/ml M-CSF, 10 ng/ml RANK-L, and 10% FBS. They were cultured for 21 days for induction of differentiation into functional osteoclasts. Besides immunophenotyping of the cells with monocyte-macrophage and osteoclast precursor cell markers (CD14, CD16, CD18, CD45, CD51/61), IF staining was performed and cell morphology was evaluated with scanning electron microscope (SEM FEI QUANTA 200F). Gene expression pro le of iPSCs derived osteoclasts was also analyzed.
The functionality of iPSC-derived osteoclasts was assessed by the presence of tartrate-resistant acid phosphatase activity (TRAP) activity in the multinucleated cells by immuno uorescence staining. Active functional osteoclasts were also identi ed by the formation of Filamanteous (F)-actin rings that were demonstrated by immunostaining of actin cytoskeleton with the Phalloidin conjugated to Rhodamine, and secretion of cysteine protease, cathepsin K.
Cathepsin K, NFATC1, Calcitonin R as osteoclast markers genes were analyzed with qRT-PCR at the indicated time points throughout the differentiation. The target gene normalized to the β-actin endogenous control, 2 ∆∆CT value was calculated considering negative control i.e. donor myeloid precursor cells.
Co-culture of hMSCs and iPSC-derived Hematopoietic Stem Cells and Migration Assay Mesenchymal stem cells (MSCs) were isolated from bone marrow samples of one healthy donor and three patients with three different TCIRG1 mutations, coculture experiments and migration assays were performed after the characterization of MSCs. The MSCs were characterized by ow cytometry (as CD29 + , CD44 + , CD73 + , CD90 + , CD105 + and CD34 neg , CD45 neg ). Adipogenic and osteogenic differentiation potential of hMSCs were evaluated for each established cell line. For induction of adipocyte differentiation, cells were treated with adipogenic differentiation medium (DMEM LG (Gibco) supplemented with 10% FBS (Gibco), 1μM dexamethasone (Sigma), 60 μM indomethacin (Sigma), 500μM 1-methyl-3-isobutylxanthine (IBMX, Sigma), and 5μg/ml insulin (Sigma-Aldrich) for 21 days. Then they were xed and stained with Oil Red O stain to visualize fat droplets within the differentiated cells. For osteoblastic differentiation, cells were treated with osteoblast induction medium (OB) containing DMEM-LG (Gibco) supplemented with 10 % FBS (Gibco), 100 nM dexamethasone (Sigma), 10 mM beta glycerophosphate (Sigma), 0,2 mM ascorbic acid (Sigma) for 21 days, and Alizarin Red staining was performed to signify the osteogenic differentiation. Quanti cation of qRT-PCR signals was performed using the (2-∆∆Ct ) method [21] , which calculates relative changes in gene expression of the target gene normalized to the β-actin endogenous control. The values obtained were represented as the relative quantity of mRNA level variations.
Migration assays were performed to evaluate the regulatory role of MSCs as osteoblast progenitor cells on migration and homing of iHSCs to the hematopoietic niche. 1X10 6 MSCs were plated in the bottom chamber of 5-µm pore size transwells (Costar) with MEM-α medium containing 10% FBS and incubated at 37°C and 5% CO 2 . After 24h, MSCs were labeled with CellTracker TM Green CMFDA (ThermoFisher), and CD34+ iHSCs labeled with CellTracker TM Orange CMTMR (ThermoFisher) were plated in transwell inserts. After 4h of incubation at 37°C and 5% CO 2 , remaining cells at the transwell inserts were removed and cells migrated to the bottom chambers were visualized and quanti ed by Olympus IX73 uorescence microscope.
Statistical analysis qRT-PCR, migration, coculture and differentiation assays were performed in three biological replicates for patients and one donor, and qRT-PCR was performed in two replicates. The median values of the patient group and the donor were compared with unpaired two-tailed T test , and patient groups compared each other with paired two-tailed T test. Friedman multiple comparisons test, and One sample Wilcoxon test were used for time-course experiments. Kruskal Wallis multiple comparisons test was applied to independent multiple groups comparisons. Analysis of variance was conducted on the replicate values of experiment groups. P values <0.05 were accepted as statistically signi cant. The data was analyzed using GraphPad Prism 8.01.

