Superparamagnetic nanoparticles for cell tracking and magnetic vectorization in ischemic stroke

Universidad de Santiago de Compostela Superparamagnetic nanoparticles for cell tracking and magnetic vectorization in ischemic stroke Nanopartículas superparamagnéticas para seguimiento celular y vectorización magnética en el ictus isquémico PhD Thesis Tesis doctoral Bárbara Argibay González 2014 El Prof. Dr. José Castillo Sánchez, Catedrático y Director del Departamento de Medicina de la Universidad de Santiago de Compostela, Jefe del departamento de Neurología del Hospital Clínico Universitario de Santiago de Compostela y Director Científico del Instituto de Investigación Sanitaria de Santiago de Compostela (IDIS), el Dr. Francisco Campos Pérez, Investigador Miguel Servet y colaborador docente del Departamento de Medicina de la Universidad de Santiago de Compostela y el Dr. Pablo Taboada Antelo, Profesor Titular del Departamento de Física de la Materia Condensada de la Universidad de Santiago de Compostela, CERTIFICAN: Que el presente trabajo titulado “Nanopartículas superparamagnéticas para seguimiento celular y vectorización magnética en el ictus isquémico” ha sido realizado bajo su dirección, por la licenciada en física Dña. Bárbara Argibay González, y se encuentra en condiciones de ser presentado y defendido como Tesis Doctoral ante el tribunal correspondiente en la Universidad de Santiago de Compostela. Santiago de Compostela, Diciembre de 2014 Prof. José Castillo Dr. Francisco Campos Dr. Pablo Taboada Dña. Bárbara Argibay Prof. Dr. José Castillo Sánchez, Professor of Neurology and Director of the Department of Medicine of the University of Santiago de Compostela, Chief of Neurology Department of the University Clinical Hospital of Santiago de Compostela and Scientific Director of Health Research Institute of Santiago de Compostela (IDIS), Dr. Francisco Campos Pérez (PhD and Miguel Servet researcher) and Dr. Pablo Taboada Antelo, Associate Professor of Condensed Matter Department of Physics of the University of Santiago de Compostela, CERTIFY: That the present research study entitled “Superparamagnetic nanoparticles for cell tracking and magnetic vectorization in ischemic stroke” has been carried out under their supervision by Bárbara Argibay González, graduated in physics, and the document fulfill all requisites for its defense under the corresponding committee proposed by the University of Santiago de Compostela. Santiago de Compostela, December 2014 Prof. José Castillo Dr. Francisco Campos Dr. Pablo Taboada Dña. Bárbara Argibay Abbreviations ACA: Anterior Cerebral Artery ADC: Apparent Diffusion Coefficients ATP: Adenosine Tri-Phosphate BA: Basilar Artery BBB: Blood Brain Barrier BM-MSCs: Bone Marrow Derived Cells C17.2: Multi-potent neural progenitor cell line C17.2 CBF: Cerebral Blood Flow CCA: Common Carotid Artery CFSE: Carboxyfluorescein Succinimidyl Ester CNS: Cerebral Nervous System Con-CC: Contralateral Common Carotid Artery CT: X-ray Computed Tomography DLS: Dynamic Light Scattering D-MNPs: Dextran-coated Superparamagnetic Nanoparticles DW: Distilled Water DWI: Diffusion Weighted Imaging EAATs: Excitatory Amino-Acid Transporters ECA: External Carotid Artery EGF: Epidermal Growth Factor EPO: Erythropoietin ESCs: Embryonic Stem cells FOV: Field of View FTIR: Fourier Transformed Infra Red GOT: Glutamate-Oxaloacetate Transaminase HUCBCs: Human Umbilical Cord Blood Cells i.a.: Intra-arterial i.v.: Intravenous ICA: Internal Carotid Artery ICP-OES or ICP: Inductive Coupled Plasma Optical Emission Spectroscopy IL-6: Interleukin-6 Ip-CC: Ipsilateral Common Carotid Artery Ip-EC: Ipsilateral External Carotid Artery iPSCs: Induced Pluripotent Stem Cells LDH: Lactate Dehydrogenase Assay MCA: Middle Cerebral Artery MGE: Multi Gradient Echo MMPs: Matrix Metalloproteinases MNPS: Magnetic Nanoparticles MRA: Magnetic Resonance Angiography MRI: Magnetic Resonance Imaging MSCs: Mesenchymal Stem Cells MSME: Multi Slice Multi Echo NePi: Neuropil nNOS: neuronal Nitric Oxide Sintase NSCs: Neural Stem Cells OA: Oleic Acid PBS: Phosphate Buffered Saline PCA: Posterior Cerebral Artery PCoA: Posterior Communicating Artery PEO: Poly(ethylene oxide) PFA: Paraformaldehyde PLL: Poly-L-Lysine P-MNPs: Pluronic F127-coated Superparamagnetic Nanoparticles PPA: Pterygopalatin Artery PPO: Poly(Prolpylene oxide) PWI: Perfusion Weighted Imaging Rb: Red Blood Cell RF: Radio Frequency Pulse rt-PA: recombinant tissue Plasminogen Activator SPIONs: Superparamagnetic Iron Oxide Nanoparticles SZV: Subventricular zone TA: Transfection Agents TE: Echo Time TEM: Transmission Electron Microscopy TGA: Thermal Gravimetric Analysis TIA: Transient Ischemic Attack TLDF: Transcranial Laser Doppler Flow tMCAO: transient Middle Cerebral Artery Occlusion TMNPs: Tetronic 908-coated Superparamagnetic Nanoparticles TR: Repetition Time VEGF: Vascular Endothelial Growth Factor VSM: Vibrating Sample Magnetometer WHO: World Health Organization XRD: X-ray Diffraction Index Section I .............................................................................................. 13 1. Introduction .............................................................................. 15 1.1. Stroke .................................................................................. 15 1.2. Stem cell therapy ............................................................. 33 1.3. Animal models of ischemic stroke ................................. 41 1.4. Non invasive imaging ...................................................... 48 1.5. Superparamagnetic iron oxide nanoparticles ............ 54 Section II ............................................................................................. 71 2. Hypothesis ................................................................................ 73 3. Objectives ................................................................................ 75 Section III ............................................................................................ 77 4. Synthesis of superparamagnetic nanoparticles coated with different polymers as contrast agents for MRI ............................. 79 4.1. Hypothesis .......................................................................... 79 4.2. Materials and methods ................................................... 80 4.3. Results ................................................................................. 85 4.4. Discussion ........................................................................... 96 4.5. Conclusions ..................................................................... 104 Section IV ......................................................................................... 105 5. Mesenchymal stem cells labeled with Dextran coated MNPs: Transfection agents evaluation ....................................... 109 5.1. Hypothesis ........................................................................ 109 5.2. Materials & Methods ...................................................... 109 5.3. Results ............................................................................... 114 5.4. Discussion ......................................................................... 125 5.5. Conclusions ..................................................................... 129 6. Dextran-, Pluronic F127- and Tetronic 908- coated superparamagnetic nanoparticles: In vitro validation ............. 131 6.1. Hypothesis ........................................................................ 131 6.2. Materials & Methods ...................................................... 132 6.3. Results ............................................................................... 137 6.4. Discussion ......................................................................... 149 6.5. Conclusions ..................................................................... 154 7. Long-term study of mesenchymal stem cells labeled with PLL- Dextran coated superparamagnetic nanoparticles ....... 155 7.1. Hypothesis ....................................................................... 155 7.2. Materials & methods ...................................................... 156 7.3. Results ............................................................................... 163 7.4. Discussion ......................................................................... 174 7.5. Conclusions ..................................................................... 177 Section V .......................................................................................... 179 8. Animal models of focal cerebral ischemia and their influence on the intra-arterial administration route for stem cell therapy ........................................................................................... 183 8.1. Hypothesis ....................................................................... 183 8.2. Materials & Methods ...................................................... 184 8.3. Results and discuss ion .................................................... 1 93 8.4. Conclusions ..................................................................... 237 9. Cell tracking of mesenchymal stem cells .......................... 239 9.1. Hypothesis ....................................................................... 239 9.2. Materials & Methods ...................................................... 240 9.3. Results & Discussion ........................................................ 247 9.4. Conclusions ..................................................................... 278 Section VI ......................................................................................... 281 10. Neuroreparation study in an animal model of ischemic stroke using mesenchymal stem cells labeled with Dextrancoated superparamagnetic nanoparticles .............................. 283 10.1. Hypothesis ....................................................................... 283 10.2. Materials & Methods ...................................................... 284 10.3. Results ............................................................................... 294 10.4. Discussion ......................................................................... 304 10.5. Conclusions ..................................................................... 307 18 1.1.2. Biochemistry of cerebral ischemia The acute obstruction of one of the large brain arteries induces an instantaneous reduction of blood flow in the corresponding irrigation area (focal ischemia). But that reduction of blood supply is not homogeneous in the affected area, and can change within minutes or hours, especially if blood supply is reinstated.[8] Two regions can be distinguished: The ischemic core is the portion of tissue closest to the affected blood vessel and where the ischemia becomes severe, and the so-called penumbra, where the reduction of blood flow is less severe, due to the blood supply carried out by collateral arteries of the non-ischemic neighbor tissue.[9] The impact of brain ischemia will depend on the level of the artery occlusion and duration of the reduction of blood flow, that is why time is a very important parameter in this disease. After the onset of brain ischemia, a sequence of molecular events are triggered in the short and the long term, initiated with an energetic failure in cells, related to the interruption of oxidative phosphorylation processes and the deficient production of adenosine tri-phosphate (ATP) (Figure 1). The cessation of transmembrane ionic gradients due to the failure of sodium-potassium-ATPase pumps, and other ATP- dependent ionic pumps, is the key step of the physiopathological mechanisms in stroke, especially of cell death in the ischemic core, when the vascular occlusion lasts for few minutes.