Transplantation of MSCs for a Longstanding Engraftment and Maintenance of Bone Marrow Stroma

3.1 Nonspecific retention of MSCs in the lungs immediately after injection?

The statement that MSCs are nonspecifically retained in lung tissue is based on the studies in which BM mesenchymal cells were labeled with a radioisotope marker and then injected intravenously into the recipient, and their fate was monitored [9]. The authors observed that after intravenous injection of labeled cells, which in this study were cultured mesenchymal stromal cells (which we will abbreviate hereinafter as “MCs,” and not primary mesenchymal stem cells of BM existing in vivo, which we denote here as “MSCs”), they, through the small circle of blood circulation, got nonspecifically stuck in the lungs, where they were detected 1 hour after intravenous injection. During the first 24 hours after infusion, MCs remained in the lungs, and after 72 hours, they could not be detected in the recipient’s body. The same authors subsequently published a review [10], in which they noted that most of the studies that investigated the ability of stromal cells to transplant used in vitro cultured MCs, which at that time were called “MSCs,” as an object of observation, and that the low ability of such cells to transplant may be associated with changes in their size, morphology, and surface phenotype, including activation of adhesion molecule expression as a result of in vitro cultivation. Indeed, it is known that cultivation of BM stromal cells on plastic leads to the fact that over time, small round-shaped cells turn into large spindle-shaped cells [11]. In another study, cultured MCs have been shown to become as large as 19 μm in diameter [12]. Since this diameter is larger than the size of pulmonary microcapillaries (5–10 μm), it is not surprising that such cells are retained in the lungs after intravenous injection. Besides, in one of the early studies, it was experimentally proved that pre-culturing reduces the ability of “MSCs” to migrate through the basal membrane [13], which creates difficulties for the migration of these cells. Considerations about the ratio of the linear size of cells and the diameter of microcapillaries in the lungs are important when considering the ability of MSCs to overcome this barrier, in light of the fact that “very small embryonic-like stem cells” (VSELs) are known. These cells have been isolated from mouse BM as a population of Sca-1⁺lin⁻CD45⁻ cells expressing SSEA-1, Oct-4, Nanog, and Rex-1 markers characteristic of pluripotent stem cells. VSELs constitute ̴ 0.02% of BM mononuclear cells and are characterized by a small size (2–4 μm), smaller than the diameter of pulmonary capillaries [14]. In addition, VSELs express CXCR4 and thus are potentially capable of homing to the BM. A wide range of differentiation of VSELs has been demonstrated: they are able to differentiate into osteoblasts, cardiomyocytes, neurons, astrocytes, oligodendrocytes, Schwann cells, and insulin-producing cells [14, 15]. Thus, mouse BM contains a relatively rare cell population with osteogenic differentiation capacity that may escape the “lung trap.” It is possible that the BM contains other cell subpopulations with similar properties. Although no direct analog of VSELs has been found in adult humans, VSELs in humans are hypothetical, but we cannot exclude the theoretical possibility of discovering cells with similar properties in the future. For example, a population of small stem cells with diameters of up to 5 μm resembling VSELs was sorted recently from human embryonic stem cell (hESC) cultures [16]. These VSEL-like cells expressed stem-cell-related markers (CD133, SSEA-4) and markers of germinal lineage (DDX4/VASA, PRDM14). These cells were comparable to similar populations of small stem cells sorted from cell cultures of normal ovaries and were the predominant cells in the ascites of recurrent ovarian cancer.

3.2 Absence of MSCs in the organism of an allogeneic recipient in the early terms after injection?

The idea that MSCs may not survive for a long time after administration is based on the data indicating that most MSCs become apoptotic after administration [17], as well as on the results of early studies on the biodistribution of MCs [9]. However, several groups of investigators have shown in different kinds of model animals that MCs can be detected for long periods of time (up to several months) in a wide range of tissues, including BM, after systemic administration, even despite prior in vitro cultivation [18, 19, 20, 21, 22, 23, 24]. In one of these studies, PCR analysis showed that few donor murine cells were present in recipient mice after 1 week, but after 1–5 months, donor cells accounted for 1.5–12% of cells in bone, cartilage, and lung, as well as in BM and spleen [18]. The results demonstrated that at least some of the mesenchymal progenitor cells of BM expanded in culture can not only avoid arrest in lung microvessels and repopulate other organs and tissues but also survive long term in bone, cartilage, and lung and, more importantly, retain their functionality, since the authors were able to show that donor cells represented 2.5% of the chondrocytes isolated from the xiphoid and articular cartilage of the recipient. Interestingly, in these studies on the biodistribution of MCs, their frequency of detection and concentration in BM were often higher than in other tissues. This indicates the propensity of these cells to repopulate BM, which is not surprising if one assumes that the cells have a tropism for the microenvironment from which they were extracted. Recently, a systematic review has been published on the issue of MCs biodistribution [25], which shows that the biodistribution of MCs depends on the route of administration, that after intravenous administration MCs are detected in a wide range of organs and tissues, and that MCs are able to accumulate in damaged organs and tissues. In the majority of studies discussed, it is shown that the BM and spleen are the preferred sites of MCs homing. At the same time, the authors of a number of papers note that the efficiency of MCs homing into the BM is low. It is shown that the ratio of donor stromal cells is estimated as 0.1% of BM nuclear cells [18, 21, 22]. This may be due to the fact that the cited studies performed infusion of pre-cultured cells, and cultivation may result in a reduced ability of cells to invade through the vascular endothelial layer, as shown [13]. In those studies, where it is stated that mesenchymal cells are absent in the recipient’s organism after a short time, the sensitivity of the applied methods could be insufficient to observe rare cell subpopulations, which include MSCs (~0.01%–0.001% of nuclear BM cells).

