Mechanobiology of mesenchymal stem cells: Which interest for cell-based treatment?

1. Introduction

Mesenchymal stem cells (MSC) were first isolated from bone marrow (BM) stem cell niche in the middle 60s. Then, these cells have been found in several adult tissues (muscle, adipose tissue … ) but also in fetal ones (umbilical cord matrix, placenta). MSC which have a mesodermal origin were defined, in 2006, according to the ISCT (International Society for Cellular Therapy) standards by (i) plastic cell adhesion in in vitro culture conditions, (ii) expression of some surface differentiation clusters (CD) such as CD73, CD90, CD105 but absence of hematopoietic markers as CD34 and CD45, (iii) osteoblast, adipocyte and chondrocyte differentiation capacity [1]. In addition, immune properties of MSC have been demonstrated both in vitro [2] and in vivo [3]. These properties rely on three main mechanisms: the arrest of the cell cycle of immune cells at the G1 phase, a direct interaction of MSC with immune cells and a paracrine effect. The main factors expressed by MSC in response to immune cell interactions are: microRNAs, Programmed death-ligand 1 (PD-L1), histocompatibility antigen G (HLA-G), Prostaglandin E2 (PGE2), cytokines (interleukin (IL) 6, IL10) and growth factors (transforming growth factor-β (TGFβ), Hepatocyte growth factor (HGF), Vascular endothelial growth factor (VEGF)), galectin(s) and one enzyme (Indoleamine 2,3-dioxygenase: IDO). Thanks to their differentiation potential and their immune properties, in 2016, MSC have been proposed in more than 650 clinical trials [www.clinicaltrials.gov]. The indications are numerous and concern either tissue engineering (bone, cartilage, ligament diseases, vascular diseases…), or MSC-based therapy. In this case, MSC are administrated to patient to treat mainly autoimmune diseases (graft-versus-host disease (GVHD), liver cirrhosis, systemic lupus, erythematsus, rheumatoid arthritis, type I diabetes…) [www.clinicaltrials.gov]. For MSC-based therapy, cell administration could be performed either by topical administration or by systemic administration. In preclinical studies and clinical trials, the most common method used is the intravenous injection for its convenience. To be efficient, MSC-based treatments still rely on adequate cell homing towards the inflamed or injured site. Once they have reached the site, MSC will have to act with immune cells while being subjected to the mechanical environment of the concerned tissue. The aim of this review is to understand how the mechanical environment could interact with MSC behavior once they are injected to patients.

2. Mechanobiology and biomechanics concepts

A fundamental property of living tissues is their adaptation to the environment and to the movement (a property described by Roux at the end of the 19th century). These movements are at the origin of local mechanical stresses and constitute stimuli which can lead to modifications of the biological behavior of cells [4]. The mechanobiology concept envisages the biological evolution of cell, tissue or organ, by taking into account the mechanical forces that it undergoes. Thus, cell mechanobiology studies the cellular processes implied in a biological situation taking into account the mechanical and chemical environment. The mechanobiology concept regroups some other concepts as mechanoactivation or mechanostimulation and mechanotransduction (Fig. 1). The mechanoactivation or mechanostimulation consists in stimulating cells or tissues with mechanical stimuli and in studying cell or tissue behavior (viability, differentiation, matrix synthesis…). The aim is to understand the role of mechanical factors in regulating cell and tissue physiology. The mechanoactivation may originate from external mechanical forces but also from substrate stiffness [5]. The use of external mechanical stimulation may help to improve or enhance tissue regeneration in vitro. The most popular mechanical stimuli used in in vitro studies are stretching, shear stress, compression or hydrostatic pressure. Together, they mimic the main mechanical stimuli present in vivo. The disadvantage of in vitro studies is based on the lack of standardization of protocols and results. Therefore, comparisons between studies are delicate. Indeed, it has been reported that different influences on cellular physiology are caused by different parameters including their magnitude and duration. This is the case for shear stress where we could define laminar, pulsatile and oscillatory shear [6,7]. Lastly, if it is well known that mechanical stimulation affects proliferation, differentiation and matrix synthesis, it seems that mechanical stimulation could also influence immune properties of cells. Recently, Becquart et al. have shown that intermittent shear stress and hydrostatic pressure affected differently the mesenchymal stem cell behavior with a different profile in secreted factor ability [8]. Thus, it was shown that PTGS2, VEGFA and FGF2 genes encoding for soluble factors (respectively PGE2, VEGF and FGF) are expressed by MSC stimulated by intermittent shear stress but not by MSC stimulated by cyclic hydrostatic pressure.

