Development of an ex vivo xenogeneic bone environment producing human platelet-like cells

The efficiency of in vitro platelet production is considerably low compared with physiological activity due to the lack of pivotal factors that are essential in vivo. We developed an ex vivo platelet production system, introducing human megakaryocytes into an isolated porcine thighbone and culturing in closed circuit. The efficiency of the ex vivo platelet production system was compared to those in vivo and in vitro. CD61+ platelet-like cells were counted by immunostaining and flow cytometry. Results showed that 4.41 ± 0.27 × 103 CD61+ platelet-like cells were produced by 1 × 103 megakaryocytes in the ex vivo system, while 3.80 ± 0.87 × 103 and 0.12 ± 0.02 × 103 were produced in the in vivo and in vitro systems, respectively. Notably, ex vivo and in vitro production systems generated cells that responded well to thrombin stimulation and expressed functional molecules, such as CD62P. Overall, our ex vivo production system was comparable to in vivo production system and produced platelet-like cells that were functionally superior to those produced in vitro. In future, the present ex vivo production system implementing xenogeneic bone marrow would offer a promising alternative for industrial-scale production of platelet-like cells.


Introduction
Several in vitro platelet production systems have been proposed by mimicking in vivo environment [1][2][3][4][5]. Physiologically, platelets are hematopoietic-lineage cells. CD34 + hematopoietic stem cells differentiate into mature megakaryocytes in the bone marrow niche after multinucleation and cytoplasm enlargement [1,2]. Subsequently, the mature megakaryocytes migrate in the proximity of bone marrow sinusoids and release proplatelets, which further mature in the sinusoids through the influence of shear stress due to local turbulences, and platelets are secreted into the circulation [3][4][5]. Along the process, cytokines, scaffold, and intercellular interactions are crucial [6][7][8][9]. The number of platelets produced in vivo ranges between 1,000 and 5,000 per megakaryocyte; however, the efficiency of in vitro production is still low despite a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 considerable efforts and the implementation of thrombopoietin, 3D bioreactors, and artificial turbulence [10,11].
Thrombopoietin, a cytokine essential for inducing differentiation of hematopoietic stem cells to megakaryocytes, was first cloned in 1994, enabling highly efficient megakaryocyte production [12][13][14][15][16]. Matsunaga et al. produced 1.68 × 10 11 platelets from 5 × 10 6 CD34 + hematopoietic stem cells in vitro by optimizing the types and amounts of cytokines added to umbilical cord blood-derived CD34 + hematopoietic stem cells [17]. Brent et al. produced 14.2 platelets per CD34 + hematopoietic stem cell by creating a 3D bioreactor, adding cytokines, and applying shear force [18]. Recently, Ito et al. produced 50-100 platelets per megakaryocyte using an immortalized megakaryocytic cell line established from induced pluripotent stem cells and a bioreactor that imitates turbulence generated in bone marrow vessels [19]. These produced platelets were morphologically like platelets in vivo and expressed functional molecules in response to induction using ADP.
However, abundant pieces of evidence showed that the platelet production efficiency was low as compared with that in vivo. Noteworthy, the incomplete reproduction of cytokines and the scaffolding environment involved in platelet production in the bone marrow niche in vivo is a cause of low production efficiency of platelets in vitro [19,20]. Therefore, improvement of production efficiency in vitro may be achieved by adding factors effective for platelet production based on analysis of the niche environment in vivo.
Herein, we aimed to develop an ex vivo production system that has both in vitro and in vivo characteristics, using isolated porcine bone as a site for production. Moreover, the system will elucidate the detailed mechanisms of platelet differentiation.
In vitro platelet production system CFSE-labeled megakaryocytes were incubated at a density of 5 × 10 6 cells/mL at 37˚C for 3 hours, and then platelet-like cells derived from the megakaryocytes in the culture supernatant were collected.