It can be concluded that at intravenous injection of pre-expanded in vitro MCs, the majority of such cells are indeed retained in the lungs of the recipient for some time. However, this does not mean that rare cell subpopulations present in MC cultures, including cells resembling real MSCs by their properties, are also nonspecifically and irreversibly retained in the lung tissue. Low efficiency of homing, together with detection of cells in other organs and tissues, raised a question about the presence or absence of the mechanism of homing to the BM.

3.3 Lack of a mechanism for homing MSCs to BM?

An example of studies on which this statement is based is the article in which the arrest of MSCs in the lungs was shown [9]. In this study, the biodistribution of MSCs was monitored for 72 hours after intravenous injection of MSCs. At none of the investigated time points were injected MSCs detected in BM [9].

At the same time, by now there is a plenty of researches in which it is shown that intravenously injected stromal cells can, similarly to leukocytes, perform rolling and adhesion to vascular endothelial cells [26] and extravasate into various tissues, including BM, from the vascular channel through the basal membrane [27], although the number of cells migrating through the membrane is usually low [28]. In one of the already mentioned studies, the authors managed to detect eGFP-labeled cells in bones and BM of two recipients using PCR DNA ELISA analysis [19], and subsequently, the presence of MCs was detected in other organs and tissues, but using a more sensitive method [20]. The authors conclude that after intravenous infusion, MCs are predominantly localized in the BM, indicating the possibility of the existence of a mechanism of directed migration (homing) of these cells to the BM. Also, in another study, it was shown [29] that after intravenous injection of cells of stromal line GB1/6 to mice irradiated at doses of 3–8.5 Gy with additional irradiation of the right hind limb at a dose of 10–12.5 Gy, the injected cells were detected in situ in the hematopoietic sinuses of the recipient BM. Donor cells were undetectable 2 months after infusion in other organs—in the spleen, liver, and lungs. This indicates that intravenously infused stromal cells can infiltrate the vascular system of the BM and selectively accumulate in it [29].

Thus, both in animal models and in humans, it has been shown that after intravenous administration, stromal cells isolated from BM are capable of homing into their “native” BM niches, although with relatively low efficiency [28]. What limits the tropism and engraftment of MSCs in the BM remains not fully understood. There have been published papers showing the ability of cultured BM mesenchymal cells to migrate through the basal membrane and endothelial cell layer in model in vitro experiments and demonstrating the role of the metalloprotease MMP2, which is able to cleave type IV collagen (a major component of the basal membrane) in this process [27]. Interestingly, it is sufficient to culture primary BM-derived mesenchymal stromal cells for 24–48 hours for their ability to home to BM to be significantly reduced [13]. This information explains why, in many experiments on intravenous injection of multipotent mesenchymal stromal cells (MMSCs), it was not possible to observe significant homing of injected cells into the BM. At the time of publication of the described study, the mechanism of this phenomenon remained unknown, but much later it was found that MMSCs cultivation leads to a decrease in the expression of CXCR4 on their surface—a key molecule in the migration of various types of stem cells. The studies on the mechanisms of directed migration of MSCs and BM-MCs can be conditionally divided into those that show mobilization of stromal cells from BM into peripheral blood and those that study the mechanisms and pathways of migration of BM-MCs after their intravenous administration. It was noted that intravenously injected BM-MCs are able to accumulate in the sites of damage, hypoxia, and ischemia [30]. This ability of MMSCs was confirmed in myocardial infarction and ischemic brain injury [31, 32]. It has been hypothesized that proinflammatory cytokines, chemokines, including SDF-1α, and hypoxia factors, such as HIF1α, are involved in the mechanisms of MSCs migration to various tissues, and, in particular, to the BM, in which hypoxic conditions are maintained physiologically [30]. The following studies are devoted to this issue.