Generation of Induced Pluripotent Stem Cells from Peripheral Blood Erythoid Progenitors of Osteopetrotic Patients
Peripheral blood mononuclear cells (PBMCs) obtained from three patients with three different mutations of TCIRG1 gene and one healthy donor were enriched for erythroid progenitor cells before reprogramming with CytoTune Sendai reprogramming kit. Enrichment e ciency of erythroid progenitors derived from the patients' samples was con rmed by ow cytometry [CD36 (95,9 % ±1,6), CD71 (98,5% ± 0,15), and CD235a (72,4% ± 8,78) (see Supplementary Figure 1A and 1B)]. The rst osteopetrotic iPSC colonies were observed at around day 12-15 after transduction. Eight to ten iPSC colonies from each iPSC line established were picked manually and expanded in culture for detailed analyses (see Supplementary Figure S1-C).
Tri-lineage differentiation potential of generated iPSC lines were evaluated with embryoid body (EB) formation and ability of iPSC-derived EBs to give rise to early progenitors belonging to each of three germ layers. Spontaneously differentiated EBs were stained triple-positive with lineage-speci c MAP2 (ectoderm, neural lineage marker), α-SMA (mesoderm, mesenchymal stromal cell marker), SOX17 (endoderm, fetal hematopoetic stem cell marker) on day 21 by immuno uorescence staining (Fig. 1d) Lineage-speci c gene expression pro les were shown in Supplementary Figure 2B, 2C, 2D.

Osteopetrotic iPSC-derived Myelomonocytic Precursors Generate Dysfunctional Osteoclasts
We next investigated whether myelomonocytic precursors derived from osteopetrotic iPSCs had the potential to differentiate further into mature osteoclasts. iPSCs-derived myelomonocytic precursors were seeded onto multiple-well plates coated with a synthetic inorganic bone mimetic and cultured in the presence of M-CSF and RANKL, which are essential cytokines for osteoclasts differentiation and survival. Change in cell morphology and size was observed throughout the culture and multinucleated cells were formed at the end of the differentiation period (Fig. 3a). Osteopetrotic-iPSCs-derived multinucleated osteoclasts were phenotypically similar to healthy donor osteoclasts and demonstrated similar expressions of CD14 (Donor 98,3%, Patient 99,1 ± 0,3%), CD16 (Donor 98,2%, Patient 88,3 ± 3,7%), CD18 (Donor 96,6%; Patient 99,5 ± 0,5%), CD51/61 (Donor 100%; Patient 78,8 ± 7,6%), which are speci c surface markers for osteoclast precursors (Fig. 3b). The identity and functionality of differentiated osteoclasts were con rmed by the presence of TRAP activity and secretion of a cysteine protease Cathepsin K, enzymes highly expressed in mature osteoclasts. We further examined the formation of Factin rings which are speci c cytoskeletal structures of resorbing active osteoclasts (Fig. 3c). TRAPpositive osteopetrotic-osteoclasts had weak actin ring formation and decreased Cathepsin K secretion compared to donor osteoclasts. The electron microscope images indicated a difference between the podosome sizes of osteopetrotic and donor multinucleated osteoclasts (Fig. 3c). There was a signi cant reduction in the expression levels of osteoclast-speci c genes; Cathepsin-K, Calcitonin-R, and NFATC1 in osteopetrotic osteoclasts as compared with donor osteoclasts (Fig. 3d). Overall, osteopetrotic osteoclasts which are phenotypically identical to healthy donor osteoclasts were found functionally defective.

