Using cynomolgus monkeys, we have previously established a new method for harvesting bone marrow cells (BMCs) with minimal contamination of the BMCs with T cells from the peripheral blood. We originally conducted this new “perfusion method” in the long bones (the humerus, femur, and tibia) of cynomolgus monkeys.
Here, we apply the perfusion method to obtain BMCs from the ilium of cynomolgus monkeys, since BMCs are usually collected from the ilium by the conventional aspiration method in humans. The perfusion method consists of two approaches: transverse iliac perfusion and longitudinal iliac perfusion. BMCs harvested by the perfusion method from the long bones and ilium were compared with those collected from the ilium by the aspiration method. The contamination of BMCs with peripheral blood, determined by the frequencies of CD4+ and CD8+ T cells, was significantly lower in BMCs obtained from the ilium or long bones by the perfusion method (CD4+ plus CD8+ T cells <4%) than in those obtained by the iliac aspiration method (CD4+ plus CD8+ T cells >20%). However, the numbers of immature myeloid cells, such as myeloblasts, promyelocytes, myelocytes, and metamyelocytes, were higher in BMCs obtained by the iliac perfusion method than in those obtained by the iliac aspiration method. The assays for in vitro colony-forming unit in culture revealed that progenitor activity was significantly higher in BMCs obtained by the perfusion method than in those obtained by the aspiration method. These findings suggest that the contamination of BMCs with peripheral blood is much less when using the perfusion method than when using the aspiration method. To determine the best site for harvesting BMCs by the perfusion method, age-dependent changes in BMCs harvested by the perfusion method from the long bones and ilium were examined. The numbers of BMCs varied in the long bones (humerus > femur > tibia) and showed age-dependent decreases, whereas they remained similar in the ilium of cynomolgus monkeys from 3 years to 6 years of age. However, in cynomolgus monkeys, BMC harvesting by the perfusion method from the ilium (but not from the long bones) is found to involve the risk of fat emboli, particularly when the BMCs are quickly perfused under high pressure. These findings suggest, even in humans, that the perfusion method is better than the aspiration method, and that the best site for collection of BMCs is the humerus.
Bone marrow transplantation (BMT) is now one of the most powerful strategies for the treatment of leukemia, aplastic anemia, congenital immunodeficiency, and also autoimmune diseases [1–, 3]. Furthermore, gene therapy and organ transplantation in conjunction with BMT have recently been carried out to treat various intractable diseases [4–, 7]. We previously have found that conventional allogeneic BMT can be used to treat autoimmune diseases in various autoimmune-prone mice except MRL/lpr mice . Furthermore, we have very recently established a new “Intra-Bone Marrow-BMT” (“IBM-BMT”) strategy for allogeneic BMT . This can be used to treat even intractable autoimmune diseases in chimeric resistant MRL/lpr mice . To apply the “IBM-BMT” method to humans, we have been attempting to find the best method for allogeneic BMT using cynomolgus monkeys.
In humans, bone marrow cells (BMCs) have usually been collected by multiple bone marrow aspirations from the iliac crest according to the method established by Thomas et al. . However, in this aspiration method, the BMCs are contaminated with the peripheral blood. Consequently, the BMCs include more than 20% T cells, and the transplantation of the BMCs causes acute graft-versus-host disease (GVHD). Very recently, using the long bones of cynomolgus monkeys, we have established a new “perfusion” method for collecting BMCs with minimal contamination with the peripheral blood . This method is simple and safe, and would, therefore, be of great advantage in obtaining pure BMCs, and result in the decreased incidence of acute GVHD in allogeneic BMT. However, the fatty marrow associated with decreased numbers of BMCs has been found to increase with age. To establish the best method for harvesting BMCs, we used variously aged cynomolgus monkeys, and compared the cellular components and progenitor activity of BMCs collected by the perfusion method from various sites, including the ilium, with those obtained by the conventional aspiration method. Here, we show that the perfusion method is better than the aspiration method, and that the best site is the humerus.