Among chemokines and their corresponding receptors, the SDF-1α/CXCR4 axis is a thoroughly studied system [30, 33, 34, 35, 36, 37]. This axis is known to be central to the homing of HSCs to the BM when administered intravenously [38], regulating the migration and homing of CXCR4+ hematopoietic progenitor cells, pre-B lymphocytes, and T-lymphocytes [33, 34, 35, 39]. More recent evidence suggests that in addition to HSCs, functional CXCR4 is expressed on the surface of various types of tissue-specific stem cells, such as neuronal stem cells [40], skeletal muscle satellite cells [41], liver oval cells [42], primordial germ cells [43], and other stem cell types including Muse cells and VSELs [44, 45].

There has been considerable evidence to suggest that the SDF-1α/CXCR4 axis is important for the migration of mesenchymal stromal/stem cells isolated from various sources [36, 37, 46, 47] and, more importantly, determines their ability to home to BM after intravenous administration [48]. Interestingly, data from previous studies suggested that a small subpopulation of cultured MSCs/MMSCs/BM-MCs initially maintains CXCR4 expression on the cell surface [48, 49], but upon ex vivo expansion, expression gradually declines and becomes barely detectable after prolonged cultivation [50]. This phenomenon may partly explain the fact that many researchers have failed to observe efficient repopulation of intravenously injected MSCs/MMSCs/BM-MCs after their preliminary expansion in vitro.

Kyriakou et al. showed that despite low surface expression of CXCR4, human MSCs/MMSCs/BM-MCs migrated in response to SDF-1α in vitro, and this was enhanced when CXCR4 was overexpressed in cells by lentiviral transduction [51]. It was also shown that artificial overexpression of CXCR4 in human adipose tissue stromal cells (hADSCs) helps to direct these cells to damaged tissues in response to SDF-1α signaling [36]. Chen et al. observed the same correlation both in vitro and in vivo for CXCR4-transduced murine MSCs transplanted into irradiated mice [52]. Moreover, Bobis-Wozowikz described that transplantation of CXCR4-transduced human MMSCs isolated from adipose tissue into NOD/SCID mice resulted in increased engraftment to BM [53]. Interestingly, CXCR4 overexpression caused a significant increase in homing to the BM and spleen of animals that had previously received irradiation [51, 52]. This indicates that prior damage to the BM stroma may be a prerequisite or at least a factor contributing to the engraftment of intravenously injected mesenchymal progenitor cells in the recipient. It has been shown that irradiation of an animal with ionizing radiation or exposure to other DNA-toxic agents leads to increased expression of SDF-1α in BM [54], and this may be one of the mechanisms behind the increased ability of MSCs to engraft in the BM of irradiated mice [55].

Enhanced survival and ability of MSCs to migrate in vitro in response to hypoxia and cultivation in the presence of SDF-1α have been demonstrated [46, 56]. The ability of BM-MCs to migrate along the SDF1a gradient was experimentally demonstrated in a mouse model of hypoxic-ischemic brain injury (HIB), and this effect was abrogated by the CXCR4 inhibitor (AMD3100) [30].

These prerequisites, namely the expression of CXCR4 in BM-MCs at basal level and its activation in response to hypoxia, the local increase in SDF1α levels under hypoxic conditions, which can occur as a consequence of ischemization or injury, are also physiological in BM [57], together with the ability of BM-MCs to accumulate in vivo in ischemic and injured tissues, were the grounds to suggest that the SDF-1α/CXCR4 axis is the key and determines the ability of homing to BM not only of HSCs but also MSCs or their progeny.

Thus, summarizing the presented data, we can conclude that the ability of MSCs to chemotaxis along gradients of SDF1α is well documented. This is primarily shown for the migration of MSCs to injury sites. Among other things, it has been shown that CXCR4 expression is increased on the surface of MSCs in response to several chemokines, including SDF-1α. Stromal cells have been shown to be able to migrate along the SDF-1α gradient and penetrate the endothelial barrier in the BM from peripheral blood. At the same time, the efficiency of MSCs homing into BM is usually low, which is associated by researchers with a low level of expression of receptors to chemokines on the surface of primary, uncultured MSCs; an even greater decrease of expression of these receptors during in vitro cultivation of MSCs/BM-MCs isolated from the organism; and heterogeneity of the BM-MCs population.