OPEN IN VIEWER

Fig. 1.

Interrelationship between mechanobiology, biomechanics, mechanoactivation and mechanotransduction.

The mechanotransduction concept consists in studying pathway signaling after mechanical stimulation of the cells. The process of mechanotransduction corresponds to a signal of adaptation of the cells and/or tissue. Today, many signaling pathways have already been described in the literature and show a real complexity to understand how mechanical forces are transmitted to cells. For Luo and Robinson, a mechanotransduction system requires at least the coupling of one sensor and one transducer [9]. The sensors, located next to the outer layer of the cell membrane, sense the mechanical stimuli, such as force, pressure and flow speed, and transmit the mechanical signals into the inside of cells. The actin cytoskeleton composed of actin filaments, actin crosslinking proteins and Non-Muscle Myosin II (NMM-II) proteins, is unambiguously involved in these mechanosensing processes and maintains a physical link between cell environment and nucleus. Indeed, a recent work indicates that mechanical stress emanating from the cytoplasmic cytoskeleton can activate pathways in the nucleus which eventually impact both its structure and the transcriptional machinery since cytoplasmic cytoskeleton, nuclear envelope and nucleoskeleton are connected via LINC (Linker of Nucleoskeleton and Cytoskeleton) complexes [10]. In this case, nucleus also acts as a mechanosensitive structure. Some others sensors have also been described as ion channels, protein kinases, integrins, membrane glycocalyx, G proteins and intercellular junction proteins [9]. Within seconds following activation of mechanosensors, a number of processes begin such as release of secondary chemical messengers (calcium, cyclic AMP, etc.), protein phosphorylation, and remodeling of cell–cell adhesions. Since each cell may express multiple mechanotransducers, it is difficult to discern the primary mechanochemical step leading to mechanical sensing.

Among all these concepts, a last concept is fundamental to establish the link between cell/tissue/organ response and the interaction with their environment. It is biomechanics concept. This technique applies the principles of mechanics with the comprehension of the processes and the biological functions. Its main aim is thus to characterize organ/tissue/cell movement, and their mechanical components (forces, deformation…) by keeping into account the loads acting on it. The resistances to the deformation or to the strains are some parameters studied in biomechanics. According to Guilak et al. biomechanics topics include measurement and modeling of the in vivo biomechanical environment; quantitative analysis of the mechanical properties of native tissues, scaffolds, and repair tissues [11]. This technique could also investigate development of rationale criteria for the design and assessment of engineered tissues; investigation of the effects biomechanical factors on native and repair tissues, in vivo and in vitro; and development and application of computational models of tissue growth and remodeling [11].

Studies based on mechanobiology and biomechanics have largely helped tissue engineering to develop. Their objectives were (i) to understand the functioning of normal or pathological tissue, (ii) to develop scaffolds closed to these tissues and (iii) to investigate how physical cues of cellular environment affect MSC fate and differentiation. The use of these concepts for cell based therapy could also help scientists to understand how MSC migration in blood circulation is.

