In Vivo Growing of New Cell Colonies in a Portion of Bone Marrow: Potential Use for Indirect Cell Therapy

Abstract

The ability of bone marrow cells (BMCs) to migrate to different organs can be used for indirect cell therapy, a procedure based on the engraftment of therapeutic cells in a different place from where they will finally move to and perform their action and which could be particularly useful for chronic illness where a persistent and long-lasting therapeutic action is required. Thus, establishing a stable colony of engineered BMCs is a requisite for the chronic provision of damaged tissues with engineered cells. Reported here is a procedure for creating such a cell colony in a portion of the bone marrow (BM). The study was performed in C57BL/6j mice and consisted of developing a focal niche in a portion of the bone marrow with focal irradiation so that it could be selectively colonized by BM cells (C57BL/6-FG-VC-GFP mice) injected in the blood stream. Both the arrival of cells coming from the nonirradiated BM (week 1 after irradiation) and the proliferation of cells in the irradiated BM (week 2) prevented the homing of injected cells in the BM niche. However, when BMCs were injected in a time window about 48 h after irradiation they migrated to the BM niche where they established a cell colony able to: 1) survive for a long period of time [the percentage of injected cells increased in the BM from day 30 postinjection (15%) to day 110 postinjection 28%)]; 2) express cell differentiation markers (90% of them were lineage committed 4 weeks after engraftment); and 3) colonize to the blood stream (with 5% and 9% of all blood cells being computed 1 and 3 months after engraftment, respectively). The intravenous injection of BMCs in combination with a previous transitory focal myeloablation is a safe and easy method for creating the long-lasting colony of modified BMCs needed for treating chronic and progressive illness with indirect cell therapy.

INTRODUCTION

Bone marrow (BM) stem cells can be differentiated into a number of cell types suitable for cell therapy (3,9,15,16,19,30,31,33), a procedure that often requires their injection (sometimes after their in vitro differentiation to a particular cell type) in the damaged organ where they should replace degenerated cells (11,16,33). However, both processes can occur spontaneously because BMCs have the potential ability to migrate from the blood to different organs where they differentiate to cell types of the host tissue (12,32,34,37,38). This opens up an alternative for cell therapy, which uses the migratory capability of BMCs injected in the blood stream to introduce new cells in damaged tissues (indirect cell therapy; ICT). This procedure could be particularly suitable for cell therapy of tissues with a difficult access or where the optimal distribution of implanted cells cannot be obtained with local injections. This is the case of the brain, where cell implants can only be performed with major surgery and where the high cell density obtained after the direct injection of cells in the brain tissue can induce markedly undesirable side effects (8,22,45). Circulating BMCs can cross the blood–brain barrier (BBB), spreading across the brain tissue (7,25) where, after their differentiation to microglia (50%) and other cell types (1,2,18,34), they integrate in the brain without inducing an excessive cell aggregation. These advantages, and the fact that BMCs are attracted to degenerating brain regions (34), suggest ICT as a particularly suitable procedure for the treatment of neurodegenerative disorders.

The low rate of BMC migration to the brain (34) limits the utility of ICT in acute brain illness (such as stroke), but not in chronic and progressive illness (such as Parkinson’s disease or Alzheimer’s disease), where a persistent and long-lasting therapeutic action is required. However, the establishment of a peripheral BMC colony suitable for providing the brain with engineered cells (e.g., able to release neuroprotecting molecules) for months or years is required for long-lasting therapies. In this work we studied the best procedure for creating this cell colony in mice using methods potentially applicable to humans. Taking advantage of the ability of BMC to migrate from the blood to the BM, a stable BMC colony can be produced by simply injecting these cells in the blood stream. Although this method has proved suitable in animals, a very large dose of BMCs (generally obtained from the BM of other animals) is needed to develop a significant cell colony (27,35,40), which limits its utility in humans where a self-transplantation needed to avoid immune reactions is highly desirable (the in vitro amplification of the number of BMCs normally modifies their characteristics). Better results could be obtained with intra-BM cell injections (20,39), but even in this case the proliferation of injected cells might not be enough to make a significant cell colony (47).