3.4 Loss of MSC functions due to cell dissociation?

One of the significant arguments that can be heard in favor of the view that transplantation of BM stroma is impossible is the notion that MSCs lose their functional abilities to regenerate stromal tissue after dissociation of intercellular contacts and transformation of BM into a single-cell suspension before intravenous administration to the recipient. The conclusion that stromal cells are not capable of transplantation in the form of a single-cell suspension was made by several independent, authoritative researchers in the 70–80s of the twentieth century who conducted experiments on animals. In similar studies by Ivanov-Smolensky and Friedenstein, BM explants from (CBA x C57BL/6)F1 mice were used to investigate this issue. F1 mice were lethally irradiated at a dose of 11 Gy and reconstituted with BM from CBA mice, and BM explants were obtained from the recipients. Colonies of fibroblasts grew from the explants, which, judging by isoantigens in the indirect immunofluorescence reaction with anti-C57BL/6 serum, were of recipient origin [58, 59]. The same conclusion about the impossibility of stroma transplantation was made by Chertkov [60, 61]. In 1983 the authors demonstrated that if BM is explanted in culture as a whole fragment without cell separation and then the stromal sublayer is collected and transplanted under the kidney capsule, an ectopic focus of hematopoiesis is formed in the implantation site, i.e., MSCs fulfill their function and build a full-fledged hematopoietic microenvironment [60]. However, if BM is transformed into a single-cell suspension before explanting into the culture, the cellular composition of the stromal sublayer of the culture is quite different, and, most importantly, when it is implanted under the kidney capsule of the recipient mouse kidney, a focus of ectopic hematopoiesis is not formed, i.e., the function of MSCs is lost due to the preliminary dissociation of cells. Then, 2 years later, following a largely similar scheme to the experiments of Ivanov-Smolensky et al. [59], Chertkov et al. studied the cells of the stromal sublayer of LTBMCs from the BM of radiation chimeras, which were obtained by irradiating mice with a lethal dose of radiation and reconstituting hematopoiesis by intravenous injection of donor syngeneic or allogeneic BM in the form of a single-cell suspension [61]. The origin of stromal cells in the LTBMCs of chimeric BM was studied using double fluorochrome staining with mouse lineage-specific antiserums. It was shown that the plastic-adherent cells of the LTBMCs of allogeneic chimeras contained only recipient stromal progenitor cells. The authors concluded that stromal progenitor cells cannot be transplanted intravenously in mice.

It is interesting to note that J. Chertkov and A. Friedenstein published the results of their other experiments, in which the strength of the basis of the hypothesis put forward is questionable. Thus, for example, in the same issue of Experimental Hematology, I.L. Chertkov publishes another article [62], in which a similar experiment of forming the foci of ectopic hematopoiesis from a whole fragment of BM and from a single-cell suspension of BM, which was pelleted by centrifugation on a small fragment of a filter, was conducted. The difference between this study and the study described above by the same authors was that in the latter study, the stage of seeding the BM cells in culture was omitted. In this arrangement of the experiment, foci of ectopic hematopoiesis were formed in both cases. To the surprise of the authors, the dissociation of BM cells not only failed to prevent the formation of foci, but other way round the proportion of donor cells in such foci was increased. Also, Friedenstein published a study in which it was shown that transplantation of fibroblasts in the form of a suspension under the kidney capsule led to the formation of elements of BM stroma, including bone, and to the presence of hematopoietic cells in the implants, which indicated that dissociation of cells in these experiments still did not lead to a complete loss of MSC function [63].

The statement that MSCs lose their functions after dissociation of intercellular contacts directly contradicts later studies, where it was shown that MSCs are able to migrate in the organism to the microenvironment of solid tumors and to the site of inflammation [64, 65, 66]. For example, MSCs have been shown to be selectively recruited to injured or inflamed tissues due to the release of multiple inflammatory factors by injured tissues [67]. This process, in particular, is essential for the resolution of infection and bone regeneration. The ability of MSCs to migrate is increased during acute inflammation, a process associated with the release of tumor necrosis factor (TNF) and activation of the NF-κB pathway [68, 69]. It is known that MSCs migrate from the BM to the gastric cancer microenvironment [70]. Thus, MSCs or some of their descendants are able to move around the organism as separate cells, accumulate in inflamed or damaged tissues, and actively participate in regeneration processes. Consequently, dissociation of BM cells before their administration to the recipient does not necessarily mean the cessation or decrease of MSC functionality.

3.5 Lack of immune privileges for MSCs?

The question of immune privileges (IPs) of MSCs and other stromal progenitor cells has a long history of study. In vitro studies of immunomodulatory properties of MMSCs have shown that these cells are able to inhibit proliferation and activation of T-lymphocytes during cocultivation, express MHC class I at a low level, and do not express MHC class II on their surface [71, 72, 73]. It has also been shown that these cells and the extracellular vesicles and individual proteins and other molecules secreted by them are able to reduce the activation of immunologic responses in a number of autoimmune pathologies and in graft-versus-host disease after alloHSCT [74, 75, 76].