Materials and Methods
Normal cynomolgus monkeys (3 to 6 years old; 3 to 6 kg body weight [BW]) were obtained from Keari (Osaka, Japan). The monkeys were free of intestinal parasites and were seronegative for tuberculosis, herpes B, hepatitis A, and hepatitis B viruses. All surgical procedures and postoperative care of animals were carried out in accordance with the guidelines of the National Institutes of Health for care and use of primates. The study protocol was approved by the Animal Experimentation Committee, Kansai Medical University (Osaka, Japan).
BMC Harvesting from Long Bones by the Perfusion Method
The bone marrow fluid was collected from the long bones as described previously . In brief, as shown in Figure 1A, one needle (Katsunuma's bone marrow puncture needle, 1.8 mm diameter; Kyoto, Japan) was inserted into the proximal side of a long bone such as the humerus, femur, or tibia, and the other was inserted into the distal side. The first needle was connected to a syringe (30 ml, Code No. SS-30ES; Terumo Co., Ltd.; Shibuya, Tokyo, Japan) containing heparin (10 U/ml), and the other needle was connected to a syringe containing 30 ml of phosphate-buffered saline (PBS). The PBS was pushed gently from the syringe into the medullary cavity to flush out the bone marrow.
BMC Harvesting from the Ilium by the Transverse or Longitudinal Perfusion Method
Cynomolgus monkeys were anesthetized using Ketalar® (10 mg/kg; Sankyo Co., Ltd.; Tokyo, Japan; http://www.sankyo.co.jp/menu.html), and the bone marrow fluid was collected by the transverse (Fig. 1B) or longitudinal (Fig. 1C) perfusion method.
The longitudinal iliac perfusion method was performed as follows: one needle (Katsunuma's bone marrow puncture needle, 1.2 mm diameter; Kyoto, Japan) was inserted into the iliac crest, and the other needle was inserted into the region between the post-dorsal iliac spine and the gluteal line. The first needle was connected to a syringe containing heparin. The other needle was connected to a syringe containing 30 ml of PBS, and the PBS was then pushed gently from the syringe into the medullary cavity to flush out the bone marrow. The bone marrow fluid was collected into the syringe containing heparin under slight negative pressure.
The transverse iliac perfusion method was performed as follows: one needle was inserted into the anterior abdominal iliac spine, and the other needle was inserted into the anterior dorsal iliac spine. The first needle was connected to a syringe containing heparin, and the other needle was connected to a syringe containing 30 ml of PBS. The bone marrow fluid was collected using a similar technique to the longitudinal iliac perfusion method.
BMC Harvesting by the Conventional Multiple Aspiration Method
Bone marrow fluid (total 20 ml) was aspirated from the iliac crest, as previously described .
Preparation of BMCs
BMCs harvested either by the perfusion or aspiration method were centrifuged and suspended in 15 ml of PBS. They were placed on 15 ml of Lymphoprep density solution (1.077 g/ml; Nycomed Pharma As; Oslo, Norway). After centrifugation for 30 minutes at 2,000 rpm at room temperature, the BMCs were collected from the defined layer at the interface.
Isolation of Peripheral Blood Mononuclear Cells (PBMNCs)
PBMNCs were isolated from heparinized blood by centrifugation (30 minutes at 2,000 rpm at room temperature) on a cushion of Lymphoprep density solution.
Analyses of Cell Surface Antigens
Cell surface antigens on the PBMNCs and BMCs were determined using fluorescein isothiocyanate- or phycoerythrin (PE)-coupled monoclonal antibodies (mAbs) against human CD4, CD8, CD20, CD11b, or CD56 (Exalpha; Boston, MA; http://www.exalpha.com), and immunoglobulin M (IgM) (Biosource; Camarillo, CA; http://www.biosource.com). These mAbs were previously examined for their cross-reactivity to the molecules expressed on the cells from cynomolgus monkeys. Flow cytometric analyses were performed using an EPICS-XL® (Coulter Co.; Miami, FL; http://www.coulter.com) and a FACScan® (Becton Dickinson; Mountain View, CA; http://www.bd.com).