[10] Neurons and glial cells suffer an extreme depolarization because of the entrance of sodium, chloride, calcium and water into the cytoplasm,[11] and in addition, potassium leaves the cell, inducing a sudden increment of its extracellular levels.[12] The energetic failure and the associated ionic changes, originate an increment 19 in glutamate, a hyperexcitability of N-methyl- D-aspartate glutamatergic receptors (NMDA), and of  -amino- 3- hydroxy-5-methyl-4-isoxazol propionic acid receptors (AMPA), which induces an even higher increase of intracellular calcium (Figure 2).[13-15] Figure 1: Sequence of main physiopathological events in cerebral ischemia. The increase of intracellular calcium does not exclusively depend on the activation of glutamate receptors, but also in the stimulation of calcium voltage-dependent channels. Hyperexcitability causes a depolarization phenomenon in the periphery of the infarct, which increases the energetic cost while the membrane tries to re-polarize itself.[9,16,17] Calcium increment, together with acidosis and peri-infarct depolarization, contributes to initiate the damage, and after it, inflammation and activation of apoptotic phenomena contribute to increment the lesion.[15,18] During ischemia, and particularly during reperfusion, free 20 radicals are generated. These are highly reactive species produced at the initial and late stages of brain ischemia, following different physiopathological mechanisms. In first place, oxygen reactive species are produced by the metabolism of arachidonic acid and the activity of neuronal Nitric Oxide (NO) synthase (nNOS). During intermediate stages, free oxygen radicals are provided by the infiltration of neutrophils in the ischemic area. At later stages, they are produced via the synthesis and activation of inducible NO synthase enzymes (iNOS) and cycloxigenase-2(COX-2).[19,20] Figure 2: Glutamate role on the stimulation of AMPA and NMDA receptors. Ischemic stroke triggers a series of complex molecular events, where the activation and the expression of genes 21 are included. Some of those events arise from the immediate reaction of neurons to damage,[21] and others are associated to cellular processes that determine the near fate of the affected neurons[22], or coordinate the repair mechanisms of tissues.[23,24] However, ischemic cellular death can take place in two different ways. The most common one is necrosis,[25] which is the result of the acute energetic failure, and it is characterized by morphology changes and, in the end, cellular lysis, which also triggers inflammatory processes.[26] On the other hand, apoptotic or programmed cell death can be observed when energy-dependent intracellular mechanisms are activated, leading to cell degradation.[18,27] 1.1.3. The peri-infarct region The introduction of the concept of ischemic penumbra by Astrup[10], and its subsequent development, has been crucial for the consideration of ischemic stroke from a preventable catastrophe to a treatable disease. The time is brain aphorism, and the consideration that stroke is a neurological emergency, as a result of the ischemic penumbra concept, has allowed patients to profit from fast attention by trained medical staff and specialized stroke units. The penumbra is classically defined as the hypoperfused tissue surrounding the ischemic core, where blood flow is too low to maintain electric activity but sufficient to preserve ion channels. However, this area is subjected to waves of deleterious metabolic processes propagated from the core to the neighboring tissue, including excitotoxicity, spreading depression, oxidative stress, and inflammatory response, which lead to the 22 expansion of the ischemic core and the subsequent worsening clinical outcome. The most relevant definition of ischemic penumbra for clinical practice is based on neuroimaging techniques. It is widely accepted that brain tissue with reduced blood perfusion, as seen in MR perfusion-weighted imaging (PWI) but not included into the lesion core, as seen in MR diffusion weighted imaging (DWI), indicates potentially salvageable tissue. Thus, the combination of PWI and DWI images has led to the PWI/DWI mismatch concept.[28] Although it is generally accepted that hyperintense signal on DWI represents the lesion core, quantitative measurements of apparent diffusion coefficients (ADC) reveal that during ischemia, ADC values decline before energy metabolism fails, indicating that the increase in DWI signal intensity is not restricted to the infarct core[29]. Therefore, the discrimination between penumbra and infarct core by DWI is not always clear, because DWI signal increases in both sites. In addition, PWI-detectable flow decrease is pathophysiologically relevant only when interferes with the adequate oxygen supply to the tissue. Thus, the PWI/DWI mismatch includes not only the penumbra periphery, but also surrounding intact tissue.[28] Nowadays, neuroscientists are able to delineate the ischemic penumbra using alternatives to this hemodynamic concept. Among others, tissue hypoxia can be determined by MR spectroscopy, anoxic depolarization depicted by manganese enhanced MRI, tissue acidosis shown by pH-weighted MRI, and others. However, none of these methodologies is able to provide clear-cut threshold values to differentiate among ischemic core, penumbra, and surrounding intact brain tissue by themselves, and a multimodality approach is generally recommended.[28,29] Most of these MRI-based methods are unfortunately restricted to experimental models and 23 state-of-the-art brain imaging systems, and must be fully validated for the human brain before they can be translated into clinical practice.[30] Although the definition of the penumbra by imaging techniques is currently of great value, in clinical terms an alternative definition of “tissue-at-risk”, that is generally called peri-infarct region, may be more suitable for the development of new therapeutic approaches in stroke. The peri-infarct region is defined as a no men’s land between the infarct core, with a front of toxic mediators of damage spreading out from it, and the healthy tissue, from where tissue’s healing and repair mediators are trying to access to the affected but still viable tissue (Figure 3). It is clear that the peri-infarct definition and spatiotemporal dynamics do not necessarily match the specifications of the ischemic penumbra. Figure 3: Peri-infarct region defined as the battlefield between mediators of damage, from the infarct core, and mediators of healing and repair from normal tissue. (Adapted from [31]) 24 1.1.4. Therapeutic approaches for the treatment of stroke 1.1.4.1. Thrombolysis and neuro-interventionism Pharmacological or mechanical (thrombectomy) thrombolysis are the strategies that report higher benefits for the patient, in terms of neurological outcome. The most common thrombolytic agent is the recombinant tissue plasminogen activator or rt-PA. Both therapies have pushed for the creation of stroke units inside hospitals, which have improved the management of stroke patients. Nevertheless only 3-7% of stroke patients are currently treated by these procedures. Such reduced numbers may be due to different factors, including the narrow therapeutic window and the high risks of hemorrhage transformation. Current neuroprotective strategies are required to work at both stages, by widening the therapeutic window and by reducing the associated risk factors.[32-36] The therapeutic window associated with intravenous thrombolytic treatment is 4.5 hours. The extension of this window may be possible by selecting candidate patients with a large penumbral area (the area of the brain susceptible to damage unless otherwise protected within the first 24-48 hours).[37] In addition, reperfusion is associated with a higher rate of hemorrhagic events and the maintenance of the bloodbrain barrier integrity (i.e. with metalloproteinase inhibitors), fluid homeostasis (by regulating aquaporin and inhibiting endothellin) and disabling free radicals after a reperfusion treatment, makes neuroprotective strategies during the acute phase of a stroke difficult to be viable.[38] An early treatment with anti-excitatory drugs or induction of hypothermia could also help to maintain the penumbra for 25 extended periods of time, hence extending the therapeutic window for reperfusion.[39] 1.1.4.2. Neuroprotection Neuroprotection is a term that conglomerates a variety of strategies focused in reducing cell death after an ischemic event, without affecting tissue reperfusion. So far, a plethora of compounds have been proposed to block the pathway leading to ischemia-induced cell death at different steps. Most of these compounds have shown positive effects in experimental studies, although none of them have shown a clear beneficial effect in clinical studies.[40,41] Neuroprotective drugs can be classified into different groups, regarding their action mechanism. a. Calcium blockers Calcium plays an important role in stroke pathophysiology. The blockage of calcium channels stops neuronal calcium intake, hence reducing cell death. Nimodipine is an example of this family of compounds. With more than 200 published experimental studies, only half of them proved the efficacy of the drug. None of the members of this family of compounds have demonstrated a clear neuroprotective activity on clinical studies.[42,43] Among the reasons for this fact, overall the studies that generated positive results, animals were mostly treated within the first 15 minutes postischemia, which is not translatable to the clinical settings. b. Glutamate antagonists Glutamate is the main excitatory amino acid in the central nervous system. After an ischemic event, an excitotoxic response is generated leading to cell death. Glutamate activates NMDA and AMPA receptors. NMDA receptors 26 mediate cellular calcium and sodium intake. Several noncompetitive blocking drugs against NMDA receptors have been synthesized, such as MK-801 (dizocilpine), which was used as a neuroprotective compound and rendered positive results in animal models of stroke. Nevertheless, the subsequent clinical trial had to be stopped due to adverse effects.[44] NMDA competitive antagonist drugs have been also synthesized and used in animal models of stroke, like CGS 19755 (Selfotel) and GV15526 (Gavestinel). Although both compounds produced encouraging results in experimental studies, but they showed no effect in clinical trials.[45,46] Natural NMDA antagonists, such as magnesium, have also been used as neuroprotective compounds. Magnesium sulphate has showed a brain infarct reduction up to 60% in animal models, treated even 6 hours after ischemia onset. Once again, these results showed no effect in clinical trials.[47] As for NMDA, AMPA receptors have also been targeted for neuroprotective strategies. AMPA antagonists like ZK200755 have shown moderate neuroprotective effects in animal models. Results of its clinical trials have never been published.[38,42,48] c. Plasmatic glutamate grabbers Under normal physiologic conditions, glutamate present at the synaptic junction is captured by neurons and astrocytes through specific excitatory amino-acid transporters (EAATs). Nonetheless, under ischemic conditions these transporters do not function properly, and the massive release of this molecule lead to an accumulation of glutamate in the extracellular space. 