Impaired Differentiation Potential, Homing and Retension of Osteopetrotic iHSCs Recover After Co-culture with Healthy MSCs
Mesenchymal stromal cells are critical components of the hematopoietic niche and they are the progenitors of bone-forming osteoblasts which are required for hematopoietic niche formation. Osteoclasts induce osteoblastic commitment and differentiation of MSCs ). Thus, alterations in the HSC compartment in the bone marrow of osteopetrosis patients could be related to modi cations in MSC compartment. To test this hypothesis, we established a co-culture system with iHSCs and bone marrow-MSCs of osteopetrosis patients to model niche compartments in vitro. First, bone marrow-MSCs were characterized by morphology, immunophenotyping and assessment of their differentiation potency. All MSCs lines expressed speci c surface markers of mesenchymal stromal cells and they were negative for hematopoietic markers (see Supplementary Figure 3A). Diminished osteogenic differentiation potential was detected in osteopetrotic BM-MSCs, while they were successfully differentiated into adipocytes (see Supplementary Figure 3B). When immunophenotypes of CD34enriched healthy iHSC co-cultured with healthy MSCs (as in healthy niche) and osteopetrotic iHSCs cocultured with osteopetrotic MSCs (as in osteopetrotic niche) were compared, we found a signi cant difference among the expressions of myelomonocytic progenitor and differentiation markers, con rming On the other hand, healthy iHSCs co-cultured with osteopetrotic MSCs showed a myeloid skewed phenotype compared to their phenotype after co-culture with healthy MSCs (CD34+CD43+CD45+, 42,7% ±2,3 vs. 27,8% ±3,1; CD14+, 37,3% ±2,7 vs. 19,7%± 2,9 ; CD18+, 46,4% ±3,4 vs. 33,5% ±3,2 P<0,05).
Altogether these data demonstrate that there is an alteration in the hematopoietic differentiation potential of osteopetrotic HSCs. This alteration could be a rescue mechanism for defective osteoclast function in osteopetrosis. Osteoclasts are tightly coupled with osteoblasts and modulate differentiation and function of osteogenic progenitor cells namely MSCs, and they are essential for healthy niche formation. The restoration of the myeloid-skewed phenotype of osteopetrotic iHSCs after co-culturing with healthy MSCs supports the impairment in osteopetrotic MSCs (Sobacchi, 2013 To gain more insight into the regulatory role of MSCs in HSC niche, we analyzed the expression of genes controlling HSC homing, survival, quiescence and adhesion such as Ang-1, Kit ligand (Kit-L), Sdf-1, Jag-1, Opn and N-Cad in osteopetrotic MSCs after co-culture with osteopetrotic iHSCs. Osteopetrotic MSCs showed lower expression of Ang-1, Jag-1, Sdf-1, Opn, and higher expression of N-Cad when co-cultured with osteopetrotic iHSCs in comparison with the healthy MSCs co-cultured with healthy iHSCs. Interestingly, the expression of Jag-1, Kit-L, and Opn was signi cantly increased and N-Cad decreased after coculture with healthy iHSCs (Fig. 4b). These results strongly suggest an impairment in MSCs function in supporting HSC homing and maintenance.
Finally, we performed in vitro migration assay to investigate the ability of osteopetrotic MSCs to attract HSCs. A dramatic reduction in attraction ability of osteopetrotic MSCs was observed for both osteopetrotic and healthy iHSCs compared to healthy MSCs. Contrarily, the migratory potential of osteopetrotic-iHCSs was signi cantly recovered after co-culture with healthy MSCs. (Fig. 4c,and d). Accordingly, these results con rm that altered MSCs do not support HSCs homing and retention in osteopetrosis.