Cytological Analyses of Bone Marrow Fluid
The PBMNCs and the BMCs were cytocentrifuged onto a slide using a Cytospin 3® (Shandon Scientific Ltd.; Astmoor Runcorn, UK) and stained with May-Giemsa solution.
The colony-forming ability of the BMCs (colony-forming unit in culture [CFU-C]) was assayed as described previously . Briefly, BMCs (104 cells/well) were plated in 12-well plates (ICN Biomedicals, Inc.; Aurora, OH; http://www.icnbiomed.com) in 10 ml of Methocult GF H4434 (StemCell Technologies, Inc.; Vancouver, BC, Canada; http://www.stemcell.com), consisting of optimal concentrations of cytokines (recombinant human stem cell factor [SCF], erythropoietin [EPO], interleukin-3 [IL-3], GM-CSF, and G-CSF), 30% fetal bovine serum, 1% bovine serum albumin, 2 mML-glutamine, 10−4 M-mercaptoethanol, and 0.9% methyl cellulose. Fourteen days later, the CFU-C were counted.
Numbers of BMCs in Long Bones or Ilia of Cynomolgus Monkeys at Various Ages
Age-dependent changes in the numbers of BMCs in the long bones of cynomolgus monkeys at 3, 4, 5, and 6 years of age were first examined using the perfusion method, which minimizes the contamination of BMCs with peripheral blood, as described previously . As shown in Figure 2A, the numbers of BMCs obtained from one humerus and femur were 5.6 ± 1.9 × 108 and 2.6 ± 0.3 × 108 cells, respectively, at 3 years of age, but were fewer at 2.7 ± 1.2 × 108 and 1.9 ± 0.8 × 108 cells, respectively, at 6 years of age; there were significantly more BMCs harvested from the humerus than from the femur. There were significantly fewer BMCs collected from the tibia than from the femur, and these rapidly decreased from 0.55 ± 0.05 × 108 cells at 3 years of age to 0.18 ± 0.02 × 108 cells at 6 years of age (Fig. 2B). These findings suggest that the number of BMCs decreases in an age-dependent manner. However, sufficient numbers of BMCs were still maintained in the humerus and femur at 6 years of age, but not in the tibia, where rapid replacement of the red marrow by the fatty marrow was observed.
In the case of the ilium, the numbers of BMCs collected by the perfusion method were similar between 3- and 6-year-old monkeys (Table 1). The numbers of BMCs collected from the ilium using the transverse and longitudinal perfusion methods were 0.6 ± 0.1 × 108 and 1.9 ± 0.9 × 108, respectively, at 3 years of age, and 0.9 ± 0.1 × 108 and 2.6 ± 0.6 × 108, respectively, at 6 years of age. The numbers of BMCs collected by the transverse perfusion method were approximately one-third those collected by the longitudinal perfusion method. This seemed to be due to the small amount of bone marrow space in the iliac transverse crest of cynomolgus monkeys (Fig. 1). That the number of BMCs collected from the ilium was similar for 3- and 6-year olds suggests that the ilium is a suitable location for collecting sufficient numbers of BMCs from either aged individuals. However, in cynomolgus monkeys, the use of the perfusion method to collect BMCs from the ilium (but not long bones) was found to involve the risk of fat emboli, particularly when the BMCs were quickly perfused under high pressure. The numbers of BMCs harvested by the conventional aspiration method were 3.5 ± 0.4 × 108 cells at 3 years of age and 3.6 ± 0.9 × 108 cells at 6 years of age. Both figures were higher than when harvested by the perfusion method, however, this was due to the contamination of BMCs with peripheral blood, as shown in the following findings.
Table Table 1.. Number of BMCs collected from an iliac crest
The results of each group are expressed as the mean ± SD of three monkeys.