27 EAATs are also present at the luminal side of endothelial cells in brain capillaries, and facilitate the passing of glutamate from the brain to blood. In this sense, it has been recently published that the increase of glutamate gradients between brain and blood facilitates the dumping of glutamate from the extracellular space to circulating blood, offering in this way an alternative strategy to protect brain cells during ischemia. Those strategies are based on metabolizing plasmatic glutamate by the action of the resident enzyme glutamate-oxaloacetate transaminase (GOT). This neuroprotective strategy has been successfully tested in animal models by our research group, leading to a significant reduction on brain infarct volume in animal models of stroke. In parallel, an association between elevated levels of GOT on blood and good outcome after stroke has been found in two clinical retrospective studies, pointing to the potential translationality of this neuroprotective strategy to the clinical setting.[49-52] d. Antioxidants Oxidative stress is another mechanism implicated on cell death after an ischemic event. Antioxidants could therefore play a role as neuroprotective drugs. The most successful antioxidant tested has been NXY-059. This drug reduced brain infarct by 66% in animal models, when injected 5 hours after occlusion. The first clinical trial generated positive results, improving patient’s functional outcome. However a second clinical trial showed negative results. [53-55] e. Phospholipid precursors: citicoline Citicoline or CDP-choline is a precursor on the synthesis of phosphate-choline, which is integrated in the membrane of neurons. It has been shown that citicoline inhibits 34 ischemic stroke recovery using stem cells are still poorly understood.[78] One of the potential mechanisms of stem cell therapy is cell replacement. Early studies have reported that after stem cell transplantation the number of grafted cells differentiated into mature neurons formed can vary. The origin of this variation lies in the source of the stem cell and also in the injection site. In this way, in recent preclinical studies, 34-60% of the injected cells were Neural Stem Cells (NSCs),[79] 30% Embryonic Stem Cells (ESCs)[80] or 2-20% for Mesenchymal Stem Cells (MSCs).[81,82] But sometimes, grafted cells remain undifferentiated close to the injury site, releasing growth and trophic factors [83,84] or even without entering the Cerebral Nervous System (CNS) inducing also benefits after stroke.[85] Regardless of this effect, this systemic cell administration can provoke that administered cells stay also in other organs, like spleen or liver, interacting with other cells from the immune system and leading into modulatory functions,[86] reducing at the same time, the post-ischemic inflammatory damage in stroke. [87-89] Nevertheless, apart from cell replacement, enhanced trophic support and immunomodulatory response, there are endogenous brain repair processes that stem cell can stimulate, such as angiogenesis, neurogenesis, synaptogenesis and white matter remodeling [62,90-93] inducing neurorestorative effects after ischemia. In contrast to other pathologies, in stroke there is affected more than one brain cell population, involving also vascular cells in the ischemic region. This heterogeneity means that not only functional neuronal connections have to be reestablished; the vascular system has to be restored as well. In consequence, there is not an unique type of stem cells for stroke therapy, many of them have been tested 35 demonstrating beneficial effects: Embryonic Stem Cells (ESCs),[94-96] Neural Stem Cells (NSCs),[91,97,98] Induced Pluripotent Stem Cells (iPSCs),[99,100] Human Umbilical Cord Blood Cells (HUCBCs)[101,102] and Mesenchymal Stem Cells (MSCs). [82,103,104] Recent studies with iPSCs have shown powerful new opportunities for modeling human diseases offering hope for personalized regenerative cell therapies. However, the field is racing ahead, and some researchers are pausing to evaluate whether induced pluripotent stem cells are indeed the true equivalents of embryonic stem cells and whether subtle differences between these cells might affect their research applications and therapeutic potential.[105,106] But, there are several parameters to take into account to elucidate the effective time point for stem cell therapy in preclinical or even clinical studies. An important factor to be considered is the use of autologous versus allogenic stem cells. The term autologous is referred to those from the same individual who is going to receive the treatment whereas allogenic cells are related to a donor. In spite of this, the less lineage specific the cells are, the less chances to have an immune response,[107] in fact, for example MSCs which can be harvested from a variety of mesenchymal tissues have different immune response depending on their origin.[108] Additionally to the immune response, several cases of tumorigenicity have been reported after stem cells therapies from different stem cell types. [109,110] Very few preclinical studies have compared different routes of administration, and the choice among them depends on the organ target. First studies for cerebrovascular diseases explored the intraparechymal route with the main goal of determine the feasibility of replacing lost neurons and 36 rebuilding neural circuits,[111] meanwhile later studies showed that stem cells were not only able to survive after injection, as well as proliferate, differentiate and even migrate to the site of the injury.[112,113] However, intraparenchymal cell delivery normally is not the first choice because the injection is focused in one region and to reach that point it is necessary to damage brain tissue. On the other hand, vascular routes, intravenous (i.v.) and intra-arterial (i.a.), solve these pitfalls and they are becoming the most common cell delivery method. Tail vein, femoral and jugular veins are the most common ones for stem cell administration. Several groups have demonstrated that i.v. administration can reduce infarct size or induce functional recovery [82,101] when administrated 24h after stroke. Later studies also studied the biodistribution after i.v. injection and found out that depend on the cell type, between 1-5% of the injected cells reached arterial circulation.[114] One of the most interesting i.v. applications was for NSCs delivery. It’s known that this kind of stem cells has the ability to differentiate into neurons, astrocytes and oligodendrocytes. At the early beginning of stem cell therapies their repairing mechanisms were assumed to be related to rebuild the neural structure, but in recent years it has been observed that the injured brain environment is not optimal for cell integration, so the recovery works through growth factor mediation or anti-immflamatory processes. [115] Nevertheless, it seems that if the brain is the target, it would be desirable to reach it with the most of the injected cells, and to achieve it the best delivery route is the i.a. administration. The first study showed that 21% of the injected Bone Marrow Stem Cells (BMSCs) were observed in the affected hemisphere, and this was also related with functional recovery in behavioral tests but not as an infarct volume reduction in the treated group. However, 37 administration route efficiency is also conditioned by the safety of the procedure and attending to this, i.a. administration has shown higher mortality than i.v. delivery. In a recent study, 41% of the animals died after an i.a. injection compared to 8% for i.v. delivery.[116] That’s why this delivery route has been more explored recently and many studies have developed different injection strategies to minimize the risks.[117-120] All these promising preclinical results progressed into early clinical trials of stem cell therapy in stroke. The effectiveness of different types of stem cells and delivery routes is also reflected in clinical studies which are exploring different combinations. So far, there is four published clinical trials of intracerebral cell therapy for stroke and one that is still ongoing,[121-123] one of intravenously administration [115,124-126] and three following intra-arterial delivery. [127- 129] The main goals of these clinical trials were to demonstrate the safety and feasibility of stem cell therapy using different types of stem cells and routes of administration. Published results have reported that there were not enough patients enrolled in these studies to show benefits in terms of functional recovery after the treatment, but all of them demonstrated that stem cells are a promising, feasible and safe therapy for stroke. However, the optimal conditions for applying stem cell therapies are still under discussion, fundamental questions related to cell type, characterization and dosage, therapeutic timing versus toxicity, or the relationship between biodistribution, fate and outcome are still on the bench, and preclinical studies should answer it first. 38 1.2.1. Mesenchymal Stem Cells The term mesenchyme is related to embryonic connective tissue, derived from mesoderm, that can be differentiated into hematopoietic and connective tissue. Mesenchymal Stem Cells (MSCs) can be derived from almost all tissues in the body (Figure 6): bone marrow, placenta, muscle, skin, adipose tissue, umbilical cord, etc.[130] The bone marrow derived cells (BM-MSCs) is the most studied type of MSCs. [131] They express a number of markers, none of which unfortunately are specific to MSCs, being CD73+, CD90+, CD105+, CD11b-, CD14-, CD19-, CD79a- , CD34-, CD45- and HLA-DR. It is generally agreed that rat MSCs do not express hematopoietic markers, as CD45, however cultured MSCs are uniformly and strongly positive for CD105, CD90 and CD73, regardless their passage or time in culture. MSCs can be differentiated into several cell types in vitro and can be expanded in vitro. These two properties make the mesenchymal stem cells attractive as a powerful source for tissue repair.[132,133] Moreover, MSCs have the potential to differentiate into adypocites, chondroblasts and osteoblasts[134] and their ability to be differentiated into myocytes, neuronal and glial cells has been already reported.[135-137] 39 Figure 6: MSCs differentiation into different cell types. MSCs are one of the best candidates for regenerative therapies, also in stroke, not only due to their multipotientiality, but mainly because their ability of releasing growth factors and their immunomodulatory abilities.[131] Several mechanisms of action of these cells in the ischemic brain have been described, mechanisms like transdifferentiation into cells of neural lineage[138-141] induction of neurogenesis,[142-144] angiogenesis,[142-144] synaptogenesis,[145] activation of endogenous restorative processes by producing cytokines and trophic factors,[142,146-148] regulating the Cerebral Blood Flow (CBF) and the Blood Brain Barrier (BBB),[146] and other neuroprotective mechanisms like the reduction of apoptosis, immflamation, demyelination and the increase of astrocyte survival.[101,142,143,149] Many preclinical studies reported functional recovery and infarct volume reduction after ischemic stroke, [131] which facilitates the leap into clinical trials. 40 1.2.2. Strengthen of endogenous stem cells and stroke recovery Several studies in rodents and also in humans have reported that after a stroke event there is an endogenous stem cells migration towards the site of damage to contribute to the functional recovery.[150,151] These neural stem cells and their progeny activation and migration occur till the cells are integrated forming new functional neurons.[152,153] However, this endogenous contribution is not enough to overcome the consequences after stroke. Several groups have demonstrated that direct administration of growth factors can enhance these endogenous processes like migration of NPCs, neurogenesis and angiogenesis. Factors such as vascular Endothelial Growth Factor (VEGF),[154-156] Epidermal Growth Factor (EGF),[157] Fibroblast Growth Factor (FGF) and Erythropoietin (EPO)[158] have been proved in stroke models with functional benefits. Moreover, the stimulation of the NPCs without affecting other cell types is being investigated. There are some approaches for this purpose by focalizing different treatments as poly(ethylene glycol) nano/microparticles carrying biomolecules[159] direct brain injection of functionalized hydrogel[158] or intraventricular infusion of biomolecules.