The efficiency of both implanting methods markedly increases after myeloablation (e.g., with body irradiation) (5), a procedure that has been extensively used since the 1960s for treating different illnesses (41–43) but whose dangerous side effects restrict its use to life-threatening diseases (24). Partial myeloablation (14) (an incomplete decrease of BMC density in all the BM by using low radiation doses, cyclophosphamide, methotrexate) (5,10,28,36) could be a more suitable procedure. Although it is a less dangerous method that could be more suitable for treating nonmalignant diseases (it has been used in hemoglobinopathies, autoimmune disorders), it is not free from fatal complications. Thus, we decided to use focal myeloablation (FMy), a procedure based on inducing a marked ablation in a portion of the BM with local irradiation (right hind limb in our case), thereby creating the local conditions needed for the development of a cell colony (“focal niche”). Although less characterized (26) [it has been mainly used to study the cell migration between two portions of the BM (23)], this procedure seems to be much less dangerous (most of the BM remains unaffected) and more suitable for treating patients with neurodegenerative disorders. The aim of this study was to determinate the most suitable conditions in mice for the creation of a stable BMC colony by using FMy and cell injections in the blood stream, and to characterize this procedure that could be applicable to humans.

MATERIALS AND METHODS

Experiments were carried out on 10-week-old male C57BL/6j and C57BL/6-FG (VC-GFP) mice homozygous for transgene transmission (Charles River Laboratories, Barcelona, Spain). Animals (body weight 28 ± 3 g) were housed at 22 ± 2°C, five per cage, under normal laboratory conditions on a standard light/dark schedule (12:12 with 0800–2000 light on) and free access to food and water. Experiments were carried out in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC) regarding the care and use of animals for experimental procedures and adequate measures were taken to minimize pain and discomfort. All the procedures were approved by the Ethics Committee on Animal Experimentation.

Bone Marrow Irradiation

Irradiations were performed under anesthesia (ketamine 80 mg/kg/xilacine 36 mg/kg, IP). A mouse group (C57BL/6j) was subject to a whole-body irradiation of 10 Gy (one session with ⁶⁰Co60; 1% lethality), but avoiding the head irradiation with a lead protection block of 70 mm/thickness (general irradiation). Mice from the other group (focal irradiation) received the same irradiation but after protecting the whole body with the lead protection block except for the right hind-limb.

Bone Marrow Transplantation

C57BL/6-FG (VC-GFP) mice were used as BM donors (n = 20). After sacrificing 8–12-week-old mice, BM cells from femurs and tibias were collected by flushing the shaft with phosphate-buffered saline (PBS) using a syringe with a 26-gauge needle. Cells were passed through 70-μm nylon mesh (to remove the remaining clumps of tissue) and washed (in PBS with a double centrifugation of 300 × g for 6 min). The erythrocytes were lysed (in 0.84% NH₄Cl) and BMCs were resuspended in DMEM (1% penicillin/streptomicin; Sigma) and adjusted to 10⁷ cells/ml.

At different times after their irradiation, recipient mice (C57BL/6j) were injected in the tail veiqn with 200 μl of the BMDC suspension (2 × 10⁶ cells). After cell engraftment, mice were housed for different time periods.

Bone Marrow Study

In order to obtain a cell suspension representative of each BM, mice were sacrificed at the end of the experiments and both femurs were quickly obtained. BM cells were then collected (by flushing the BM shaft with 1 ml PBS injected with a 25-gauge needle), filtered (to remove the remaining clumps of tissue with a 70-μm nylon mesh), and washed (by centrifuging at 300 × g during 6 min in PBS). Erythrocytes were removed by lysis (pellets were resuspended in 2 ml of PBS and 6 ml of a solution of 0.84% NH₄Cl, 0.1% KHCO₃, and 0.037% EDTA was added for 10 min), followed by a double centrifugation (300 × g for 10 min, resuspended in 6 ml PBS, centrifuged at 300 × g for 6 min and resuspended in 0.5 ml of PBS). At this time, cell suspensions were used to quantify the cell number per BM (hemocytometer) or to study the characteristics of their cells with fluorescence-based methods.

The counting of cells with the hemocytometer was performed by adding one volume (0.1 ml) of PBS/4% paraformaldehyde (PFA) to 100 μl of the cell suspension. After 15 min at room temperature, fixed cells were counted in a hemocytometer (Neubauer counting chamber), distinguishing implanted (GFP⁺) and resident (GFP⁻) BM cells by using fluorescence microscopy (DM2500; Leica; Barcelona).