Cell n (× 108)
3 years old (n = 3)
6 years old (n = 3)
0.6 ± 0.1
0.9 ± 0.1
1.9 ± 0.9
2.6 ± 0.6
3.5 ± 0.4
3.6 ± 0.9
Cell Surface Antigen and Cytological Analyses
We next compared the percentages of T cells (CD4+ plus CD8+ cells) in the BMCs collected from the long bones or ilia by the perfusion method with those harvested by the conventional aspiration method. As shown in Table 2, more than 40% of the PBMNCs were found to be T cells (CD4+ plus CD8+ cells), and more than 20% of the BMCs collected from the ilium by the conventional aspiration method were found to be T cells. In contrast, less than 4% of the BMCs collected from the long bones (the humerus and femur) and ilia by the perfusion method were found to be T cells (p < 0.01, perfusion versus aspiration method). Furthermore, to examine the contamination of the BMCs with peripheral blood, we compared the cytological findings of the BMCs harvested by the conventional aspiration method with those harvested by the perfusion method. Table 3 shows the differential counts of myeloid cells in the BMCs. The percentages of myeloblasts to metamyelocytes were higher in the BMCs collected by the iliac perfusion method than in those collected by the iliac aspiration method. This was confirmed by the cytospin profiles; myeloid, erythroid, and megakaryocytic cells were observed in the BMCs harvested by the iliac perfusion method, whereas mature lymphocytes were dominant in the specimen obtained by the iliac aspiration method (data not shown). Table 4 shows the RBC:WBC and lymphocyte: granulocyte ratios. Due to the contamination of BMCs with peripheral blood, both ratios in the BMCs obtained by the iliac aspiration method were significantly higher (p < 0.01) than those obtained by the iliac perfusion method: RBC:WBC = 273.1 ± 90.0 versus 41.5 ± 16.0, respectively, and lymphocyte:granulocyte = 1.3 ± 0.1 versus 0.3 ± 0.1, respectively.
Table Table 2.. Analyses of cell surface antigens on BMCs
Cells were stained with FITC- or PE-conjugated mAbs and analyzed using an EIPCS-XL.
The data of the iliac (transverse or longitudinal) perfusion method, long bone perfusion method, and iliac aspiration method are expressed as the mean ± SD.
Statistical analyses were performed by t-test: *p < 0.01; perfusion method versus aspiration method.
Cell surface antigens (%)
1.0 ± 0.2*
2.7 ± 1.1*
2.3 ± 0.6
2.9 ± 0.9
26.7 ± 9.1
5.0 ± 1.6
1.0 ± 0.4*
2.8 ± 1.5*
3.4 ± 2.2
2.7 ± 1.3
25.7 ± 8.7
6.6 ± 3.5
0.9 ± 0.6*
2.2 ± 1.9*
2.7 ± 1.7
5.5 ± 1.8
28.1 ± 5.5
6.9 ± 3.1
(humerus and femur)
Asiration method (ilium)
10.5 ± 0.8
12.6 ± 0.2
5.1 ± 1.8
7.1 ± 3.3
23.0 ± 4.1
5.5 ± 2.0
19.9 ± 5.1
22.4 ± 4.8
17.3 ± 4.0
22.6 ± 9.1
10.8 ± 4.4
3.5 ± 2.2
Table Table 3.. Cytological analyses of myeloid cells
The numbers of myeloid cells in different maturational stages were counted under a microscope after staining cytospin samples with May-Grünwald's solution. Five hundred cells were counted, and the percentages of cells in each stage were calculated.
*p < 0.05; perfusion method versus aspiration method.
Aspiration method (%)
1.7 ± 0.7
0.5 ± 0.2
0.0 ± 0.0
3.7 ± 1.8
0.9 ± 0.3
0.0 ± 0.0
5.6 ± 3.8
2.4 ± 0.8
0.0 ± 0.0
14.2 ± 2.5*
2.8 ± 1.0
0.0 ± 0.0
37.7 ± 4.8
28.3 ± 7.4
17.5 ± 4.8
36.8 ± 4.5
65.1 ± 8.7
82.5 ± 9.8
Table Table 4.. RBC/WBC and lymphocyte/granulocyte ratios in BMCs
The ratios of RBC to WBC were determined by an automated blood counter (Sysmex K1000).