However, it was subsequently found that pre-cultured BM-MCs do not possess immune privileges, and their main part causes immune responses in the organism and is eliminated by the immune system in MHC-incompatible transplantation [77, 78]. However, even in the case of cultured cells, the possibility that among the total cell population, there is a rare population of cells still possessing IPs cannot be excluded. Indeed, compelling experimental evidence for IPs of BM-MCs in vivo has subsequently emerged and includes studies performed in mice, large primates, and humans. We will describe here only a few of them.

Long-term presence of donor BM stromal cells in the body of recipients after alloHSCT has been demonstrated: in the already mentioned study of the ability of BM-MCs to home to BM in primates, in addition to autoBMT, MHC-incompatible allogeneic combined infusion of HSCs and genetically labeled BM-MCs was performed [19]. It was shown that markers of donor MCs were detectable in the BM up to +76 days. The same authors published another study on primates, in which they injected autologous GFP-labeled MSCs into a baboon previously lethally irradiated and recovered by its own hematopoietic cells mobilized into PB before irradiation and a baboon that had not been irradiated or exposed to any cytotoxic therapy [20]. MSCs expressing immunogenic protein GFP persisted in the recipient’s organism for several months (up to 21 months) without being eradicated by the immune system of the organism [20]. For human allogeneic MSCs, their long-term (for several months) presence in the BM of a recipient allogeneic to the donor after intraosseous injection of a suspension of donor MSCs has been demonstrated [79, 80].

The IPs of MSCs have proven to be so strong that they allow these cells to survive in a xenogeneic organism. One study demonstrated the presence of xenogeneic human MSCs in the BM of immunodeficient (nude) mice for several months after intravenous infusion [23]. In another study, the authors transplanted a well-characterized population of human MSCs to sheep embryos at the early stages of pregnancy, before and after the expected development of immunological competence [81]. In this xenogeneic system, human MCs engrafted and persisted in various tissues for 13 months after transplantation. The transplanted human cells underwent site-specific differentiation into chondrocytes, adipocytes, myocytes, cardiomyocytes, BM stromal cells, and thymus stroma. Long-term engraftment was unexpectedly recorded even when cells were transplanted after the development of immunocompetence.

Recently, it was shown that MSCs have pronounced IPs in vivo in adult mice: MSCs expressing immunogenic marker GFP survived and maintained their functionality for 6 weeks in immunologically complete adult mice [82, 83]. It is known that in vitro cultivation significantly changes the properties of cells extracted from the organism. For example, it has been shown that MMSCs cultivation leads to the induction of MHC expression, changes in the expression of genes encoding proinflammatory cytokines [73, 84], changes in the global DNA methylation and gene expression profile, and the occurrence of chromosomal rearrangements [85, 86]. As a consequence, it is legitimate to suggest that the immunological characteristics of cells in vivo and in vitro differ significantly.

Thus, it can be concluded that in mammalian BM, there are one or more distinct cell populations that have IPs and can be successfully transplanted to an MHC-compatible or incompatible recipient and survive in their body despite an active immune system.

If we think about stroma transplantation in patients as a component of alloBMT, we can assume that the probability of survival of these cells is much higher than in other settings, because at the moment of infusion of donor cells and for a certain period of time after it, the patient is in deep agranulocytosis, and the activity of the remaining circulating T-lymphocytes of the recipient is significantly reduced. Humoral immune response to MSCs and their differentiating progeny is also unlikely in such a situation. If combined transplantation of MSCs and HSCs is successful, the recipient becomes a “chimera:” his hematopoiesis and immune system become donor and, consequently, syngeneic with injected MCs, so there is no reason for a systemic immune response to injected donor MCs. However, local immune reactions presumably can be observed due to the preserved resident macrophages of the recipient residing in organs and tissues and residual circulating T-lymphocytes.

It can be concluded that for today there are reliable and numerous experimental confirmations that at intravenous injection of whole BM cell suspension containing true multipotent, not subjected to pre-cultivation MSCs, the latter are able to survive for a long time in the recipient’s organism, migrate into BM using specialized molecular mechanisms, and perform their direct function—to participate in regeneration and maintenance of BM stromal tissue.

4.1 The strategy

This alternative concept consists of the following key steps/stages (Figure 1).

A hypothetical timeline for the four-step protocol is as follows. Eradication of BM stroma could be done after induction therapy and simultaneously with consolidation chemotherapy. Infusion of donor MSCs should be done in advance, for example, at day-14, in order to give the cells some time to home to the BM, engraft, and reconstitute BM stroma. Infusion of HSCs on day 0. Supportive and regenerative therapy for at least 3 months after HSCs infusion. This timeline is speculative and theoretical; a real-time course should be established in future experiments.