[160] Nevertheless, there is another way of growth factors delivery: stem cells. If the matter is target the proximity of the injury, as it has been specified at the beginning of this section, intra-arterial route of administration is the most efficient in terms of cell engraftment. But this binding can be also enhanced either by selecting a specific population of cells that endogenously express adhesion molecule or by modifying them to induce their expression.[161] However, it 41 is still necessary to demonstrate if those cells which have been bound to the endothelium, are able to migrate and target stroke lesions. 1.3. Animal models of ischemic stroke Stroke is a focal neurologic deficit caused by an alteration in the cerebral blood circulation. Animal models for cerebral ischemia have been used for over 150 years trying to mimic the human pathology,[162] improving our understanding of the physiopathology of this disease. Numerous preclinical studies, mostly with neuroprotective drugs, have shown promising results for the treatment of stroke, however not all positive experimental outcomes were transferred to patients, in terms of a phase III clinical study so far.[163,164] However, the success of stroke studies in animals depends on the choice of the experimental model species. An inadequate selection of the specie and/or the experimental model may lead to limitations that compromise results and analysis.[165,166] Numerous animal species have been used to study stroke, although the most common are rodents, due to the ease for inducing the cerebral ischemia, ease to monitor physiologic conditions and the reduced costs to obtain and maintain the animals for long periods of time.[163,166] We may not forget that in animal models of cerebral ischemia, most of variables are controlled in laboratory experiments and these conditions are quite different compared to a human going through a ischemic stroke process.[165] 42 Among rodents, the arterial and nerve supply of the rat is essentially similar to that of humans,[167] so rat is an appropriate animal for preclinical studies of brain ischemia. Firstly, and for a better understanding of cerebral ischemic models, it is necessary to have an anatomic knowledge about cerebral circulation in the rat. Basically, the head blood circulation depends on two systems: the carotid and the vertebrobasilar systems.[168] o Carotid system. Right and left common carotid arteries (CCA) supply the head and the neck. Left arises from the arch of the aorta and right common artery is a continuation of the innominate artery. Each common carotid yields two branches, the external and internal carotid, both follow the same general direction to the head. External carotid (ECA) leads to occipital, superior thyroid, ascending pharyngeal, lingual and ascending palatine, meanwhile internal carotid artery (ICA) continues to reach the base of the skull. On its way to the brain, it branches into pterygopalatine (PPA) artery, which corresponds to a portion of the internal maxillary branch of the external carotid artery in humans. The cerebral portion of the internal carotid artery branches into the posterior communicating artery (PCoA), hypothalamic artery, anterior choroid artery, middle cerebral artery (MCA) and anterior cerebral artery (ACA). The main branch of posterior communicating artery is the posterior cerebral artery (PCA), which supplies the surface of the brain hemisphere, including the medial and lateral surfaces of the occipital lobe. o Vertebrobasilar system: vertebral arteries arise from the anterior surface of the subclavian artery. When 43 entering the skull, both vertebral arteries join to form the basilar artery (BA). This artery is anastomosed to posterior communicating arteries from the carotid system. Arteries from carotid and vertebrobasilar systems form a circle of communicating arteries, which protects the brain when the blood flow circulation is interrupted. This arterial circle, present in humans and also in rats, is called circle of Willis. The circle of Willis in rats is a buttonhole-like structure, formed by the anterior cerebral and posterior communicating arteries. Normally, another extra artery, the anterior communicating artery (ACA), bridges anterior cerebral arteries from both hemispheres closing the circle, however is quite common that the circle is not complete. Probably one of the most interesting things about the circle of Willis is that the blood flow circulation is overlapped, i.e., if one of the main arteries is occluded, the distal smaller arteries that it supplies can receive blood from other arteries, which is called collateral circulation.[167,169] (Figure 7) Depending on the region affected by the diminution of cerebral blood flow (CBF), cerebral ischemic experimental models can be classified in global, focal or multifocal ischemia,[166] and attending to the duration of the ischemic insult, ischemia can be classified in transitory, if the artery occlusion is followed by reperfusion, or permanent ischemia, without artery reperfusion.[163] In this work we will focus on the focal cerebral ischemic model, which normally results in an ischemia in the cortex and striatum from occlusion or the middle cerebral artery, and it mimics the most common site of human stroke. In animal models of focal cerebral ischemia, the degree and distribution of CBF, and therefore infarction, depend on the duration and the 50 electrical signal that can be detected and further processed to generate an image. Figure 8: T 2 weighted magnetic resonance image. There are two independent and simultaneous physical processes responsible for the relaxation of excited proton spins, but before, it would be necessary to define what spin magnetization vector is. To describe MRI from a macroscopic point of view, we can suppose that we allocate a tissue into an external magnetic field (B0) and that such tissue (Figure 9a.) can be divided in different parts, depending on how these parts are affected by the magnetic field (Figure 9b.), or in other words, in groups of spins that are experiencing the same magnetic field strength. Each of the delineated groups of spins can be represented by a global magnetization vector (Figure 9c.), which, at the same time contribute to form the net magnetization of the whole tissue (Figure 9d.). 51 Figure 9: Description of spin magnetization vector. Under B0 condition, protons have an equilibrium longitudinal magnetization vector, parallel to this magnetic field (we consider B0 and the net magnetization aligned along the Z axis). In this equilibrium, transversal components of the magnetization (perpendicular to B0) are not detectable because nuclear spins are processing out-of-phase and their transversal components cancel each other. However, after an excitation of the spins by a RF excitation, the system is out of equilibrium in terms of longitudinal and transversal magnetization, and in this state, the transversal component of this vector can be detected. As we have already mentioned, there are two relaxation processes that bring the system back to equilibrium, each of them described by time constants, the Spin-Lattice relaxation or longitudinal relaxation (T1) and the Spin-spin or transversal relaxation (T2). a. Spin-Lattice Relaxation T1: This phenomenon is associated with the recovery of the thermodynamic equilibrium magnetization along B0 and therefore with transitions between the Zeeman energy levels. To get back to equilibrium there is needed some energy dissipation into the surrounding tissue (lattice), and the emission is induced by fluctuations of the magnetic field, as a 52 result of the magnetic dipoles in the neighborhood variations (Brownian motion). As a consequence, T1 is dependent on the strength of the static field and not only on the molecular mobility of the observed nucleus, on the surrounding dipoles as well. Mathematically, spin-lattice relaxation can be expressed as:     󰇩1e 󰇡   󰇢 󰇪 C is a constant dependent on the flip angle of the RF pulse. b. Spin-spin relaxation T2 and effective spin-spin relaxation T2*: Transversal magnetization is related to the loss of phase coherence in the nuclear precession after a RF excitation. If the net magnetization is placed in “xy” plane, it will rotate following one defined frequency called “Larmor frequency”. But, apart from rotating, the net magnetization will start dephasing and each spin packages would have their own Larmor frequency. In this way, the time constant which describes the return to equilibrium of the transverse magnetization (Mxy) is called the spinspin relaxation time (T2).     󰇩 󰇡   󰇢 󰇪 Two factors also contribute to this decay, molecular interactions (pure T2 effect) and variations in B0 (inhomogeneities). The combined time constant is called effective spin-spin relaxation and represented by T2*. 53 1  ∗ 1   1   Here, T1 and T2 have been described separately but both processes occur simultaneously. The full description of the magnetization vector behavior under different conditions is implemented in Bloch equations. MR images are acquired using a pulse sequence (formed for radiofrequency pulses (RF) and gradient pulses). Depending on the pathologic condition or the composition of the tissue of interest, it will be necessary the use of one or another sequence to distinguish between what is called T1 and T2 weighted images, the two most common modalities of MRI in clinical practice. The differences between sequences rely on the choice of the acquisition parameters, mainly the repetition time (TR) and the echo time (TE), but also the contribution of the intrinsic tissue values of T1 and T2. However, sometimes for a more specific tissue detection or a change in an earlier stage of the disease, the combination of intrinsic relaxation times and sequence parameters is not enough for being visualized in MRI. It is needed an extra tool: MRI contrast agents. 1.4.2. Contrast agents for magnetic resonance imaging. MRI contrast agents are mainly used for two purposes, enhance the natural contrast and /or obtain dynamic information about biological processes. There are several options of MRI contrast agents, and parameters as pathology under study, target organ, therapeutic or diagnosis options, have to make the choice among them. 54 Particularly, for stem cell therapy, it would be ideal to monitor where and how these cell distribute. For different pathologies, also in ischemic stroke, the selected contrast agent to label stem cells has to achieve several goals: - Specificity of the type of cell - Ability to be detected at low concentrations - Reasonable clearing period - Stability - Reduced toxicity The most used contrast agent for cell tracking purposes verifying all previous conditions are magnetic nanoparticles (MNPs). 1.5. Superparamagnetic iron oxide nanoparticles Cells are too small to be detectable using MRI, however the use of contrast agents to label them leads us to an in vivo real-time image of those cells. Superparamagnetic iron oxide nanoparticles are an excellent tool for stem cell tracking, not only for the high quality of the MR image, for the low toxicity and the ease for tagging as well. The term nanoparticle is normally referred to particles with a size in the nano range, which corresponds to ca. 1-100nm. This reduced size gives them extraordinary properties compared to their counterparts, bulk materials. In this way, magnetic nanoparticles (MNPS) are nanoparticles made of ferro /ferrimagnetic materials, which exhibit superparamagnetic properties below certain dimensions 55 and above certain temperatures. To explain these properties several preliminary concepts will be introduced. 