Discussion
Enchondral ossi cation which is essential for the initial formation of hematopoietic stem cell niches in bone marrow is tightly controlled by bone modeling/remodeling involving bone-forming osteoblasts and bone-resorbing osteoclasts [10,11]. The absence of bone resorption is associated with a decrease in bone marrow space for HSCs as implicated by different osteopetrotic mouse models [11,22,23]. But a decrease in bone marrow space is not the only reason of reduced BM hematopoiesis and extramedullary hematopoiesis in osteopetrosis patients since they differ by their osteoclast and HSC contents despite presenting with defective bone resorption [6,24]. Recently, osteoclasts have been described as the potential regulators of HSCs niche, besides their bone-resorbing function which provides space for hematopoiesis, since they also provide signals which affect cellular and molecular components of the hematopoietic niche [16]. They regulate differentiation and function of mesenchymal cells representing osteoblast progenitors, and MSCs are one of the essential components of HSC niche. Extramedullary hematopoiesis associated with osteopetrosis strongly supports a link between osteoclast function and HSC niche [8,25]. The investigation of the roles of osteoclasts and osteoblasts in HSC niche formation in oc/oc mouse model demonstrated an increased proportion of mesenchymal progenitors with reduced osteoblastic commitment leading to impaired HSC homing which was recovered with the restoration of osteoclast function [10]. Considering clinical heterogeneity in osteopetrosis especially among patients with different mutations, patient-derived iPSCs carrying disease-causing mutation within the context of the patient's whole genome could be an ideal platform to investigate osteoclastogenesis and HSC niche in osteopetrosis [1].
In this study, we reprogrammed peripheral blood mononuclear cells of osteopetrosis patients carrying different mutations in TCIRG1 gene and successfully generated osteoclasts from patient-derived iPSCs.
They were phenotypically identical to healthy donor osteoclasts but functionally defective as expected since TCIRG1 mutation affects osteoclast function due to impaired ru ed border formation [2,26].
NFATc1 is a master transcription regulator which is responsible from terminal osteoclast differentiation. It promotes osteoclast-speci c gene expression which is required for differentiation and bone resorption, such as Calcitonin receptor, Cathepsin K and TRAP [27,28]. We observed thinner actin ring formation, weak TRAP activity and decreased Cathepsin K secretion correlated with decreased NFATC1 expression in osteopetrotic-osteoclasts comparing with donor osteoclasts. This data indicates that osteoclasts from patients with TCIRG1 mutation could not differentiate into actively resorbing mature ostoclasts [29]. Interestingly, when patient-derived iPSCs were differentiated into hematopoietic stem/progenitor cells, they showed an increase in macrophage colony-forming ability (CFU-M) compared with healthy donor iHSCs. Further differentiation of osteopetrotic iHSCs into myeloid progenitors gave rise to more monocyte-macrophage progenitors, revealing myeloid-skewed hematopoietic differentiation potential of osteopetrotic iPSCs which leads to enrichment of monocyte-macrophage progenitors. Augmented intraand extramedullary myelopoiesis together with an increase in osteoclastogenesis was reported in oc/oc mice [22]. The upregulation of the myelomonocytic differentiation of osteopetrotic iHSCs may present a compensatory mechanism related to the lack of bone-resorbing active osteoclasts. Since hematopoieis is regulated by bone marrow niche and stress conditions such as irradiation, infections or drugs change heterogeneity of hematopoietic stem/progenitos cells and hematopoietic hierarchy [30,31] myeloid skewing in osteopetrotic iHSCs probably develops as a response to perturbation in the balance between bone-resorbing and bone-forming cells in osteopetrosis.
Osteoclasts control the fate of mesenchymal cells by inducing their differentiation into osteoblasts which is essential for healthy niche formation in bone marrow, as evaluated by the improvement of hematopoiesis after reconstitution of osteoclast activity in osteopetrotic mice [8,25,32]. To better understand how defective osteoclast activity can cause alterations in HSC niche contents and impairment of hematopoiesis, we performed hematopoietic niche modeling with an ex vivo co-culture system of iPSC-derived HSCs and bone marrow-MSCs of osteopetrosis patients carrying TCIRG1 mutation.