The numbers of granulocytes and lymphocytes were microscopically counted after staining cytospin samples with May-Grünwald's solution.
*p < 0.01; perfusion method versus aspiration method.
Perfusion method (n = 6)
Aspiration method (n = 6)
PBMNC (n = 3)
41.5 ± 16.0*
273.1 ± 90.0
559.0 ± 79.8
0.3 ± 0.1*
1.3 ± 0.1
2.4 ± 0.2
These findings indicate that the contamination of BMCs with peripheral blood was significantly lower when using the iliac perfusion method than when using the iliac aspiration method.
In vitro CFU-C assays were carried out to examine the progenitor cell activity in the BMCs collected by the iliac perfusion and aspiration methods. The BMCs harvested by the iliac perfusion and aspiration methods were cultured in methylcellulose containing a combination of cytokines (SCF, EPO, IL-3, GM-CSF, and G-CSF). As shown in Table 5, the BMCs collected by the perfusion method generated a significantly higher (p < 0.01) number of CFU-C than those harvested by the aspiration method when assayed on day 14 of culture (ilium: 26.0 ± 4.0 versus 14.0 ± 2.0 /104; and long bone: 22.0 ± 4.0 versus 8.3 ± 0.6 /104). This indicates that the frequency of progenitor cells was higher in the BMCs collected by the perfusion method than in those collected by the aspiration method.
Table Table 5.. Colony-forming unit in culture
CFU-C assays were determined after culturing BMCs for 12 days with SCF, IL-3, GM-CSF, and G-CSF. CFU-C counts in the PBMNCs were also examined as a control: 3.3 ± 1.5/104 cells.
*p < 0.01; perfusion method versus aspiration method.
Ilium (n = 3)
Long bone (humerus) (n = 3)
26.0 ± 4.0*
22.0 ± 4.0*
14.0 ± 2.0
8.3 ± 0.6
Primates have recently been used for preclinical studies on allogeneic BMT, organ transplantation, and gene therapy, and these studies have substantially benefited human applications [7,, 11]. However, a serious problem associated with BMT is the contamination of the BMCs with T cells derived from the peripheral blood when the BMCs are collected using the multiple aspiration method. Although various treatments with anti-T cell Abs and immunosuppressants, etc., have been used in an attempt to reduce the functions of these contaminant T cells, they have failed to prevent GVHD. In the cynomolgus monkey, the BMCs collected from the ilium by the conventional aspiration method included more than 20% T cells (CD4+ plus CD8+ cells). However, BMCs collected by the perfusion method from the long bones (the humerus and femur) contained less than 4% T cells. Therefore, there was significantly less contamination of the BMCs with peripheral blood when using the perfusion method versus the aspiration method. This was the case when the BMCs were obtained from the ilium by either the transverse or longitudinal perfusion method (Table 2). The reduced contamination of BMCs with peripheral blood containing mature T cells was confirmed by the RBC:WBC and lymphocyte:granulocyte ratios (Table 4). It has been reported that the T cells (particularly CD8+ T cells) originally present in the bone marrow of mice facilitate the engraftment in allogeneic BMT, and that these T cells do not induce GVHD [12–, 14]. Indeed, we have confirmed that no GVHD develops when the BMCs (T cells <4%) harvested from the long bones by the perfusion method (but not by the aspiration method) are used for allogeneic BMT in cynomolgus monkeys (manuscript in preparation).