1.5.1. Magnetic materials All materials are magnetic, but their response under an applied external magnetic field depends on their atomic structure and the temperature conditions. When a material is placed under a magnetic field of strength H, the individual atomic moments contribute to the overall magnetic response (B), which mathematically is expressed   󰇛󰇜 [Eq.1] Where is the permeability of free space, and M is the magnetization per unit of volume. It is possible to classify the materials according to their magnetic response under a magnetic field, or what is the same, to the volumetric magnetic susceptibility (). Formally, this magnitude is defined as the induced magnetization in a material by a magnetic field H.  [Eq.2] In this way, depending on the magnetic susceptibility it is possible to distinguish between: a. Diamagnetism: it is a property common to all materials, but normally is very weak. In diamagnetic materials there are not unpaired electrons in their atoms and consequently, there is not net magnetic moment. When these types of materials are placed under a magnetic field, the magnetization produced is negative, so < 0. As it is shown in Figure 10a, for zero field, the magnetization is also null. 56 b. Paramagnetism: The net magnetization of these materials is different from zero due to the unpaired electrons of their atoms. However, due to the nonmagnetic interaction between the individual magnetic moments, the global process is randomized, and if the external field is removed, the magnetization is also zero. Under the influence of applied H, there is a partial alignment of the atomic magnetic moments in the direction of the field, resulting in a positive magnetization and as a consequence, the susceptibility becomes positive as well. (Figure 10b) Furthermore, this alignment is opposed to the randomizing effect of temperature, resulting in a correlation between temperature and magnetic susceptibility (Curie Law). Figure 10: Magnetic responses for a diamagnetic material (a.), paramagnetic material (b.) and ferromagnetic material (c.)[186] Hence, both diamagnetic and paramagnetic materials, exhibit the same behavior; no magnetization when the applied field is zero. Nevertheless, some other materials remain magnetic even without magnetic field presence. Those are ferromagnets, ferrimagnets and antiferromagnets. a. Ferromagnets: these materials are characterized by the parallel alignment (Figure 11a.) of their atomic 57 moments, which results in a large net magnetization even when the external magnetic field is zero (Figure 10b). Two characterizing aspects of ferromagnetic materials are their spontaneous magnetization and the existence of a magnetic ordering temperature, called Curie temperature (Tc). b. Ferrimagnets: In these materials the opposing atomic moments are unequal so that there is a remaining magnetization (see Figure 11c.). This phenomenon occurs when there are different ions in the material (such as Fe2+ and Fe3+). In this type of materials, the spontaneous magnetization is maintained below Tc and they behave as paramagnetic materials above it. c. Antiferromagnets: As in ferromagnetic materials, the atomic moments are opposed and equal, so the net moment is zero (see Figure 11b.). In this case, the critical temperature is TN (Néel temperature): above TN the magnetic susceptibility decreases with higher temperature and below it the susceptibility increases with temperature. The behaviors previously described are also strongly dependent on the size of the material, so modifying the size it is possible to tune the magnetic response at a certain temperature. 58 Figure 11: Alignment of magnetic dipoles for different types of magnetism. (a.) Ferromagnetism. (b.) Antiferromagnetism. (c.) Ferrimagnetism. As presented before, a paramagnetic material is characterized by randomly oriented magnetic dipoles, which under an external magnetic field are aligned along the direction of the field. But when the field is switched off, all internal dipoles randomize again and no extra energy is needed to demagnetize the material, it goes back to equilibrium spontaneously, i.e. this type of material has no coercivity and no remanence. However, ferromagnetic materials present the individual magnetic spins align parallel one to the other exhibiting a collective response even in the absence of the external applied magnetic field. In this absence of an applied magnetic field, a ferri or ferromagnetic material is divided into macroscopic magnetic domains. There is a specific direction of magnetization within each domain, but different from one domain to the other. When a magnetic field is applied to a multi-domain material, the domain walls move in response to the applied magnetic field, increasing the net magnetization of the crystal. As the applied field is increased, the magnetization increases until saturation is reached, when the crystal consists of a single magnetic domain with its magnetization aligned to the field (Figure 12). In this way, the saturation magnetization (Ms) is a 59 property of the material and can be related to the magnetic distribution and ordering. If the applied magnetic field is now reduced to zero, the magnetization will decrease to MR, which is the remnant magnetization, and in order to reduce the magnetization to zero, a reverse field has to be applied. The value of this reverse field is called coercivity Hc. Increasing the reverse field saturates the sample in the reverse direction, and so on, giving a hysteresis loop, which describes the relationship between the applied field and the magnetization.[187] Figure 12: The behavior of the domain distribution in a multidomain grain during various stages of magnetization. M s is the saturation magnetization where the grain consists of a single domain magnetized in the direction of the applied magnetic field. (Adapted from [187]) The difference between a magnet and the material to which magnets are attracted is the size of the magnetic domains. Thus, the shape of the M-H curves is partially determined by the size of the material, in this case, the size of the particle. In large particles (>m) there is a multidomain ground state, which causes a narrow hysteresis loop due to the small energy needed to make the domain 66 addition is in general performed by using an amphiphilic molecule, which hydrophobically interacts with the nonpolar surfactant. On the other hand, ligand exchange involves total replacement of the non-polar ligands by a polar surfactant.[197] There are 3 different approaches to stabilize magnetic nanoparticles by adding a ligand: monomeric stabilizers (carboxylates, phosphates or sulfates), inorganic materials (silica, gold or gadolinium, providing not only stability but also helping in binding various biological ligands to the surface as well) and polymeric stabilizers. In this approach, it is possible to distinguish between in situ coatings, where nanoparticles are coated during the synthesis, and postsynthesis coatings. In the literature, the most common coatings are dextran, carboymethylated dextran, carboxydextran, arabinogalactan, glycosaminoglycan, starch, sulfonated styrene-divinylbencene, PolyEthylene Glycol (PEG), PolyVinyl Alcohol (PVA), poloxamers and polyoxamines,[196] although the selection of the coating will be conditioned by the biological application of the coated nanoparticle. 1.5.3.3. Superparamagnetic nanoparticles for cell labeling and cell tracking MRI contrast agents for in vivo cell tracking are being widely investigated. Transplanted cells not only have the potential to replace damaged cells, but also to produce growth and trophic factors, or stimulate the release of such factors from host brain cells, enhancing endogenous brain repair processes. Many issues for stem cell therapy are still in the bench like therapeutic window, cell type, delivery route and in vivo monitoring of their migration pattern. The possibility to track the transplanted cells could help to 67 clarify all these questions and improve stem cell therapies for different pathologies.[198] MRI contrast agents for cell tracking require nanoparticles with special characteristics as high magnetization values and small and narrow size distribution. Thermal decomposition and chemical coprecipitation methods are the most extended procedures. Coprecipitation technique is probably the simplest and most efficient chemical route to obtain magnetic nanoparticles. As it was previously mentioned, the main advantage is that a large amount of particles can be synthesized easily and economically.[199] But for biological applications not only the magnetic properties are important, a specific surface coating of the magnetic particles is also needed, which has to be biocompatible.[196] Among all the coating possibilities, dextran is one of the most widely polymers used as magnetic nanoparticle coating due to its biocompatibility. It is a polysaccharide polymer composed exclusively of D-glucopyranosyl units with varying degrees of chain length and branching. In 1982, Molday and Mackenzie were the first who reported the formation of magnetite in presence of dextran 40000.[200] The synthesis of dextran coated nanoparticles is an in situ procedure, and the effect of reducing the terminal glucose of dextran upon the formation and stability of dextran-coated superparamagnetic nanoparticles has demonstrated to be significant for particle size, coating stability and magnetic properties.[199] For cell labeling applications, dextran coated nanoparticles normally need to be combined with transfection agents or the use of electroporation techniques to increase the tagging efficiency.[201-204] 68 Other methods to synthesize coated magnetic nanoparticles need more than one step. Jain et al. [205] developed a novel oleic acid (OA)-Pluronic-stabilized iron oxide magnetic nanoparticle formulation, and characterized it as a drug carrier system for anticancer agents. The hydrophobic drugs would partition into the OA shell surrounding the iron oxide nanoparticles, and the polymer (Pluronic) would anchor at the interface of the OA shell to confer an aqueous dispersity to the formulation. [205] Pluronic (Poloxamer) and Tetronic (Poloxamines) are a family of amphiphilic nonionic block polymers. More in detail, Pluronic block polymer consists of a hydrophobic poly(ProPylene Oxide) (PPO) chain flanked on each side with hydrophilic Poly(Ethylene Oxide) (PEO) chains, meanwhile Tetronic block polymer contain four PPO-PEO chains connected together at the PPO subunits to form a star shape structure (Figure 15).[206] Because Pluronic and Tetronic are a family of polymers, they vary structurally, as well as in the ratio of hydrophobic PPO and hydrophilic PEO segments in the polymer chain and in molecular weight [207]. Moreover, FDA and EMEA have approved the use for some of the linear PEO-PPO-PEO triblocks in food, pharmaceutical and agricultural industries[208] and also Pluronic has been already used as iron oxide nanoparticle coating, and their applications on cell labeling experiments point out that no transfection agents are needed.[205,209] 69 Figure 15: Schematic of MNPs. Each particle contains an iron-oxide core coated with OA and is coated with either Pluronic (single PEO-PPO-PEO subunit) or Tetronic (two PEO-PPO-PEO subunits). The PPO subunit from the copolymers adsorbs onto the OA rendering the MNPs dispersible in aqueous solution. (Pluronic and Tetronic are registered trademarks of BASF SE, Ludwigshafen, Germany.) Adapted from [207] In addition, when considering MNPs enabled MRI cell tracking, there are other parameters with high impact on the success of applying this method:[210] the amount of nanoparticle uptake,[211,212] reactive oxygen species, [213-215] the specific nanoparticle localization in the cell, [210] and the impact of transfection agents (TA).