MSC administration could be done by either topical administration or systemic administration. Topical administration consists in intralesional injection and local vascular injection which present the advantage that MSC arrive directly at the target tissue with little loss during migration. This is not the case for systemic administration including intravenous injection, intraperitoneal injection and intra-arterial injection. In systemic administration routes, cells will end up in the blood circulation and will have to migrate to the target site. The most common method used, in clinical trials relative to cell-based treatment, is the intravenous injection. MSC dosages vary from 1 to 3 × 10 6 cells per kg of patient body weight, according to different protocols [12]. Once injected, cells are found in the blood circulation and will have to migrate to the site of injury. When they have reached their target, MSC have to adhere to the tissue vasculature, and then to transmigrate from the vasculature into the injured tissue. This process is called homing. However, the lack of data describing the exact positioning of the MSC following infusion makes it difficult to determine if the cells have stayed within the vessels or have undergone transendothelial migration [13]. Several authors have described that MSC administered via intravenous injection are more easily trapped in lung, liver or spleen because of their larger size and expression of cell adhesion molecules like integrin CD49f or CD49d [14]. In turn, a decrease of cell number delivered to target tissue is observed which could weaken the effectiveness of the treatment. The other systemically administered MSC have been shown to preferentially accumulate at sites of tissue damage and inflammation [15,16]. As leukocytes, neutrophils or lymphocytes, MSC are highly motile cells and use chemotaxis to reach their target sites [17]. They have the ability to sense chemoattractant gradients (chemokines) through members of G Protein-Coupled Receptors (GPCR) that link to heterotrimeric G-proteins. The GPCR/G-protein sensing machinery acts as sensor and transduces the signal into the cell which reorganizes the actin cytoskeleton that drives cell movement [17]. The main chemoattractants described in literature are stromal cell derived factor 1 (SDF-1) [18,19], macrophage inflammatory protein 1a (MIP-1a) [20], monocyte chemoattractant protein 1 (MCP-1) [21], placenta growth factor (PlGF) [17] and platelet-derived growth factor (PDGF) [22]. Among these chemoattractants, the SDF-1 with its receptor CXCR4 expressed by MSC plays an important role in their recruitment into injured sites [23,24]. However, considering the mechanical environment of MSC during their migration, this chemotaxis is certainly influenced by blood circulation, like leukocytes. Indeed, it has been described that vascular adherence and transendothelial migration of leukocytes was largely modulate by forces deriving from blood flow shear [25–27]. Several studies have shown that MSC are able to detect and respond to shear stress due to blood and interstitial fluid flow through mechanotransduction pathways [28–30]. Depending on the vessel diameter and other physical parameters, MSC could be exposed to a wide variety of forces: from 0.1–0.6 Pa in veins to 1–7 Pa in arteries [31]. Studies have shown that shear stress can regulate MSC proliferation and osteoblastic, cardiomyocytes differentiation [32–35]. For migration, MSC response seems to depend on shear stress intensity since lower shear stress (0.2 Pa) induces migration while greater one (>2 Pa) hinders it [29]. From the mechanotransduction point of view, the fluid shear generates intracellular forces with actin and NMM-II proteins which allow modifying cell differentiation, cytokinesis, cortical tensions and migration. As very often, signaling pathway is a very complex process and implied several ways. Yuan et al. demonstrated that ERK1/2, c-jun and p38MAPK proteins are involved in the MSC migration under lower shear stress [29]. Moreover, mechanical strains could enhance the expression of CXCR4 on MSC [36,37]. The SDF-1/CXCR4 pathway known to be applied in the recruitment of MSC to the injured site seems to be augmented by mechanical stimulation. Yuan et al. also demonstrated, in another study, that lower shear stress upregulated the expression of CXCR4 and promoted MSC migration through an upregulated SDF-1/CXCR4 axis involving JNK/p38 MAPK pathway [30]. However, most of the functional CXCR4 chemokine receptors are lost upon in vitro culturing. Culturing of MSC for more than two passages is associated with the loss of CXCR4, and lack of chemotactic response to chemokines [23,38]. Thus, MSC mechanobiology knowledge must be taken into account to adapt the in vitro culture conditions and maintain the migration potential of MSC.