It has generally been known that more than 2 × 108 BMCs/kg are necessary for human BMT. As shown in Table 3, BMCs collected by the perfusion method contained more immature cells, such as myeloblasts and promyelocytes, than those obtained by the aspiration method. Furthermore, in CFU-C assays, the BMCs collected by the perfusion method showed higher progenitor activity than those collected by the aspiration method (Table 5). Thus, a large number of pure BMCs can rapidly be harvested using the perfusion method, and it may be possible to reduce the number of BMCs required for BMT because of the high percentage of hemopoietic progenitor cells (due to the low level of contamination with the peripheral blood) when using this method. Furthermore, it should be noted that the enriched progenitor activity (CFU-C in Table 5) in the BMCs collected by the perfusion method is advantageous for the recipients, since short-term reconstitution by donor cells is attributed to these progenitors. Indeed, we have confirmed that the injection of BMCs (3 × 108), obtained from the humerus of a cynomolgus monkey (3 kg BW), directly into the bone marrow of another cynomolgus monkey (3 kg BW)—“IBM-BMT” (as previously described )—leads to the quick recovery (within 10 days) of hemopoiesis (granulocytes, erythrocytes, and thrombocytes) (manuscript in preparation).
Although the number of BMCs in the long bones decreases with age, an adequate number can still be obtained from the humerus and femur of 6-year-old monkeys (Fig. 2). Furthermore, the number of BMCs obtained from the ilium of 6-year-old monkeys was similar to that collected from 3-year-old monkeys, although perfusion must be carefully performed so that fat emboli do not develop when the BMCs are obtained from the iliac crest of cynomolgus monkeys by the perfusion method. Since it has been reported that there is red bone marrow in the ilium of even elderly human individuals , BMCs harvested from the ilium of aged donors by the perfusion method can be used for BMT across major histocompatibility complex (MHC) barriers and organ transplantation (in conjunction with BMT) if the donors are brain dead.
Nowadays, the number of patients for whom BMT is necessary is rapidly increasing, and HLA-mismatched BMT must therefore be carried out. However, there are several problems, such as GVHD, graft rejection, and incomplete T cell recovery in BMT across MHC barriers [16–, 19]. In the present study, we have succeeded in collecting pure BMCs using the perfusion methods; cellular components are close to those of original BMCs present in the bone marrow (minimal T cell contamination and enriched hemopoietic progenitor cells). In addition, we have found that the BMCs thus collected contain stromal cells (including mesenchymal stem cells): fibroblastic adherent cells grow in vitro after culturing the BMCs (manuscript in preparation). We have previously found that MHC restriction exists between pluripotent hemopoietic stem cells and stromal cells , and that the recruitment of donor stromal cells facilitates the engraftment of donor BMCs [21–, 24]. If the BMCs contain both hemopoietic stem cells and mesenchymal stem cells, the injection of whole BMCs obtained from the humerus of healthy donors by the perfusion method directly into the bone marrow of recipients (“IBM-BMT,” as described in our recent paper ) would bring about great benefits for human allogeneic BMT, gene therapy, organ transplantation in conjunction with BMT, and also autologous BMT for regeneration therapy and cancer therapy.
Whole BMCs contain not only hemopoietic stem cells and stromal cells (including mesenchymal stem cells) but also various BMT-facilitating cells, such as T cells, natural killer (NK) cells, NK T cells, macrophages, dendritic cells, and their progenitors. In addition, whole BMCs contain granulocytes, platelets, erythrocytes, and their progenitors. Therefore, these cells can prevent the graft failure of donor BMCs and facilitate hemopoietic recovery. However, in human BMT, in order to remove T cells, the use of CD34-positive selection is globally prevalent, although this procedure leads to the elimination of BMT-facilitating cells. We are now in the process of establishing a new safe strategy for allogeneic BMT using the perfusion method in the cynomolgus monkey.
We thank Ms. Y. Tokuyama, Ms. M. Shinkawa, and Ms. S. Miura for their expert technical assistance, and Mr. Hilary Eastwick-Field and Ms. K. Ando for their help in the preparation of the manuscript.
This work was supported by: a grant from “Haiteku Research Center” of the Ministry of Education; grant-in-aid for scientific research (B)11470062; grants-in-aid for scientific research on priority areas (A)10181225 and (A)11162221; a grant from the “Millennium” program of the Ministry of Education, Culture, Sports, Science and Technology; a grant from the “Science Frontier” program of the Ministry of Education, Culture, Sports, Science, and Technology; and also a grant from Japan Immunoresearch Laboratories Co., Ltd. (JIMRO).