The first step is “Eradication of the recipient’s BM stroma.” The goal of this step is to eradicate the host’s MSCs in the BM and/or significantly damage the patient’s BM stroma. This is necessary so that the donor MSCs, which will be injected into the recipient at the next step, would get into the organism in which the function of the BM stroma is lost/damaged and, reacting to this, carry out their direct duty—to regenerate the lost/damaged tissue. A direct analogy can be made here with alloHSCT. This procedure is aimed not only at the elimination of tumor cells but also at almost complete destruction of the patient’s own hematopoiesis, including his immune system, as well as emptying hematopoietic niches in the BM. Destruction of the immune system and emptying of hematopoietic niches create favorable conditions for the introduced donor’s HSCs to engraft by populating empty niches without experiencing pressure from the recipient’s immunity, activating their regenerative program, and fully restoring hematopoiesis. It is likely that successful engraftment of donor HSCs and reconstitution of hematopoiesis is not in the least due to the fact that HSCs enter the body with the lost/damaged hematopoietic tissue and activate their regenerative function to replenish it. At this step, various strategies and approaches can be used, the detailed development of which is still to come. One such approach is nonspecific stromal damage by radiation or polychemotherapy (PCT). Several studies that have demonstrated successful engraftment of the donor’s stroma have used high doses of radiation [29, 89], so in principle, this approach is feasible. However, it should be remembered that stromal cells are much more resistant to ionizing radiation [96, 98] than hematopoietic cells. This means that if ionizing radiation is considered as a factor capable of damaging MSCs and applied as TBI, it may require too high doses of radiation to effectively suppress the stromal component of BM. The side effects of using such doses may be acute radiation sickness, high cytotoxicity, and the occurrence of secondary tumors. It is doubtful that in the case of TBI it will be possible to achieve doses that would, on the one hand, lead to destruction of BM stroma or at least MSCs within the stroma and, on the other hand, be acceptable in terms of toxicity/complications profile. Radiation therapy is actively developing, and methods of precision irradiation of cells with a precision of up to 1 millimeter are available today, which is a promising alternative to TBI. The development of 3D-4D conformal radiation therapy approaches, such as CyberKnife radiosurgery technology, which is used for radiosurgical treatment of solid tumors, primarily brain tumors and vascular malformations, has become more accessible nowadays. This technology allows irradiation of areas with an accuracy of less than a millimeter and, importantly, affords continuous tracking of the target position and correction of the course of irradiation when the target shifts with respiration or when the patient moves intentionally or not. This makes it possible to selectively and precisely apply ionizing radiation to all areas where BM is concentrated—bones of the skull, ribs, spine, pelvis, and epiphyses of limb bones. Another method of stromal eradication may be the administration of chemotherapeutic drugs that have a cytotoxic effect on stromal cells. It is difficult to think of any significant selectivity with respect to them, although it can be assumed that those drugs used for the treatment of malignant bone tumors and, in particular, osteosarcomas—tumors arising from osteoblasts—may be more effective here than other substances [99]. These drugs include cisplatin, doxorubicin, ifosfamide, etoposide, epirubicin, and methotrexate [100]. This list is certainly not exhaustive; additional information can be drawn from the Russian national clinical guidelines for the treatment of malignant bone tumors or the US National Comprehensive Cancer Network (NCCN) national clinical cancer guidelines [100]. Other chemotherapeutic agents with distinct effects on stromal cells can be selected in functional tests on stromal cell lines (e.g., OP-9) [101, 102], fibroblasts [103], or on primary cultures of BM-derived MMSCs, adipose tissue, or other sources [104], although it is most appropriate to test drugs specifically on primary, previously uncultured BM stromal cells, since this tissue is the therapeutic target in this case, and cultured MSCs isolated from different sources other than BM vary in their ability to proliferate, differentiate, migrate, and secrete extracellular matrix, which may result in different sensitivity to the drugs being tested [105], and cell culturing itself leads to changes in cell characteristics, as mentioned above. It is also worth noting that the tumor