Cell suspensions were filtered through 40-μm nylon mesh, washed in 5 ml of PBS, and resuspended in 400 μl before using flow cytometry. This new suspension was used for either identifying BM cells committed to the hematopoietic lineage (lineage detection kit) or, after fixing in PBS/2% PFA (adding 600 μl PBS and 1 ml of PBS/4% PFA), to quantify the percentage of GFP⁺ cells and the percentage of proliferating cells with a propidium iodide-based method.

The lineage detection was performed by mixing a lineage detection cocktail-biotin (containing monoclonal antibodies for T cells, B cells, monocytes/macrophages, granulocytes, and erythrocytes) and the cell suspension (MACS® Miltenyi Biotec Inc., USA) in a proportion of 1:10 (v/v). After 40 min at 4°C, the cells were washed (by adding 2 ml PBS and centrifuging for 6 min at 300 × g) and incubated for 15 min at 4°C with 1:10 anti-biotin-PE. Finally, the cells were washed (by adding 2 ml PBS and centrifuging 6 min at 300 × g) and fixed in PBS/2% PFA.

In addition to the hemocytometer count, the BM cells harvested were analyzed by flow cytometry; intrinsic and extrinsic cells were distinguished according to their natural fluorescence.

The percentage of proliferating cells was determined by labeling the DNA with propidium iodide in order to identify cell cycle phase, with the S and G₂ phases being used as indicators of cell proliferation and G₁ as the indicator of nonproliferation (6,13). In this procedure, the fixed cell suspension (500 μl) was centrifuged at 300 × g for 6 min, resuspended in 500 μl of the cool propidium iodide (Sigma, St. Louis, MO) staining solution (50 μg/ml propidium iodide, 0.1% Triton X-100, 5% glycerol in PBS), and incubated at 4°C for 4 h.

All cell suspensions were filtered (50 μm) just before their acquisition and analyzed with the flow cytometer (EPICS-XL Backman coulter and Expo32 software; Isaza, Madrid). The optic filters used were log FL-1 (for GFP⁺ cells), log FL-2 (for lineage⁺ cells), and linear FL-3 (for propidium iodide-labeled cells).

Blood Study

Smears of blood samples obtained at the end of the experiments were analyzed by fluorescence microscopy (DM2500; Leica; Barcelona) using 40× objective. For each sample, five randomly selected fields were used for counting GFP⁺ and GFP⁻ cells. The percentage of GFP⁺ cells was calculated as the average number of GFP cells versus total cells counted from each field.

Experiments

The first experiment was aimed at evaluating the recovery of the number of BM cells after irradiation. Two experimental conditions were tested, one with a selective irradiation of BMs of the right hind-limb (focal irradiation) and other with an irradiation of all BMs (general irradiation; irradiation of the whole-body except for the head). In both groups, the number BM cells of the right femur were computed at different time intervals after irradiation (1, 2, and 4 weeks after treatment vs. controls submitted to the same procedure, including anesthesia, but which were not irradiated).

The time course of the cell response to focal irradiation of the right hind-limb was studied in the second experiment. Thus, the number of BM cells of the right (irradiated) and left (nonirradiated) femur was computed 1, 2, 3, 7, 14, 35, and 56 days after irradiation. In addition, the cell proliferation in both femurs was computed 2, 3, 7, and 11 days after irradiation.

The third experiment was aimed at studying the BM colonization of focally irradiated C57BL/6j mice and which were injected in the tail vein with GFP⁺ BMCs obtained from GFP-expressing mice (C57BL/6-FG). The cell colonization was studied in two conditions, without previous myeloablation and after myeloablation of the right hind-limb. In order to determine the optimal time window for cell colonization, mice were injected with GFP cells 2 h after irradiation (day 0), and 1, 2, 7, and 14 days after irradiation. Data were compared with those obtained from two control groups, one injected with cells but not irradiated and the other irradiated but not injected with cells. Thirty days after cell implants, the total number of cells in the irradiated and nonirradiated BM was computed. In order to evaluate cell colonization degree, the percentage of GFP cells in both femurs was determined. In order to study the functional status of implanted cells, the percentage of lineage-committed cells was quantified in the GFP cells obtained from the irradiated and nonirradiated femurs of mice injected with cells 2 days after focal irradiation and studied 30 days later. The lineage commitment degree of implanted GFP cells was compared with that computed for nonirradiated animals (C57BL/6j and C57BL/6-FG VC-GFP).