In vivo platelet production system
The protocol for the in vivo experiment was conducted with the approval of the Laboratory Animal Ethics Committee of the National Center for Child Health and Development (IRB number: A2000-001) based on the Japanese Guideline for Animal Experiments of Ministry of Health, Labour and Welfare.
We used micro-mini pigs not exceeding 30 kg in weight at the age of 2 years [21]. Twelvemonths-old female micro-mini pigs of were purchased from Fuji Micra, Inc., Shizuoka, Japan. Animals were treated per the Animal (Scientific Procedure) Protection Act 1986 of the United Kingdom. The pigs were housed in cages under temperature and light-controlled conditions (12-hour light/dark cycle) and were provided with food and water ad libitum. The pigs fasted for 12 hours before surgery with free access to water. Immunosuppressed pigs were prepared as follows: the micro-mini pigs were intravenously administered mycophenolate mofetil at a dose of 60 mg/kg B.W. daily 5 days before the experiment and tacrolimus at a dose of 0.5 mg/ kg B.W. daily 3 days before the experiment. Sedation with a mixture of midazolam/medetomidine/butorphanol was followed by endotracheal intubation and mechanical ventilation. Anesthesia was maintained with inhalational isoflurane. Midazolam and medetomidine were added according to the depth of anesthesia. For the formation of the perfusion pathway to introduce the cells, a hole of 2-mm diameter was drilled at two points in the epiphysis of the thighbone of anesthetized pigs. An 18-G needle attached to a 10-mL syringe with saline was pierced into the hole, and the saline was perfused into the bone marrow with positive pressure while applying negative pressure from another side (Fig 1). Next, 500 μL of CFSE-labeled megakaryocytes (2.5 × 10 6 cells) were introduced into the thighbone using a 1-mL syringe, followed by 1 mL of saline to remove megakaryocytes in the perfusion line. The introduced megakaryocytes were incubated for 3 hours and subsequently harvested. Buprenorphine was administered as an analgesic for intraoperative pain management. After intravenous administration of pentobarbital saturated potassium chloride was rapidly administered intravenously to euthanize.

Ex vivo platelet production system
The protocol for the ex vivo experiment was conducted with the approval of the Experimental Animal Ethics Committee of Keio University (IRB number: 14709-(0)) based on Institutional Guidelines on Animal Experimentation at Keio University. The thighbone used for the ex vivo production system was harvested from livestock pigs at 3 months old. All animals were also treated in accordance with the Animal (Scientific Procedure) Protection Act 1986 of the United Kingdom as described above in "In vivo platelet production system". The excised thighbone was covered with quick-drying epoxy putty (DHP-482, Loctite) and dried for 20 minutes at room temperature. A hole of 1.3-mm diameter was drilled at two points in the epiphysis of the thighbone. Then, an 18-G needle attached to a 10-mL syringe (syringe B) with perfusion medium was pierced into one hole, and 300 mL of perfusion medium was perfused into the bone marrow with positive pressure at a flow rate of 7 mL/min using a syringe pump while applying negative pressure from the other side (syringe A) using a syringe pump (Fig 2). Next, 500 μL of CFSE-labeled megakaryocytes at a density of 10-40 × 10 6 cells/mL was introduced into the thighbone and incubated for 3 hours at 37˚C for production of platelet-like cells. Following ex vivo incubation, 120 mL of perfusion medium was perfused through one hole in the thighbone, and platelet-like cells were collected from the other hole.

Immunofluorescence of megakaryocytes and immunohistochemistry of platelet-like cells in porcine thighbone
Immunomorphological characterization evaluation of megakaryocytes derived from CD34 + cells on the 19th day, cytospin preparation began with 2×10 5 cells in PBS containing 5% bovin serum albumin and centrifuged at 120 × g for 3 minutes and then transferred onto glass slides. Cytospin preparation was fixed by PBS containing 4% paraformaldehyde (PFA) for 10 minutes at room temperature and then incubated with 100 μL of antibody solution (Alexa Fluor 647-labeled anti-CD61 antibody (1:100, BioLegend, clone: VI-PL 2), containing 0.1% Hoechst33342 (Dojindo) in PBS for 10 minutes at room temperature. After washing with PBS, the cytospin preparation was embedded with 50% glycerol in PBS and observed under fluorescence microscope (BZ-X 700, Keyence).