[210,216] TAs are macromolecules possessing an electrostatic charge and normally are used for non-viral transfection of DNA into the nucleus.[216,217] There are different types of TAs available, including polyamines (as PoLy-Lysine –PLL-), lipidbased agents (as lipofectamine) and heat-activated dendrimers. Usually, TAs have net positive z-potential, but it depends on the type of agent and their molecular weight.[218] However, TAs when added to cell medium are toxic to the cells, and the toxicity is proportional to the concentration in culture medium. Their importance to tag non phagocytic cells with dextran coated nanoparticles 70 has already been shown [201,204,219,220] but it is necessary to determine the appropriate ratio nanoparticles/TAs to efficiently label cells and minimize toxic-side effects.[216,221] Several studies have demonstrated the migration capacity of injected stem cells under normal and pathological conditions. Specific cell tracking is a successful application of SPIONs in MRI. The ability of load cells with enough MNPs in vitro has provided a useful technique to label and track cells in vivo by MRI. The detection limit will be conditioned by the type or MNPs and the cell load.[222] 1.5.3.4. Magnetic vectorization Recent works have studied the influence of the administration route on the efficacy of stem cell therapy, suggesting that improved methods for cell delivery to targeted areas would lead to better outcomes.[223] Several studies have already demonstrated its applicability for cell therapy for spinal cord injury [224] and magnetic vectorization of labeled endothelial progenitor cells after an intra-arterial injection in the common carotid artery in mice.[225] In order to increase the amount of injected cells close to the ischemic lesion, magnetic vectorization of cells tagged with superparamagnetic nanoparticles could be a good approach. However, the hypothesis of directing most of injected cells to the brain in order to decrease infarct size and increase functional recovery after an ischemic event was not yet demonstrated.[183] 71 Section II 72 73 2. Hypothesis Cell based therapies show great promise in the treatment of certain, so far incurable, pathologies like stroke. To determine the success of these cell therapies, non-invasive monitoring of the cells is required, and MRI is one of the most appropriate tools to do it because of its excellent soft tissue contrast and high resolution. By tagging cells with MRI contrast agents such as superparamagnetic nanoparticles, MRI can provide anatomical and pathological information and also the possibility of visualize and track cells in vivo. In this thesis we hypothesized that it is possible to synthesize biocompatible superparamagnetic nanoparticles for in vivo cell tracking. Superparamagnetic nanoparticlestagged cells can be monitored in vivo by MRI and guided with magnetic fields without harmful effects, providing information of the cellular fate after different routes of administration. Moreover, the nanoparticle labeling will allow us to study therapeutic effects in an animal model of ischemic stroke based on the localization of the cells after delivery. 74 75 3. Objectives The present work will be structured in 4 sections, and the main objectives are: 1. Superparamagnetic nanoparticles: o Synthesis and characterization of superparamagnetic nanoparticles as contrast agents for MRI. 2. In vitro validation: o Biological validation of our synthesized superparamagnetic nanoparticles: In vitro evaluation of cellular well-being after labeling and MRI cell detection. o In vitro feasibility of magnetic vectorization of labeled cells. o Long term evaluation of labeled cells: harmful effects in the cells and MR signal after proliferation. 3. Cell tracking in animal models of ischemic stroke: o Animal models of ischemic stroke and its relationship with administration routes for stem cell delivery. o In vivo feasibility of magnetic vectorization of labeled cells. o In vivo cell tracking of mesenchymal stem cells following intra-arterial and intravenous administration. 82 (Coherent Inc., CA, USA) and operated at 2 W, and combined with an ALV SP-86 digital correlator with a sampling time of 25 ns to 100 ms. Measurements were made at an angle θ = 90° to the incident beam, as appropriate for particles smaller than the light wavelength. The intensity scale was calibrated against scattering from toluene.. Each experiment was repeated at least three times. Sampling time was 5-10 min for each run in order to define an optimal correlation function. The correlation functions were analyzed by the CONTIN method to obtain the intensity distributions of decay rates (Г).From the decay rate distributions the apparent diffusion coefficients Dapp=Г/q2, q=(4πns/λ)sin(θ/2) were derived, being ns the solvent refractive index. Values of the apparent hydrodynamic radius were calculated from the Stokes- Einstein equation. Transmission Electron Microscopy (TEM): For measuring the size and determining the shape of the MNPS, TEM samples were prepared by evaporating a drop of the nanoparticles dispersion of the particles on a carbon-coated cooper grid without staining and imaged using a TEM microscope model Philips CM-12 operating at 120 kV. The NIH ImageJ software was used to calculate the mean particle diameter from TEM images. Diameters of ca. 40 particles were measured. X-Ray Diffraction (XRD): A structural analysis of dextran coated and non-coated nanoparticles were performed by X-ray diffraction (XRD) data. The samples were measured on a Philips powder diffractometer fitted with Philips PW1710 control unit, vertical Philips PW1820/00 goniometer and FR590 Enraf Nonius generator. The instrument was equipped with a graphite diffracted beam monochromator and copper radiation source ((K)=1.5406Å), operating at 40 83 kV and 30mA. The X-Ray powder diffraction pattern (XRPD) was collected by measuring the scintillation response to CuK radiation versus the 2 value over a 2 range of 10- 80º, with a step size of 0.02° and counting time of 2 s per step. Zeta-potential: Polymer-coated MNPs were measured using a Zetasizer NanoZS instrument (Malvern) equipped with a red laser (633 nm) in backscatter mode. Vibrating Sample Magnetometer (VSM): Magnetic properties of SPIONs were recorded on a vibrating sample magnetometer (VSM) from Quantum Design TM. Hysteresis loops (magnetization vs. applied external magnetic field plots) of synthesized SPIONs were recorded at 5 and 300 K and a maximum applied magnetic field of 50000 Oe. Saturation (Ms), and remanent (Mr) magnetizations, and coercitive field (Hc) were subsequently derived from experimental data. Fourier-transform infrared (FTIR): 59.5 mg of potassium bromide and 1mg of MNPs were mixed. This mixture was dried in a vacuum for 3 days and subsequently pressed to form a pellet and mounted in FTIR spectrometer (Varian FTIR 670). The spectra were obtained at a resolution of 4 cm−1 and 64 scans were accumulated for each sample. Inductive coupled plasma optical emission spectroscopy (ICP-OES): The iron concentration was determined by ICPOES (Varian Inc., Palo Alto, USA). Hereto, the nanoparticles were dissolved in 1mL of 37% HCl after which 4 mL of DW were added. The intensity of the emission wavelength for iron at 238.204 nm was measured and compared to a standard solution. ICP_OES was performed at the Material Science Institute (MTM), KU Leuven. 84 Thermal gravimetric analysis (TGA): This analysis was performed at imec (Leuven, Belgium) on a Q5000IR (TA instruments, New Castle, DE, USA) under nitrogen atmosphere to study the organic coating of the MNPs. First, 100 μL of MNPs were slowly heated to 80ºC to remove all solvents. Next, temperature was increased to 850ºC at a heating rate of 20ºC/min. Using the software provided by the manufacturer (Universal Analysis 2000, 4.5 A), the weight difference was analyzed before and after burning off the organic coating. Magnetic Resonance Imaging (MRI): Samples were prepared using an agar phantom holder following the protocol described elsewhere [227]. Triplicates of dextran, Pluronic F127 and Tetronic 908 coated nanoparticles, 4 concentrations for each type of nanoparticle were measured. MRI studies were conducted on a 9.4 T MR system (Bruker Biospin, Ettlingen, Germany) with 440 mT/m gradients, using a combination of a linearly polarized birdcage resonator of 7 cm of diameter for signal transmission and a 2x2 arrayed pick-up surface coil for signal detection. T2-weighted images were typically acquired using a Multi-Slice Multi-Echo (MSME) sequence (3 averages with a train of 16 echoes of 7.32 ms Echo time and a repetition time of 5.4 s. The whole agar phantom was covered with 14 consecutive axial slices of 1 mm thickness with a field-of-view (FOV) of 75x75 mm (with saturation bands to suppress the signal outside the FOV) and an inplane resolution 0.75mm/pixel. Post-processing of all images was performed using ImageJ software (Rasband, W. NIH). 85 4.3. Results 4.3.1. Dextran coated magnetic nanoparticles. 4.3.1.1. Iron oxide core Magnetic nanoparticles were synthesized in the presence of dextran through the chemical coprecipitation method. TEM micrographs (Figure 16) confirmed the spherical shape of the dextran-coated magnetic nanoparticles. The mean particle size using this technique was 3.70.8 nm. The crystal structure (Figure 17) of MNPs determined by XRD showed peaks at 2θ positions of ca. 30.2º; 35.6º; 43.2º; 57.1º; 62.7º, corresponding to the (220), (311), (400), (511) and (440) planes of magnetite respectively with a lattice parameter of 8.330.02 Å. It was also possible to determine the core size of the dextran coated nanoparticles from XRD data. The calculated cristallite size (Figure 17) was 4.80.5 nm.  86 Figure 16: TEM image of dextran Coated Nanoparticles. Figure 17: XRD spectrum of dextran coated nanoparticles 87 4.3.1.2. Dextran coated nanoparticles. The hydrodynamic size measured by DLS was 943 nm. Z- potential of these coated particles dispersed in DW was negative, with values of ca. -113 mV. The presence of the polymeric coating on the nanoparticles was further confirmed by TGA and FITR. The TGA data showed two peaks corresponding to mass losses at approximately 280ºC and 322ºC, close to polymer loss temperature of dextran (289ºC), demonstrating that about 66 wt % of dextran can be found in the nanoparticles (Figure 18). The FTIR spectrum exhibits several polysaccharide characteristic absorption bands, at 3351 cm-1 due to the O-H stretching. Vibrational modes (C-H) and (C-H) correspond to 2906 cm-1, 1420 cm-1 and 1380 cm-1. Water molecular bending is assigned to 1620 cm-1; 1145 cm-1 and 1010 cm-1 peaks due to C-O vibrations and the small peaks at 908 cm-1, 877 cm-1 and 746 cm-1 corresponding to -glucopyranose ring deformation modes (Figure 19). There are also two absorption bands at 578-1 and 430 cm-1 corresponding to Fe-O vibration modes. Figure 18: TGA spectrum of dextran polymer and dextran coated NPs. 88 Figure 19: FTIR spectrum of Dextran coated nanoparticles. VSM data showed that Dextran coated nanoparticles exhibited superparamagnetic behavior (Figure 20). MRI contrast in T2* weighted images of 4 different concentrations of dextran coated nanoparticles were measured: 0.19mM; 0.09mM; 0.05mM and 0.02mM (Figure 21). The relaxation rate (R2) was plotted versus the concentration of iron. A linear relationship was found between R2 and the MNPs concentration, enabling the calculation of the transverse relaxivity (r2) (the slope). The r2 for dextran coated magnetic nanoparticles resulted in 699±4 mM-1s-1. 89 Figure 20: VSM results of dextran coated magnetic iron oxide nanoparticles. Figure 21: MRI T2 weighted of dextran-coated nanoparticles for each concentration. 4.3.2. Polymer coated magnetic nanoparticles: Pluronic F127 and Tetronic 908. 