The fourth experiment was aimed at studying the persistence of the BMC colonies in the host. Two days after the focal myeloablation of the right hind-limb, C57BL/6j mice were injected in the tail vein with GFP cells (obtained from the BM of C57BL/6-FG mice), and the number of GFP and non-GFP cells in blood samples and in the BM of the irradiated (right hind-limb) and nonirradiated (left hind-limb) femur were computed 30 and 110 days after cell injection.

Statistics

Statistical analysis was performed using one- and two-way ANOVAs followed by the Tukey honesty-test for post hoc comparisons. Analysis was performed using the Statistica program (Statsoft; Tulsa, USA). A level of p < 0.05 was considered as critical for assigning statistical significance.

RESULTS

Nonirradiated BM Is Involved in the Cell Recovery of Irradiated BMs

The first experiment showed a marked myeloablation in the right femur of both the focal-irradiated and general-irradiated mice. One week after irradiation, the number of BM cells was about 15–20% of those computed for the nonirradiated control (Fig. 1). This effect progressively decreased during the following weeks, with the number of BM cells being completely restored in both groups 1 month after irradiation. However, the recovery rate during this time period was not the same in the mice that received focal and general irradiation [two-way ANOVA focal radiation × general radiation/days after irradiation showed a significant interaction effect; F(2, 33) = 10.62, p < 0.001], with the focal-irradiated mice showing a more marked recovery of BM cells than the general-irradiated mice 2 weeks after irradiation.

Figure 1.

Effect of focal and global irradiation on the BM cells. Values are mean ± SE of the number of the BM cells of the right femur of animals that received focal irradiation (only in the right hind-limb) or general irradiation (irradiation of the whole body except for the head). The number of animals is shown at the top of each column. The mean ± SE of the cell number of control nonirradiated mice is shown on the left of the figure.

The femur focally irradiated in the second experiment showed a progressive myeloablation (Fig. 2A) during the first 2 days after irradiation [one-way ANOVA, with F(7, 59) = 26.7, p < 0.001]. From day 2 to day 7 the number of BM cells was maintained at about 20% of their normal level (data in Fig. 2A have been normalized as percentages of the mean number of BM cells computed for the nonirradiated control), and was restored during the second postirradiation week to 70% of control.

Figure 2.

Effect of focal irradiation of the right hind-limb on the number of cells (top) and cell proliferation (bottom) of BM-cells of the right femur (irradiated) and left femur (non-irradiated), at different time intervals after irradiation. Values in (A) and (B) were normalized as percentages of the mean value of the number of cells computed for the control nonirradiated mice. Values in (C) are the percentages of proliferating cells computed in the BM of the right femur. (D) and (E) show examples of the distribution of cells according to the intensity of the propidium iodide labeling in animals implanted 2 dats (D) and 11 days (E) after irradiation. Cells with more intense labeling (small peak in the right side of figures) were considered as proliferating cells (S and G₂ phases) and those with less intense irradiation (high peak in the left side of figures) were considered as nonproliferating cells (G₁ phase). The number of animals is shown in (A). Values in (A), (B), and (C) are mean ± SE.

No significant modifications in the number of BM cells were observed in the nonirradiated femur during the first days after irradiation (Fig. 2B). However, a marked increase in the cell number was observed 7 days after irradiation [F(7, 58) = 6.27, p < 0.001], an affect that completely vanished 2 weeks after irradiation and which was not detected on the days 35 and 56 after irradiation.

BM cell proliferation markedly changed after irradiation (Fig. 2), an effect that was different in the irradiated and nonirradiated BM. Cell proliferation in the irradiated femur decreased during the first week (Fig. 2C, D) and increased in the second week[FM(4, 43) = 33.6, p < 0.001] (Fig. 2C, E). The nonirradiated femur showed a marked cell proliferation during the first 3 days after irradiation, a transient effect that completely vanished 1 week after irradiation [F(4, 62) = 77.1, p < 0.001] (Fig. 2C, D).