The cancellous bone marrow tissue harvested from the porcine thighbone were embedded in Tissue-Tek OCT compound (Sakura Finetek) and then frozen in dry ice. Frozen tissue sections with a thickness of 10 μm were prepared using a cryostat (Leica Biosystems), immersed in saline for 10 minutes, and then fixed by PBS containing 4% PFA for 10 minutes at room temperature. After washing with saline, the sections were incubated with 100 μL of antibody solution (Alexa Fluor 647-labeled anti-CD61 antibody (1:100), containing 0.1% Hoechst33342 in PBS for 10 minutes at room temperature. After washing with PBS, the sections were embedded with 50% glycerol in PBS and observed under fluorescence microscope, and the number of CD61 + platelet-like cells was counted.

Flow cytometric analysis of platelet-like cells
This study was approved by Sysmex Ethics Committee. All participants provided their written informed consent to participate in this study according to the study protocol. To define a platelet FSC-SSC scattergram, platelets from healthy volunteer donors were first evaluated by flow cytometry. Blood samples were placed into blood collection tubes containing acid-citratedextrose (Becton, Dickinson and Company) and then centrifuged at 200 × g for 10 minutes to prepare platelet-rich plasma (PRP). Then; 5 μL of PRP was added to 95 μL of PBS and evaluated using FACSVerse (BD Biosciences).
For platelet marker analysis, the cells harvested from ex vivo, in vitro, and in vivo production systems were fixed with 1% PFA in PBS for overnight and centrifuged at 200 × g for 10 minutes. After the supernatant was collected, the supernatant was centrifuged at 1500 × g for 10 minutes and washed with PBS. Then 5 μL of antibody (allophycocyanin-labeled anti-CD42b antibody, Alexa Fluor 647-labeled anti-CD61 antibody, allophycocyanin-labeled mouse IgG1 antibody, or Alexa Fluor 647-labeled mouse IgG1 antibody (BioLegend, clone: MOPC-21)) was added to 100 μL of cell suspension that was incubated for 15 minutes at room temperature. After the reaction, the cells were washed with PBS, and evaluated using FACS-Verse. The number of platelet-like cells was calculated using counting beads as external standard (C36950, Thermo Fisher SCIENTIFIC).

Results
CD34 + cells were differentiated into megakaryocytes and measured megakaryocyte marker over time. CD42b + CD61 + cells were detectable at day 7 and increased to 56.7% at day 19 (S1A, S1C and S1D Fig). In addition, on day 19, the megakaryocytes had undergone polyploidization, and some were proplatelets (S1B Fig). Thus, we used the CFSE-labeled cells at day 19 as megakaryocytes to evaluate the platelet differentiation by in vitro, in vivo, and ex vivo production system.
In all experiments, platelet-like cells derived from CFSE-labeled megakaryocytes were defined as cells that are same size as platelets and positive for CFSE (Fig 3A). Initially, cells collected by in vitro production system were assessed by flow cytometry. In vitro, the rate of platelet-like cells among all collected CFSE + cells was 66.3 ± 11.9% (Table 1, Fig 3B and 3C). In addition, the rates of CD61 + and CD42b + platelet-like cells in all platelet-like cells were 79.7% and 5.8%, respectively. As for platelet function, the rate of CD62P + platelet-like cells with or without thrombin stimulation was determined. CD62P + platelet-like cells were increased from 32.2% under no stimulation to 45.3% under thrombin stimulation (Table 1). Furthermore, the production number of CD61 + platelet-like cells was 124 ± 22 per 1 × 10 3 megakaryocytes (Table 1).