4. MSC transendothelial migration (TEM)

Classically, four distinct steps have been described in the process of homing: tethering and rolling mediated principally by the selectin group of adhesion molecules (step 1), GPCR/G-protein sensing machinery activation (step 2), firm adhesion by integrins (step 3) and TEM (step 4). MSC transmigrate actively into inflamed tissues via leukocyte-like mechanism but with some specificities. Nowadays, there is an absence of a clear mechanism for MSC homing. We know that MSC lack expression of various adhesion receptors that mediate homing step like the absence of L-selectin and a no functional E-selectin ligand (CD44) modifying step 1 interactions [39]. For step 3, if α 4 / β 1 (VLA-4, CD49d/CD29) and α 5 / β 1 (VLA-5, CD49e/CD29) integrins are expressed on MSC, these cells do not expressed the α L / β 2 (LFA-1) and α M / β 2 Mac-1) integrins. Moreover, it is certain that MSC must interact with the blood vessel wall, arrest their migration and mostly resist the generated shear stress to undergo the TEM. Segers et al. demonstrated that adhesion of MSC to endothelial cells depended on the flow velocity of the superfusing fluid and seems to be optimal between 1 and 2 dynes/cm² (0.1–0.2 Pa) which corresponds to the level of stress observed in veins and capillaries [40]. It would seem however that, under physiological shear stress, this adhesion would not pass through the rolling step due to the lack of expression of L-selectin [39,41]. So how MSC could decelerate and arrest to allow firm adhesion? Chamberlain et al. proposed a mechanism by which the leukocyte-like rolling is not an obligatory prerequisite. They think that the large size of MSC may reduce their velocity due to physical interactions with narrower capillaries leading to arrest and passive entrapment. This entrapment could be sufficient to enable chemokine presentation and GPCR/G-protein sensing machinery activation [41] which then enhance integrin binding (step 3). This firm adhesion of MSC to endothelial cells implied β 1 integrin particularly VLA-4 which can interact with VCAM endothelial ligand [42]. The question of interaction between MSC and endothelial cells need also to be elucidated because if forces are exerted on MSC, there are also forces exerted on endothelial cells and basement membrane below endothelial cells. In their review, Stroka et al. reported that stiffness of the basement membrane has a significant impact on both endothelial cells and leukocytes behavior [43]. Thus, it was shown that neutrophil TEM speed is biphasic with regard to substrate stiffness [44]. These cells respond very differently to substrate stiffness in the range 3-3kPa. Concerning MSC, it is well known that cell fate can be influenced by several mechanical properties, including elasticity, geometry and adhesion. So it could be possible that stiffness of the basement membrane modify MSC TEM. The results obtained from published studies on TEM show a real complexity to understand how MSC could interact with endothelial cells and the basement membrane. This is due to the difficult to obtain in vitro conditions that reflect physiological situation. In function of in vitro conditions used, authors obtained some results that differed from other studies which did not use the same experimental conditions. For example, under static conditions, Teo et al. observed that MSC diapedesis could occurred over a much longer timeframe than leukocyte one (respectively 30–120 and 3–6 min) with the absence of significant polarization, directed lateral migration and formation of cellular protrusions like lamellipodia or invadosomes [45]. In contrast, under physiologic shear flow, it was observed that MSC diapedesis implied slow lateral migration on the endothelium while maintaining similar durations before spreading on the subendothelial matrix [41]. Some described differences for MSC highlight the need to better understand static adhesion and adhesion under flow conditions as two distinct regulatory processes.

5. Impact of patient on MSC migration

6. Conclusion

An understanding of the mechanisms of MSC migration is of paramount importance for cell-based treatment. Modulation of the homing properties of MSC could allow for their efficient recruitment to sites of injury and inflammation. However, it appears that mechanical forces as fluid shear stress influences several steps involved in migration and homing of MSC. It is so necessary to take into account mechanical environment of MSC to reduce the gap between cell culture and cell physiological environment and to understand the behavior of these interest cells for both the in vitro cell culture and the in vivo situation. Moreover, biomechanical experiments are still required to test and characterize the shear-induced responses in MSC. These investigations could ultimately improve MSC-based cell therapy and regenerative medicine outcomes by adapting for example the dose of infused cells required.