Injected Cells Colonize the Irradiated BM and Were Present in the BM and Blood at Least for 3 Months

GFP cells were identified in the BM 1 month after their injection in the blood stream (Fig. 3). Only irradiated mice showed populations of GFP⁺ cells (Fig. 3A). Thus, 15% of the cells found in the irradiated BM of mice injected 1 day after irradiation were GFP⁺ (Fig. 3B), a proportion similar to that computed in mice injected 2 days after irradiation. The proportion of GFP⁺ was much smaller in animals injected 2 h (4%), 7 days (<1%), or 14 days (practically absent) after irradiation (Fig. 3C). The nonirradiated portion of BM also accepted GFP⁺ cells but, in this case, the new colony was about 8 times lower than that observed in the irradiated BM. Even in the nonirradiated BM, the best engraftment was found in mice injected 1 or 2 days after the contra-lateral irradiation [two-way ANOVA irradiated femur × nonirradiated femur/days after irradiation showed a significant interaction effect, with F(4, 55) = 19.59, p < 0.001]. Taken together, these data show an optimal window for BMC engraftment between 24 and 48 h after irradiation, with the cell viability (see the standard error in Fig. 3C) being more variable for engraftment performed 24 h after the irradiation than those performed 48 h after irradiation. Therefore, 48 h after irradiation proved to be the optimal time latency for cell engraftment.

Figure 3.

Colonization of BM and blood by BMCs injected in the blood stream: influence of the postirradiation time interval. Bone marrow: An example of the distribution of cells according to the intensity of the GFP fluorescence is shown in (A) and (B); the peak on the left shows cells not expressing GFP (intrinsic cells) and the peak on the right shows cells expressing GFP [injected cells which were observed in mice irradiated 1 days later (B) but not in nonirradiated mice (A)]. (C) The mean ± SE of the percentage of GFP cells (vs. total cells) computed 1 month after cell injection in the BM of the irradiated and nonirradiated femur. Blood: images used to compute the percentage of cells derived from the GFP-injected cells (green) in the blood are exampled in (D). (E) The mean ± SE of the percentage of GFP cells (vs. total cells) computed 1 month after cell injection, and (F) is an example of negative transplant in nonirradiated mice. The percentages (mean ± SE) of intrinsic (GFP⁺) and injected (GFP⁺) BM cells expressing lineage antigens of T cells, B cells, monocytes/macrophages, granulocytes, or erythrocytes (cocktail of monoclonal antibodies grouped as lineage⁺ cells) in the BM 1 month after cell injection are shown in the bottom left corner of (G). (H) Examples of the distribution of lineage-committed GFP⁺ cells and GFP⁻ cells obtained from a BM sample. The number of animals is shown in (C).

Identification and quantification of GFP cells in the blood of injected mice 1 month after the cell injection showed very different results in the five groups of the study. A high persistence of injected cells in the blood was observed in animals injected 1 or 2 days after irradiation (5% of blood cells were GFP⁺) (Fig. 3D), an effect that markedly decreased in animals injected 2 h after irradiation and 7 or 14 days after irradiation [F(3, 56) = 23.16, p < 0.001] (Fig. 3E). GFP⁺ cells were practically absent in nonirradiated mice (Fig 3F).

GFP cells injected 2 days after irradiation remained in the blood and BM for more than 3 months (Fig. 4). In both the blood (Fig. 4A) and BM (Fig. 4B), the percentage of GFP cells increased from day 30 to day 110, an effect that was observed in the irradiated and nonirradiated femur in the BM (Fig. 4B). Taking into account the fact that the percentage of GFP cells in the nonirradiated BM versus the irradiated BM did not change between day 30 and day 110 postirradiation (Fig. 4C shows the % of GFP cells in irradiated vs. nonirradiated BM), this increase is probably induced by the local proliferation of GFP cells in both irradiated and nonirradiated BM more than by the migration of GFP cells from the irradiated to the nonirradiated BM.

Figure 4.

Evolution of the BM colony at different time intervals after the BMC injection in the blood stream. (A) (blood) and (B) (bone marrow) show the mean ± SE of the percentage of injected BMCs (vs. total cells) computed 30 and 110 days after cell injection. (C) (bone marrow) shows the mean ± SE of the percentage of injected BMCs computed in the nonirradiated femur versus the irradiated femur. The number of animals is shown in (A).

Extrinsic Cells That Colonize the BM Are Able to Express a Lineage-Committed Phenotype

In order to study the functional status of implanted cells, the percentage of hematopoietic lineage-committed cells was quantified in GFP⁺ (implanted) and GFP⁻ (intrinsic) cells. In nonirradiated control animals, the percentage of lineage-committed cells was similar in C57BL/6j and C57BL/6-FG (VC-GFP) mice (about 80% of BM cells) (Fig. 3G). In the irradiated/implanted mice, the percentage of lineage-committed cells was higher in the implanted cells than in the cells of the host mice, a fact observed in both the irradiated and the contralateral nonirradiated femur [two-way ANOVA irradiated GFP⁺ × GFP⁻ cells/irradiated × nonirradiated femur showed no significant interaction, with F(2, 28) = 1.46, p = 0.24, but significant differences between GFP⁺ and GFP⁻ cells, with F(2, 28) = 13.26, p = 0.001].