We demonstrated that osteopetrotic MSCs had reduced osteogenic differentiation potential, and expressed Ang-1, Sdf-1, Jag-1, Opn, the genes associated with HSC homing and maintenance in bone marrow at low levels [32][33][34][35][36]. Furthermore, their ability to attract HSCs dramatically impaired. The decreased expression of main regulators of HSC proliferation, quiescence, and homing in osteopetrotic MSCs as well as their impaired attraction ability over HSCs are most likely the mechanisms behind defective HSC homing, and loss of hematopoietic progenitors in bone marrow of osteopetrosis patients.
Interestingly, N-cadherin expressed in bone marrow stromal progenitors that are essential for reservation of HSCs in niche especially under stress, was found overexpressed in osteopetrotic MSCs. This could be a stress response to impaired hematopoiesis in osteopetrosis in order to support HSC maintenance in bone marrow niche [37]. In addition, the expression of Jag-1, Kit-L, and Opn in osteopetrotic MSCs was signi cantly increased and N-Cad decreased upon co-culture with healthy iHSCs. These results strongly suggest a functional impairment in MSCs regarding their role in HSC homing and maintenance and con rm their role in defective hematopoietic niche formation in osteopetrosis through their interaction with dysfunctional osteoclasts. The restoration of function of osteopetrotic MSCs and altered phenotype of osteopetrotic hematopoietic progenitors after co-cultured with their healthy counterparts point out that osteoclasts play a key role in all these modi cations. This data may be a proof of concept for clinical studies which evaluate the feasibility/safety of MSC co-transplantation in osteopetrosis patients undergoing allogeneic hematopoietic stem cell transplantation in order to facilitate engraftment through their support on the recovery of abnormal bone marrow niche and HSC homing. Declarations 32. Singh, P., K.S. Mohammad, and L.M. Pelus, CXCR4 expression in the bone marrow microenvironment is required for hematopoietic stem and progenitor cell maintenance and early hematopoietic regeneration after myeloablation. Stem Cells, 2020. 38 (7) Figure 1 Characterization of established iPSC lines derived from patient and donor peripheral blood erythroid precursors. a) Immuno uorescence and ALP staining of three different clones of each of the selected iPSCs lines (n=12). b) Flow cytometry analysis with pluripotency markers, OCT4, SSEA4 and erythroid progenitor cell surface markers, CD36 and CD235a after reprogramming (n=12). c)Veri cation of persistence of three different TCIRG1 mutations in osteopetrotic iPSCs (n=3). d) Whole-Mount immunostaining of spontaneously differentiated embryoid bodies with lineage-speci c markers; MAP2 (ectoderm), α-SMA (mesoderm), SOX17 (endoderm) (n=4).

Figure 2
Assessment of hematopoietic differentiation potential of iPSC lines. a) Differentiated blood-forming iHSC colonies with the formation of radial sac-like structures. b) Flow cytometry analyses of osteopetrotic iPSCs that are differentiated to hematopoietic progenitor cells as indicated by an increase in CD34+CD45+ cell fraction by day 10 following hematopoietic induction coupled with low expression of myelomonocytic markers (CD14, CD16, CD18, 11b, CD38, and CD43). Hematopoietic differentiation potential of patient -and donor-derived iPSCs was evaluated with unpaired two-tailed T test. c) Con rmation of the clonogenic potential of iHSCs by Colony Forming Unit assay Donor iPSCs derived iHSCs colonies (upper row), patient iPSCs derived iHSCs colonies (lower row), the mean number of different types of hematopoietic colonies were given at the right corner. d) Con rmation of hematopoietic lineage commitment of cells forming colonies by ow cytometry after methocult assay. e) Flow cytometry analysis of myeloid differentiated osteopetrotic iHSCs at Day 14 as indicated by an increase in myelomonocytic cell surface markers following myeloid induction (Unpaired two-tailed T test for comparison of patient and donor iPSCs derived myeloid cells, paired two-tailed T test for comparison of patient-iHSCs and patient iHSCs derived myeloid cells. P<0.05). f) Morphological evaluation of myeloid differentiated cells (left-top 10x magni cation, main gure 4x magni cation).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.