In in vivo, the production number of platelet-like cells was evaluated by immunohistochemical staining of the harvested thighbone. CD61 + platelet-like cells were observed around introduced megakaryocytes (Fig 4). Two porcine were used for the experiment, and the number of CD61 + platelet-like cells produced from the 1 × 10 3 megakaryocytes introduced for each was 4666 and 2923 (Table 1). When sections of bone marrow at both ends and center were observed, megakaryocytes and platelets were present at all sites (S3 Fig). Therefore, this suggest that the introduced megakaryocytes are distributed throughout the bone marrow. Besides, platelet-like cells were not detected from peripheral blood.
We developed an ex vivo platelet production system utilizing a natural biological environment, porcine thighbone, in which a perfusion line was formed to introduce and collect cells. To verify the production efficiency of platelet-like cells from megakaryocytes in the ex vivo, the properties and number of produced platelet-like cells were evaluated and compared with those of the in vitro and in vivo. Initially, cells collected by the ex vivo production system were assessed by flow cytometry. Most of the collected cells were pig-derived blood cells, and the introduced megakaryocyte-derived CFSE + cells were 0.85% ± 0.73% (Table 1 and S2 Fig). The rate of platelet-like cells in all collected CFSE + cells was 63.0 ± 19.9% (Table 1 and Fig 3D). In addition, the rates of CD61 + and CD42b + platelet-like cells in all platelet-like cells were 76.9% and 14.0%, respectively ( Table 1). The number of CD61 + platelet-like cells produced from Table 1  1 × 10 3 megakaryocytes was 65 ± 34, and the rate of CD62P + platelet-like cells in all of plateletlike cells increased from 53.8% to 62.5% by thrombin stimulation (Table 1). Therefore, the platelet-like cells produced by the ex vivo production system responded to the stimulation. Next, platelets in the bone marrow were evaluated. The number of platelet-like cells was evaluated by immunohistochemical staining of bone marrow after ex vivo production. Similar to the results of the ex vivo production system, CD61 + platelet-like cells were observed around introduced megakaryocytes, and the number of CD61 + platelet-like cells produced from 1 × 10 3 megakaryocytes was 4411 ± 271 (Table 1, Fig 5). When the bone marrow sections at both ends and center were observed, megakaryocytes and platelets were present at all sites (S3 Fig). Therefore, this suggest that the introduced megakaryocytes are distributed throughout the bone marrow. These results showed that platelet-like cells produced by the ex vivo production system had a CD61 + rate equal to that obtained by the in vitro production system and a higher CD42b + rate than that achieved with the in vitro production system. Furthermore, the number of platelet-like cells produced by the ex vivo production system was higher than that produced by the in vitro production system and equal to that produced by the in vivo production system (Table 1). These results suggested that the ex vivo production system more efficiently produces CD61 + and CD42b + platelet-like cells than the in vitro production system, and it is possible to collect and analyze platelet-like cells produced with the same efficiency as the in vivo production system.

Discussion
In this study, we compared the efficiency of in vitro, in vivo and ex vivo platelet production systems. The number of platelet-like cells produced in vitro were counted by flow cytometry, while the platelet-like cells produced in vivo and ex vivo were counted under a microscope. It is not common to count the number of produced platelet-like cells using microscopic images. When platelets of volunteer blood were introduced and collected in the ex vivo production system, the collection efficiency was 0.91%. On counting by the flow cytometry method and the recovery efficiency, the number of platelet-like cells produced in the thighbone was 7.10 × 10 3 per 1 × 10 3 megakaryocytes. This number is similar to the number counted by microscopy. Therefore, the method of counting the number of platelet-like cells produced using a microscope image is considered to be valid. The number of CD61 + platelet-like cells produced per Thus, the in vivo and ex vivo and production systems can produce platelet-like cells with higher efficiency than the in vitro production system. This result suggests that factors that promote production from megakaryocytes to platelets exist in the bone marrow environment used in the in vivo and ex vivo production systems. In addition, in the ex vivo production system, cells are quiesced after cell introduction to induce differentiation, and thus, it is considered that the cells are not subjected to mechanical stimulation such as turbulence. Therefore, we believe that biological factors in the porcine bone marrow environment promoted platelet production. The ex vivo production system produced platelet-like cells with a higher CD42b + rate than the in vitro production system and at a number comparable to that of the in vivo production system. In the in vitro production system, CD42b is known to be cleaved by ADAM-17 when the production of megakaryocytes to platelets is carried out at 37˚C. Therefore, it is challenging to obtain platelets highly expressing CD42b [22]. On the other hand, because the ex vivo production system produced platelet-like cells with a high CD42b + rate, there may have been factors inhibiting the cleavage of CD42b, including by ADAM-17, in the bone marrow environment or factors that enhanced the expression of CD42b.