Taken together, all the present data show that GFP⁺ cells injected in the blood stream 48 h after the irradiation of right hind-limb migrate to and colonize the previously irradiated portion of BM, where they survive for a long period of time, are able to express differentiation lineage markers, and they are probably released back to the blood stream from the host BM.

DISCUSSION

Here we report a procedure for establishing a stable colony of BMCs in a portion of the BM. This appears to be a safe and easy method, which could be particularly useful for exploiting the migrating capabilities of BMCs to provide a stable source of genetically modified BMCs.

The irradiation of a portion of the BM induced a local cell- epression in the BM (to <20% of the normal level), an effect that was established 24 h after treatment and that progressively vanished during the following 2 weeks. This focal myeloablation was longer lasting when other portions of the BM were also irradiated, suggesting that the nonirradiated BMs are involved in the cellular recovery of the irradiated BM [BM cells can migrate between BMs (46), particularly after partial BM irradiations (23)]. The nonirradiated BM showed a cell proliferation increase during the days that followed the contralateral irradiation, and an increase in the total number of cells at the end of the first postirradiation week. During this time the number of cells and the cell proliferation remained at very low levels in the irradiated BM. During the second postirradiation week, the cell increase observed in the nonirradiated BM decreased to normal levels (a fact not explained by a low proliferation rate, which at this time was similar to that observed under normal conditions), whereas the cell population of the irradiated BM increased.

These observations suggest that the initial cell recovery in the irradiated BM begins with a cell proliferation in the nonirradiated BMs and a subsequent migration of their cells to the irradiated BMs. During this initial phase, the spontaneous recovery of the irradiated BM is probably minor, as suggested by the low proliferation rate observed in the irradiated BM at this time. After the first week, the cell proliferation in the irradiated BM increased over the normal basal level, suggesting that the cell activity in the irradiated BM is enough to promote its own recovery. However, also at this time it is not clear if the cell recovery in the irradiated BM is promoted by local cells that survived irradiation or by cells coming from other BM. The fact that even 2 weeks after irradiation the cell recovery was more marked in focal irradiated mice than in global irradiated mice supports the second possibility, with the cells coming from nonirradiated BM being mainly those involved in the cell proliferation of the irradiated BM during the second postirradiation week. The normalization of cell proliferation together with the increase of the number of cells observed in the nonirradiated BM 1 week after irradiation suggests that the cell flux from nonirradiated to the irradiated BM have finished at this time, a fact that also could be at the basis of the transitory accumulation of cells observed in the nonirradiated marrow at this time. Finally, 4 weeks after irradiation both the focal irradiated mice and the global irradiated mice showed a normal number of cells in the irradiated and nonirradiated BMs, suggesting that after the second postirradiation week the proliferation of cells surviving irradiation is sufficient to completely recover the basal number of BM cells.

Our data indicate a short time window for BMC engraftment after focal myeloablation. Early studies suggested that the optimal time interval between irradiation and engraftment is 24 h or less (44). We observed that these short intervals are low efficient (in the case of the 2-h interval) or present a high homing variability (with some mice injected with BMCs 1 day after irradiation showing a high homing degree but others showing a very low homing degree). Although the basis for this less effective response remains unknown, our data support the idea that the initial deterioration of BM tissue induced by recent irradiation and the local response to it (inflammation) do not facilitate the homing of BMCs injected so soon. Cells injected 1 week after irradiation showed a very low colonization of the BM, despite the fact that at this time the myeloablation degree was still very marked. Our data suggest that the cells which proliferated and accumulated in nonirradiated BMs during the first postirradiation week are released to the blood where they could compete with the experimentally injected cells, thereby obstructing their homing and proliferation in the BM niches. During the second week after irradiation, the high cell proliferation in the nonirradiated BMs vanished, and the colonizing activity of these cells decreased to levels that are probably not above the levels that normally occur under basal condition. However, a high cell proliferation was observed in the irradiated BM during the second week after irradiation (studied on day 11 postirradiation), which could explain the very low colonizing activity of cells injected at this time. On the other hand, cells injected 2 days after irradiation showed an intense colonization of the irradiated BM. At this time, proliferation in the irradiated BM was not particularly high and proliferation in the nonirradiated BM did not produce a high enough number of cells to compete with injected cells (the above commented data suggest that cells from nonirradiated BMs are probably being accumulated at this time and will not be massively released to the blood until the end of the first week). Thus, the optimal time window for BMC homing is about 48 h after irradiation, an effect observed in the irradiated but also in the nonirradiated BM.