microenvironment (TME) is currently the focus of attention of oncologists and oncology researchers, and there are actively developing therapeutic approaches that target TME. Tumor-associated fibroblasts (CAFs) are one of the significant components of TME, and attempts are being made to targetedly eradicate these cells in order to improve the effectiveness of the treatment of primary and metastasizing solid tumors [106]. The most effective drugs, which were discovered in these studies, can be considered as promising candidates for the eradication of BM stromal cells within the proposed new approach to alloBMT and treatment of stromal tissue diseases. Another therapeutic tactic for stromal eradication may be targeting certain cell subpopulations, in particular, MSCs in BM or their differentiated progeny involved in niches for HSCs, such as osteoblasts. Such tactics require knowledge about the surface markers specific to these cells. Unfortunately, to date, there is no unambiguous opinion on the surface immunophenotype of BM MSCs: different researchers mention different surface markers and believe that cells expressing them satisfy the definition of MSCs. Some authors argue that MSCs are cells expressing Lepr and Cxcl12 [107]. It has also been shown that the population of stem/progenitor mesenchymal cells in BM is significantly enriched in a subpopulation of CD271⁺ cells [108, 109]. Other researchers consider MSCs to be pericytes characterized by the expression of smooth muscle α-actin (Acta2) [107, 110]. Finally, CD45⁻Nes⁺ cells are considered MSCs [93]. They are exclusively located along BM arterioles and are a component of hematopoietic niches that regulate the quiescent state of HSCs [111]. The mentioned subpopulations can partially overlap with each other [111], but the ability to transfer hematopoietic activity in vivo has been clearly demonstrated so far only for Nes⁺ MSCs [93]. In addition, previously described Muse cells are characterized as SSEA3⁺ cells [112], and VSEL cells are described as Sca1⁺CXCR4⁺SSEA1⁺lin⁻CD45⁻ [14]. By definition, MSCs are capable of differentiating in at least three classical directions: osteogenic, adipogenic, and chondrogenic. This ability ensures the formation and maintenance of the basic cellular elements of the BM stroma. As a result of MSC differentiation in the osteogenic direction, osteoblast-osteocyte precursors are formed, the main function of which is the formation of new bone tissue due to active secretion of extracellular matrix substances. Some studies claim that osteoblasts are direct participants of niches for HSCs and their progeny in the BM [113, 114, 115]. Considering the important role of osteoblasts in HSC maintenance, a special term, “osteoblastic niche” of HSCs, was even introduced. However, it should be noted that in a number of papers, authors have disputed the importance of osteoblasts in HSC maintenance [116]. Osteoblasts are exclusively present in BM, lining the inner surface of bone. Osteoblasts are characterized by the expression of the surface membrane proteins CD10 [117], CD44 [118], CD46, CD55, and CD59 [119], as well as the recently discovered A7 antigen, which has strong specificity [120]. There is reason to suggest that osteoblasts are a favorable target for chemotherapy aimed at damaging the stroma as part of the preparatory step before its transplantation, because osteoblasts express a number of soluble extracellular matrix proteins, such as osteocalcin, osteopontin, procollagen, procollagen peptidase, and bone sialoprotein. The content of these proteins can be measured in peripheral blood, and their levels can serve to assess the efficiency of osteoblast destruction along with soluble forms of the osteoblast surface antigens described above. Moreover, since osteoblasts are present specifically in bones, while MSCs are distributed throughout the body, this makes the former a more specific target, and their destruction, presumably, may have fewer undesirable side effects on the body than the destruction of MSCs. Thus, targeted chemotherapy can be directed at osteoblasts either separately or in addition to targeted therapy directed at MSCs. The arsenal of methods for targeting specific cell types is currently quite wide and includes not only chemical agents but also various types of antibodies (monoclonal antibodies, bispecific antibodies, and mini-antibodies conjugated to a cytostatic agent) [121, 122], and T-lymphocytes or NK cells with a chimeric T cell receptor (CAR-T, CAR-NT cells, respectively) engineered to specifically recognize a particular target [123, 124, 125, 126]. Such CAR-T or CAR-NT cells can be generated against a single target in a cell population selected for exposure. CAR-T and CAR-NT cells can be used in cocktails for a more efficient cytotoxic effect against a selected cell subpopulation [127].