Matsunaga et al. reported that a single megakaryocyte differentiated from umbilical cord blood-derived CD34 + cells produced 4 platelets in 5 days by using in vitro production system [17]. Nakamura et al. reported that a single megakaryocyte differentiated from iPS cells produced 3-10 platelets in 5 days from megakaryocytes induced to differentiate by using in vitro production system [23]. Tozawa et al. reported that a single megakaryocyte differentiated from adipose-derived mesenchymal stem/stromal cells produced 5-10 platelets in 12 days by using in vitro production system [24]. These are calculated by the production efficiency per 3 hours, and it is shown that they produce 100, 70-250, and 52-104 platelets per 1 × 10 3 megakaryocytes, respectively. In other words, the platelet production efficiency was comparable to that of the in vitro production system. Moreover, Matsubara et al. reported that a single megakaryocyte differentiated from adipose-derived mesenchymal stem/stromal cells transplanted to mice produced 5-10 platelets 3 hours after transplantation [25]. The production efficiency was suggested to be about the same as in vivo production system in this study also produces 2-5 platelets per megakaryocyte in 3 hours. Therefore, ex vivo production system has almost the same platelet production efficiency as existing in vivo production system and can produce more platelets at the same time, as compared with existing in vitro production systems. It is known that platelets collected from a living body deteriorate in about 4 to 5 days and fail to meet blood transfusion criteria. Therefore, when platelets are produced ex vivo, a system for producing them in copious amounts in a brief time is required. The platelet production efficiency per 3 hours of the ex vivo production system we developed is 35 times that of the in vitro production system, so it is useful as a system for producing functional platelets. However, the collection efficiency of the produced platelet-like cells was low with the ex vivo production system. By comparing the number of platelets collected from the thighbone by perfusion with that introduced into thighbone, the collection efficiency of platelets by perfusion was found to be only 0.91%. Furthermore, CFSE + platelet-like cells rate of all collected cells from ex vivo production system was 0.44 ± 0.32%. Therefore, even if the collection efficiency is 100%, only 33% of the cells are human platelet-like cells, and most of them are cells derived from pig. To collect only human platelet-like cells, a technique for separating pig cells and human platelet-like cells is required. Therefore, the ex vivo production system is expected to improve platelet collection efficiency and to develop cell separation techniques. Alternatively, efficient platelet production could be achieved by elucidating two factors that may play a vital role in the mass production of platelet-like cells and inhibiting the cleavage of CD42b in the ex vivo production system and adding them to the in vitro production system. In the ex vivo production system, it is possible to verify which elements are essential for the production by selectively removing them from the tissue or adding of factors using tissue engineering techniques, such as decellularization and perfusion culture [26]. Therefore, unlike the in vitro production system, in which platelet production factors are supplemented to cells, we believed that the ex vivo production system is useful for screening factors that influence platelet production, as it is possible to exclude factors that are verified not to be related to production [27].
In conclusion, we verified the usefulness of the ex vivo production system for mass production of platelet-like cells from megakaryocytes. We clarified that there are some issues in ex vivo production system that need to be resolved before realizing industrial-scale production of platelet-like cells. Besides, identification of critical production factors using this system will enable the improvement of current in vitro production systems.