The BM colonization degree observed here after a single low dose of cells (2 × 10⁶ cells) was similar to (15% of BM cells 30 days after the implant) or longer than (30% of BM cells 110 days after the implant) previously reported after injecting very large doses (10⁸ BM cells infused in different sessions) (27,35,40). These studies used a similar procedure to inject BM cells but did not make a focal niche in the BM to facilitate cell homing. Such an approach has the advantage of avoiding myeloablation but the disadvantage of requiring very large cell doses, which prevent the possibility of autologous transplantation (i.e., such doses need an in vitro expansion of the number of cells, which normally modifies the BMDC characteristics). The present procedure has the advantage of obtaining a similar colonization degree using a low number of cells without the need of in vitro expansion. In addition, the present method focuses cell colonization into a portion of the BM, which could be used to withdraw cell engraftment if necessary. In addition, the myeloablation used here was restricted to a portion of the BM and only persisted for a few days. Therefore, focal myeloablation is much less dangerous than generalized myeloablation and probably less dangerous than partial myeloablation (14) (i.e., no mouse showed secondary complications of focal myeloablation). These features and the simplicity of the procedure used for cell engraftment (a direct injection of cells in the blood stream that could easily be repeated if necessary) facilitate the potential use of this method in humans. Although intra-BM injection of cells is a more complex method (which limits their routine use for repeated administrations), it could be an alternative way to produce focal colonies of BMCs (20,39), particularly if future studies can develop significant colonies with low cell doses and without using a previous myeloablation (47).

Bearing these considerations in mind, the present method based on combining a focal transitory myeloablation with the intravenous injection of BMCs (time locked to the BM ablation) turned out to be an easy, low cost, and repeatable method for producing a stable colonization of the BM, which could be particularly useful for ICT in humans. The colony of BMCs persisted for more than 3 months (a time period equivalent to about 7–8 years in humans), and the percentage of GFP cells increased from day 30 to day 110 after cell injection, thus suggesting that the new cell colony could last for a very long time period. In any case, this time window could be long enough for using ICT in patients with neurodegenerative disorders associated with the elderly (e.g., Parkinson or Alzheimer disease) and whose life expectancy is often no longer than 15–20 years. Because these illnesses progress during years under the action of different factors that continuously promote cell degeneration (29), an efficient cell protection only can be obtained if neuroprotective molecules are permanently available. The brain injection of cells able to release neuroprotective proteins is probably not the better way to get this protection because grafted healthy cells can gradually develop the same pathology as host cells in the disease tissue (4,17,21). This recommends a continuous replacement of neuroprotective cells in the disease tissue. ICT could be a suitable way to get this replacement, avoiding also other disadvantages of direct injections of cells on the goal tissue such as the low spatial homogeneity their protective effects (34). In mice, no side effects were observed after focal myeloablation. If this were also the case in humans, the reported procedure could be a very simple way to introduce and maintain cells in the body when necessary for the cell treatment of chronic illness. Future technical developments to facilitate the production, modification, and storage of BMCs could accelerate the development of ICT, particularly in the case of illnesses which, as is the case of neurodegenerative disorders, could be treated with a constant and long-lasting supply of engineered cells capable of reaching the brain or other tissues by their own means.

In summary, circulating peripheral bone marrow-derived cells can migrate to the brain, where they differentiate into a number of cell types including neurons and glia. We have previously reported that neuronal degeneration can promote the BMC migration across the blood–brain barrier and their penetration into the degenerating brain areas, which represents a promising opportunity for neuroprotective cell therapies. The present study offers a method for creating a long-lasting colony of modified BMCs suitable for continuously replacing the circulating bone marrow-derived cells with modified BMCs, a replacement particularly necessary for treating chronic and progressive neurodegenerative disorders. The present study does not provide information about the migratory capabilities of the long-lasting BMC colony created with present methods; this will be the next step in our studies. The development of BMCs able to prevent neuronal degeneration (e.g., by promoting their release of neuroprotective proteins) and the evaluation of their action in animal models of neurodegenerative illness are necessary before the ITC can be tested on patients.