4.3.2.1. Core Pluronic and Tetronic-coated MNPs were prepared following a two step procedure. First, magnetic cores were synthesized and next the polymer coating was added. Magnetic cores coated with oleic acid were obtained 90 through the chemical coprecipitation method, obtaining spherical MNPs (Figure 22). The core size was 8.61.4 nm determined through TEM. The core size was also determined by XRD spectrum data and by DLS, being 7.70.8 nm and 7.30.8 nm respectively. Figure 22: TEM picture of superparamagnetic nanoparticle cores. The core crystal structure determination using XRD (Figure 23) showed peaks at 2 positions of approximately 30.2º; 35.6º; 43.2º; 57.1º; 62.7º, corresponding to the (220), (311), (400), (511) and (440) planes of magnetite respectively with a lattice parameter of 8.3610.004 Å. Data obtained through VSM measurements showed that the cores exhibited superparamagnetic behavior at room temperature (Figure 24). 91 Figure 23: XRD spectrum of magnetic nanoparticle cores synthesized through chemical coprecipitation method. Figure 24: Magnetization versus magnetic field plots for magnetic cores of superparamagnetic nanoparticles. The presence of oleic acid attached to the surface of the nanoparticle was evaluated through TGA. In Figure 25A two 98 curvature of the particles. This amorphous layer is detected by TEM, but does not contribute to the XRD diffractograms. Secondly, crystal defects can form within the MNPs crystal structure distorting the XRD data. Additionally, a MNP can be made up of several crystals, all of course being smaller than the size measured through TEM. A possible reason for the unexpected larger grain size can be attributed to the both methods, where size analysis based on TEM has a limited sample size, whereas the XRD data is taken from larger samples. Overall, both experimental values are in concert with each other. It is also important to mention that the standard deviation derived from the TEM measurement and therefore the size distribution is adequate considering that they were synthesized following the chemical coprecipitation method. The crystalline structure from XRD data reflects a magnetite and/or maghemite crystal with a cubic spinel structure. Magnetite can be easily oxidized in air to form maghemite,[228] making it difficult to distinguish between them. This even can occur during sample preparation. The maghemite phase is known to exhibit few extra peaks at 23.77º (210) and 26.10º (211), and these peaks may be used to distinguish it from magnetite phase, however the intensity of these peaks are very weak (5%) for the positive identification of the maghemite phase.[229] Our XRD experimental spectrum of synthesized MNPs did not exhibit those diffraction peaks, but this fact does not guarantee that the sample is a single-phase maghemite, considering the previous point made. With regard to the calculated lattice parameter from XRD spectrum, the value 8.330.02 A is closer to maghemite (8.3515 A –JCPDS 39 1346-) than to magnetite (8.396 A –JCPDS 19 629-), what suggests that the obtained cores through chemical coprecipitation method 99 in presence of dextran are mostly composed of maghemite. Moreover, the size of the dextran-coated nanoparticles was measured by DLS with a mean size 943 nm is in agreement with previous works.[230] Apart from the size, the surface charge is an important factor to take into account in terms of nanoparticle stability. The synthesized dextran-coated nanoparticles are stable in distilled water and negatively charged (-113 mV), a remarkable aspect to take into account for further in vitro and in vivo experiments in terms of not only cell labeling, but also of protein binding as well.[231,232] To study the nature of the nanoparticle coating, FTIR and TGA were performed. FTIR spectrum demonstrated the presence of dextran and magnetite. The peak at 3351cm-1 is attributed to the stretching vibrations of –OH, which is assigned to OH- absorbed by iron oxide nanoparticles or leftover water from KBr preparation. Also vibrational modes (C-H) and (C-H) at 2906 cm-1, 1420 cm-1 and 1380 cm-1, water molecular bending due to C-O vibrations at 1620 cm- 1, 1145 cm-1 and 1010 cm-1 and small peaks at 908 cm-1, 877 cm-1 and 746 cm-1 corresponding to -glucopyranose ring deformation modes (Figure 18) demonstrated the presence of dextran in the sample. The two absorption bands at 578 cm-1 and 430 cm-1 corresponding to Fe-O vibration modes verify the presence of the iron oxide core. However, it is necessary to look of TGA data for a better understanding of the dextran coated magnetic nanoparticles. TGA data indicates the presence of two close peaks (280ºC and 322ºC), demonstrating that the 66% of total weight of the sample was dextran, which points towards a thick and dense coating of the MNPs, this result agrees with other published works.[199,233] However, those two peaks also 100 show that the high stability of the sample is also conditioned for the presence of free dextran, which would be more easily detached (therefore 280ºC) and nanoparticleattached dextran (more difficult to unbound, 322ºC). In this sense, 32.44 wt % would correspond to free dextran and 33.57 wt% to nanoparticle bonded one. Together, the results of FTIR and TGA, confirm the presence of dextran coating on the magnetic nanoparticles and also free polymer increasing the stability of the suspension. Magnetic characterization of the MNPs showed the relative magnetization curve as a function of magnetic field at 5K and 300K. The absence of hysteresis loop at 300K indicated the characteristic superparamagnetic behavior of the particles. This means that these particles have noninteracting domains, so no magnetic clustering and high magnetization saturation, an important aspect for MRI contrast enhancement.[199] To assess the transverse relaxivity measured by MRI and therefore that those particles are suitable as MRI contrast agents, T2-weighted images were acquired. Four different concentrations of dextran coated nanoparticles were measured and all parameters, such as echo time, repetition time and resolution were adjusted to obtain a well-defined exponential decay. The higher concentration (0.19mM) of particles could not be adjusted to an exponential decay because there were not enough data points to perform the adjustment. The relaxivity r2 of dextran-coated particles from this data was 6994 mM-1s-1. 101 4.4.2. Pluronic F127 and Tetronic 908 coated superparamagnetic nanoparticles. The iron oxide nanoparticles formation through chemical coprecipitation method is a fast reaction, and happens when the ammonium hydroxide is added to iron salts mixture. After the formation of iron oxide nanoparticles, oleic acid is added in order to create an OA shell surrounding the iron oxide nanoparticle, which will be used as an anchor for further polymer functionalization. However, due to the excess of OA for the coating process, a multilayer of OA on the surface of the nanoparticles is formed, making them hydrophilic like. This multilayer is not stable, so a deep DW wash process will turn them into their natural hydrophobic state, and easily dispersible in organic solvents. This full procedure, including chemical coprecipitation, OA coating and nanoparticle washing results in OA-coated iron oxide nanoparticles nicely dispersed in chloroform.[205] The oleic acid coated nanoparticles were found to be spherical in TEM micrographs, and it was also possible to measure their size, being this one about 8.61.4 nm. DLS and XRD nanoparticle sizes were 7.70.8 nm and 7.30.8 nm respectively. It is expectable that standard deviation (SD) of TEM size determination is higher than XRD and DLS due to the number of total counts taken into account to calculate the mean. However, OA-coated nanoparticles size determination was similar for all techniques, and also consistent with previous studies.[205,234] In accordance with dextran-coated MNPs XRD spectrum, the position and intensities of OA-coated nanoparticles peaks matched well with magnetite and/or maghemite crystals with a cubic spinel structure. Apart from the fact 102 that the peaks at 23.77º (210) and 26.10º (211) are not observed, it is not possible to reassure that the sample is pure magnetite. The calculated lattice parameter from XRD spectrum is about 8.3610,004 Å, which is in between maghemite (8,3515 A –JCPDS 39 1346-) and magnetite (8,396 A –JCPDS 19 629-), exhibiting, as a consequence, the presence of both phases. The magnetization curve of these OA-coated nanoparticles at room temperature showed not coercivity and remanence, indicating that the sample is superparamagnetic at room temperature. With regard to the functionalization with Pluronic F127 and Tetronic 908 of OA-coated nanoparticles, DLS size determination were performed, as it was not possible to measure through TEM pictures the size of the polymer coated nanoparticles. Tetronic coated MNPs are slightly bigger compare to Pluronic, 1206 nm and 1546 nm respectively and these values are in consonance with other block copolymers coated nanoparticles.[207] Z-potential was negative for both, Pluronic F127 coated MNPs were - 183 mV and -243 mV for Tetronic 908 coated MNPs, which confirms the stability of the nanoparticles and prevents aggregation for storage[235]. TGA of OA-, Pluronic F127- and Tetronic 908- coated nanoparticles was performed to quantify OA and polymer contents. For all formulations the percentage of oleic acid present in the MNPs went from 19% to 26%, and the amount of polymer was 35.5% for Pluronic F127 and 42% for Tetronic 908. Because these preparations were washed through magnetic separation, these percentages correspond just to attached polymer. Nevertheless, FTIR spectrums showed that OA-, Pluronic F127- and Tetronic 908-coated MNPs 103 have all peaks at same wave numbers except for 1100 cm- 1, which is not present in OA-coated nanoparticles. It has been published that broad bands around 1250-1000 cm-1 are due to the CH2 rocking and C-O-C stretch vibrations of Pluronic. The well-defined band at 1100 cm-1 is typical for a block copolymer in the optimal formulation in which OA completely covers the iron oxide nanoparticle surface.[205] MRI T2* weighted images were acquired to calculate the transverse relaxivity and evaluate if these particles are suitable as MRI contrast agents for cell tracking. Four different concentrations of each polymer coated nanoparticles were measured and all parameters, such echo time, repetition time and resolution were adjusted to obtain a well-defined exponential decay. As happened before for dextran coated MNPs, the higher concentrations (0.36 mM for Pluronic F127- and 0.45 mM for Tetronic 908- coated MNPs) of particles cannot be adjusted to an exponential decay because there were not enough data points to perform the adjustment. On the other hand lower concentration data points described a nice adjustment. With this data it was also possible to determine the relaxivity r2 of Pluronic F-127 and Tetronic 908- coated particles, being 398±12 mM-1s-1 and 33020 mM-1s-1 respectively. It was expected that both values are similar because they have common magnetic cores, similar coatings and almost equal DLS sizes. 104 4.5. Conclusions In conclusion, 3 types of superparamagnetic nanoparticles coated with different polymers have been developed. After a careful chemical coprecipitation synthesis processes, spherical-shaped maghetite/maghemite nanoparticles exhibiting superparamagnetic behavior at room temperature were obtained. Two methods were described for core synthesis and polymer coating; in situ dextran coated MNPs and two step procedure for Pluronic F127 and Tetronic 908 coating. Both processes resulted in clustered, well-coated and nicely dispersed nanoparticle suspensions in aqueous solutions, being biocompatible and suitable as a MRI contrast agent. Overall, this has resulted in an easy method to produce water-stable superparamagnetic nanoparticles with different size and coatings, which will be essential for next biological and biomedical applications. 