Various combinations of nonspecific and targeted drugs directed at different types of target cells and at different antigens on the surface of these cells are possible. This part requires detailed development in preclinical studies.

Theoretically, such a strategy can radically transform the schemes of induction and consolidation therapy before alloBMT: such schemes and regimens can be developed that will be primarily directed not so much at hematopoietic, tumor, and immune cells but at BM stroma. No niches for HSCs—no hematopoiesis and no leukemia. At the same time, it is likely that PCT regimens will retain drugs aimed at reducing the total mass of leukemia cells in the blood/BM in order to reduce the negative impact of these cells on the body. An additional option for promising regimens may be to affect T regulatory cell (Treg) populations in the recipient, as it is these cells, together with MSCs, that confer immune privileges to HSCs in the BM, keeping the latter dormant, and thus may contribute to the persistence of leukemia stem cells in the BM [128, 129]. On the other hand, Tregs may affect the manifestation and severity of GVHD [130], so a balance between benefit and risk must be maintained when targeting Tregs.

The second step of the proposed concept is “Preparing the foothold,” which is an infusion of donor whole BM cell suspension or selected subpopulations of MSCs. The aim of this stage is to repopulate the vacant BM territory with donor MSCs so that they begin to recreate the hematopoietic microenvironment in the BM and build niches for HSCs. It is important to emphasize that the aim of this stage is not to reconstitute hematopoiesis at the expense of donor hematopoietic cells. The efficiency of engraftment of donor HSCs and, as a consequence, donor chimerism of hematopoietic tissue will be low due to the fact that at this time point in the recipient’s body, there will be significantly damaged or destroyed niches for them. Consequently, at this stage, not the whole BM but only its stromal component or separate stromal cell subpopulations, in particular, MSCs, if it is possible to isolate them in any way, are feasible to be infused. On the other hand, infusion of whole BM implies infusion of various subpopulations of immune cells, which may support the recipient’s immunity for some time at the expense of granulocytes and memory lymphocytes.

The third stage is “Airborne.” This stage consists of the infusion of donor HSCs, as is currently done in alloHSCT. This (second) infusion should probably be performed some time (what time, it is to be determined) after the first one so that MSCs have time to build new niches for HSCs. This will allow the HSCs to inhabit the niches/territories recreated for them and engraft effectively. Probably due to the fact that the stroma recovery may not be complete or the efficiency of its transplantation will not be too high, it will be necessary to adjust the therapeutic doses of CD34⁺ cells administered to the recipient toward their increase.

The fourth stage is “Support/Supply Pathway Establishment.” Damaging the BM stroma can have serious negative consequences for the patient’s body. The injected donor MSCs may not be able to cope effectively with the task of building a new hematopoietic territory throughout the body. They may need support. What kind of support is the best, should be subject of further research. It could be supportive infusions of CAR-MSCs, which have pronounced homing ability to BM and significant immune modulatory effects [131, 132], or agents that stimulate regeneration of nerves in the BM [133], or chemotherapeutic agents used for prophylaxis and treatment of GVHD.

At each of these stages, it is advisable to utilize factors that can increase the efficiency of BM stromal transplantation; their combined application can result in a significantly increased likelihood of high-quality, stable, complete BM stromal transplantation. These factors may include effective eradication of the recipient BM stroma. It has been shown that irradiation of an animal with ionizing radiation or exposure to another DNA-toxic agent, such as 5-fluorouracil (5-FU) or cyclophosphamide (Cy), leads to increased expression of SDF-1 in the BM [54]. The ability of BM-MCs to migrate along the SDF1a gradient was experimentally demonstrated [30]. Together with overexpression of CXCR4 on BM-MCs, it could facilitate the engraftment of injected cells [52]. Another factor that can increase the efficiency of MSC transplantation is the introduction of an increased amount of these cells. In those studies, in which it was possible to demonstrate long-term presence of donor stromal cells in the recipient’s organism, the authors attributed the success to the increased dose of injected cells [28, 55, 89, 90]. Another factor that is likely to be significant in the engraftment of stromal cells is the use of vasodilators. It has been shown that intravenous injection of rat BM-MCs after vasodilator administration increased the efficiency of accumulation of radioactive label embedded in the cells by 50% compared to a group that was not injected with a vasodilator [21]. Finally, after infusion of mesenchymal cells into the recipient, it is advisable to consider supportive therapy that will promote engraftment of the injected cells, their proliferation, and differentiation for complete reconstruction of the stromal component of the bone marrow. Many cytokines have a stimulatory effect on mesenchymal progenitor cells. For example, it has been shown that the basic fibroblast growth factor FGF2 stimulates the proliferation of BM-MCs and their differentiation into tenocytes [134]. Transforming growth factor beta (TGFb) takes part in adipogenic differentiation of MSCs [135]. Also, immunity factors have an important influence on MSCs. For example, it has been shown that interleukin-1 beta is a growth factor for stromal progenitor cells [136, 137], and tumor necrosis factor (TNF) and interferon gamma can influence the immunomodulatory properties of BM-MCs [138, 139]. It is reasonable to consider the possibility of supportive therapy aimed at the regeneration of BM stroma and hematopoiesis. Sensory neuronal signaling is known to modulate the functionality of MSCs and osteoclasts in BM [140]. It has been shown that active proliferation of tumor cells in BM in acute myeloid leukemia (AML) has a deleterious and persistent effect on the sympathetic neuronal component of HSC niches in human BM, which can lead to long-term hematologic dysfunction [133].

4.2 Risks of the proposed strategy

Despite the considerations presented about the potential benefits of BM stroma transplantation, it is necessary to evaluate all possible risks to the patient and to ensure that the safety and clinical efficacy of this procedure are appropriate. The risks associated with the proposed procedure of MSC transplantation after prior eradication therapy are highlighted in this chapter. Given the need for high-dose radiation therapy and/or potential intensification of induction and consolidation protocols to achieve profound damage to the stromal component of the BM, the risks of secondary tumors, a higher incidence of infectious complications, and associated mortality as well as some other risks should be evaluated [141, 142, 143].