105 Section IV in vitro validation 106 107 In this section we present the in vitro validation of the superparamagnetic nanoparticles synthesized in the previous chapter. In vitro studies are divided in three chapters, where different aspects of the cell labeling process will be detailed. As we have mentioned in the introduction section, several studies have demonstrated the need of transfect agents for an efficient labeling process. In previous test (data not shown) we have observed that Pluronic F127- and Tetronic 908-coated MNPs do not need Transfect Agents (TAs), while TAs are indispensable for dextran-coated labeling. Thus, in the first chapter of this section (Chapter 4) we have evaluated the use of TAs for dextran-coated MNPs previously synthesized and its influence in mesenchymal stem cells. Considering these transfection agent results, in Chapter 5 we have studied the MNPs influence in the well-being of two types of cells. Rat mesenchymal stem cells and the multi-potent neural progenitor cell line C17.2. were used for the in vitro evaluation due to their therapeutic potential. Here, different concentrations and incubation times were evaluated to elucidate which of our synthesized MNP is the most suitable for further in vivo cell tracking experiments. However, prior to in vivo studies, in the last chapter of this section we have studied the long-term evolution of the MRI signal of labeled mesenchymal stem cells. We have evaluated in vitro how the proliferation rate affects the MRI detection limit and if the labeling has a long-term harmful effect in the cells. In addition, we have also evaluated the magnetic vectorization in vitro with MSCs labeled with our MNPs. 114 5.2.12. Statistical analysis The statistical analyses of the data were performed using GraphPad Prism 5.01(GraphPad Software, Inc) and mean and standard deviation are presented. 5.3. Results 5.3.1. Cellular internalization of the MNPs The amount of internalized iron oxide nanoparticles was determined by ICP and results are plotted in (Figure 28). PLL alone does not increase the Fe signal measured by ICP, which can be observed in control (2.32±0.04 pg/cell) and control+PLL (2.34±0.02 pg/cell). D-MNPs internalization is PLL dose dependent; no PLL labeling group showed similar values than controls, 2.75±0.02 pg/cell for 200 g/mL D- MNPs and 2.57±0.05 pg/cell for 100 g/mL D-MNPs. Increasing the PLL concentration in cell medium enhances the uptake, being for 0.75 g/mL of PLL about 6.1±0.2 pg/cell for 200 g/mL D-MNPs and 5.27±0.14 pg/cell for 100 g/mL D-MNPs. For the highest PLL concentration studied (1.5 g/mL), experimental values were 10.4±0.5 pg/cell for 200 g/mL D-MNPs and 9.9±0.3 pg/cell for 100 g/mL D- MNPs. 115 Figure 28: Iron content of D-MNPs labeled and non-labeled MSCs. 5.3.2. Magnetic resonance imaging of labeled cells To assess the D-MNPs internalized contrast, MRI of labeled and control cells was performed. T2 and T2* maps were calculated from T2 and T2* weighted images respectively and R2 and R2* relaxivities were obtained for each concentration using ImageJ software. Hence, R2 were similar for Control, No PLL 200 g/mL D-MNPs and No PLL 100g/mL D-MNPs, being 5.00.2 s-1, 4.91.1 s-1 and 5.21.2 s-1 respectively. For 200 g/mL D-MNPs, R2 = 7.11.4 s-1 for 0.75g/mL and R2 = 8.60.8 s-1 for 1.5g/mL. And for 100 g/mL D-MNPs, R2 = 7.20.8 s-1 for 0.75g/mL and R2 = 7.11.9 s-1 for 1.5g/mL. Experimental values are plotted in Figure 29, and MSCs T2-weighted images are shown in Figure 30. 116 Figure 29: R 2 values [s -1 ] of D-MNPs labeled and non-labeled MSCs. Figure 30: T2-weigthed images of MSCs. (A) Control; (B) 200 g/mL of D- MNPs and No PLL, (C) 200 g/mL of D-MNPs and 0.75 g/mL of PLL, (D) 200 g/mL of D-MNPs and 1.5 g/mL of PLL, (E) 100 g/mL of D-MNPs and No PLL, (F) 100 g/mL of D-MNPs and 0.75 g/mL of PLL, (G) 100 g/mL of D-MNPs and 1.5 g/mL of PLL. Accordingly to T2* maps, R2* for Control cells was 154 s-1. For 200 g/mL D-MNPs was: 245 s-1for No PLL, 385 s-1 for 0.75 g/mL of PLL and 64.8 0.6 s-1 for 1.5 g/mL of PLL. For 100 g/mL D-MNPs, R2 values were: 205 s-1 for No PLL, 356 s-1 for 0.75g/mL of PLL and 519 s-1 for 1.5g/mL of PLL. Experimental values are plotted in Figure 31 and MSCs T2- weighted images are shown in Figure 32. 117 Figure 31: R2* values [s-1] of D-MNPs labeled and non-labeled MSCs. Figure 32: T2*-weigthed images of MSCs. (A) Control; (B) 200 g/mL of D- MNPs and No PLL, (C) 200 g/mL of D-MNPs and 0.75 g/mL of PLL, (D) 200 g/mL of D-MNPs and 1.5 g/mL of PLL, (E) 100 g/mL of D-MNPs and No PLL, (F) 100 g/mL of D-MNPs and 0.75 g/mL of PLL, (G) 100 g/mL of D-MNPs and 1.5 g/mL of PLL. 5.3.3. Optical microscopy and TEM Two microscopic techniques were used to assess the cellular labeling. The most common is Prussian Blue staining, which shows the MNPs in a blue color in bright field microscopy imaging. Control and No PLL conditions showed no Prussian Blue Figure 33((A), (B), (C), (D)), therefore no cell labeling. Under PLL presence, Prussian Blue staining can be observed, revealing also a dependence between amount of PLL and blue staining as can be seen in Figure 33 (E, F, G, H). For 100 g/mL of D-MNPs results were similar (Data not shown). 118 Apart from optical microscopy, transmission electron microscopy was performed. There were no particles in the cells for the No PLL group (Figure 34 (A)). For the rest of the groups, D-MNPs were engulfed in cellular compartments and distributed along the cell. MSCs labeled with 1.5 g/mL of PLL and 200 g/mL of D-MNPs are shown in Figure 34 (B), where black circular dots can be observed within the cell cytoplasm. It was also possible to zoom into one labeled cell (see Figure 35 (A), (B), (C)), and the particle pattern inside the endosomes can be observed in Figure 35(C). One important result is that all particles observed were fully internalized and not attached to the cell membrane. 119 Figure 33: Prussian Blue staining of MSCs. (A) Control cells in phase contrast, (B) Bright field of Control cells, (C) Phase contrast 200 g/mL D- MNPs and NO PLL, (D) 200 g/mL D-MNPs and NO PLL in Bright field, (E) Phase contrast 200 g/mL D-MNPs and 0.75 g/mL, (F) 200 g/mL D-MNPs and 0.75 g/mL in bright field, (G) Phase contrast 200 g/mL D-MNPs and 120 1.5 g/mL PLL and (H) 200 g/mL D-MNPs and 1.5 g/mL PLL in bright field. Figure 34: TEM micrographs of (A) No PLL 200 μg/mL D-MNPs, (B) 1.5 μg/mL PLL and 200 μg/mL of D-MNPs. 121 Figure 35: TEM micrographs of 1.5 g/mL of PLL and 200 g/mL of D-MNPs (A) one labeled cell, (B) 200x zoom of A, (C) 400x zoom of A. 122 5.3.4. Effect of the MNPs internalization on the cells Proliferation rate experimental data is referred to control group (100%). There are no significant differences between all groups in terms of proliferation 12h after full D-MNPs internalization, see Figure 36. Figure 36: Proliferation rate for D-MNPs labeled and non-labeled MSCs. Moreover, LDH viability assay performed at same timepoint showed that there are not significant differences between control group and the rest of the conditions. As it is shown in Figure 37, experimental values of control are slightly increased for the presence of PLL and/or D-MNPs but there is not a statistical difference between conditions. 123 Figure 37: Viability rate [% respect control] for D-MNPs labeled and nonlabeled MSCs. To evaluate phenotypic changes in MSCs after cell tagging, CD45 and CD90 markers for flow cytometry were used to characterize labeled and non-labeled MSC. Figure 38 shows the results for the control and for highest concentration of D-MNPs, 200 g/mL. There were no differences between all labeling conditions; all groups were negative for CD45 and positive for CD90, as can be observed in Figure 38 (blue). Conditions of 100 g/mL showed the same behavior (data not shown). It can also be noted that the shape of side scattering (SSC), in Figure 38 is different depending on the condition, and this can be correlated with the amount of PLL and therefore with internalized MNPs. Control and No PLL are more similar between them than 0.75 g/mL PLL, and 1.5 g/mL PLL is the most different one. SSC is influenced by the complexity of the sample, pointing out the amount of internal organelles. The more PLL combined with D-MNPs, the more cellular uptake and as a consequence, the higher will be the SSC in flow cytometry. 130 131 6. Dextran-, Pluronic F127- and Tetronic 908- coated superparamagnetic nanoparticles: In vitro validation 6.1. Hypothesis The aim of the present chapter is to evaluate the biocompatibility of our synthesized dextran-, pluronic F127- and tetronic 908-coated MNPs in 2 different cell types, MSCs and C17.2., combining D-MNPs with PLL for an efficient cellular uptake. Moreover, we will study the influence of the incubation time and MNPs concentration in the well-being, in the uptake, and therefore in the MRI signal produced by these labeled cells. 132 6.2. Materials & Methods 6.2.1. Cell lines Mesenchymal Stem Cells (MSCs) were purchased from Cultrex, Trevigen and cultured in IMDM (78%), Fetal bovine serum (10%), Horse serum (10%), Penicillin-Streptomycin (1%) all from Gibco Invitrogen, and Amphotericin-B (1%) from Sigma-Aldrich. Cell passage numbers between 7 and 18 were used in these experiments. Multi-potent neural progenitor cell line C17.2 were characterized by Markus Auswendt and kindly donated by Prof. Mathias Hoehn (Max-Planck Institute, Cologne, Germany) and cultured in DMEM (78%), Fetal bovine serum (10%), Horse serum (5%), Penicillin-Streptomycin (1%) all from Gibco Invitrogen, and Amphotericin-B (1%) from Sigma- Aldrich. 6.2.2. Dextran-, Pluronic F127- Tetronic 908-coated superparamagnetic nanoparticles Dextran coated MNPs (D-MNPs), Pluronic F127 coated MNPS (P-MNPs) and Tetronic 908 coated MNPS (T-MNPs) were synthesized in the Clinical Neuroscience Research Laboratory following the protocols described in chapter 3. 5.2.3. Preparation of Poly-L-lysine hydrobromidecoated D-MNPs One hour prior to cell incubation, different concentrations of D-MNPs (500 and 250 g/mL) were mixed with MSCs or C17.2 mediums and with 1.5 g/mL PLL (Sigma-Aldrich) and vigorously shaken. 133 6.2.3. Cell labeling with MNPs Cells were labeled with 3 different types of MNPs, as it has been mentioned before. Two concentrations (500 and 250 g/mL for D-MNPs and 35 and 15 g/mL for P-MNPs and T- MNPs) respectively; and two incubations times (24h and 6h) were studied in this chapter (n=6 for each condition). Labeling protocol was the same for all groups. Cells were labeled following the protocol described elsewhere [227] with slight modifications, already used in the previous chapter. Cells were incubated with different concentrations of MNPs for 6h or 24h. The MNP containing medium was removed and the cells were 3 times washed with 1.5 mL of Phosphate Buffered Saline (PBS without Mg2+ and Ca2+, Gibco Invitrogen) to remove non-attached MNPs. After washing, the cells were left overnight (12h) in 1 mL of fresh medium. The following day the medium was removed, cells were washed once with 1.5mL of PBS and 0.5mL of EDTA-trypsin (Gibco Invitrogen) was added to detach the cells from the well. The trypsin was neutralized with fresh medium and the detached cells were collected in a Falcon tube. After a mild centrifugation, the supernatant was discarded and the cells were resuspended in fresh medium. 6.2.4. Proliferation To determine the proliferation rate, total cell count was performed with Trypan Blue staining and a Neubauer chamber. Samples were diluted 1:5 with PBS and 1:2 with Trypan Blue. 134 6.2.5. Cell viability assay: Lactate Dehydrogenase Assay (LDH) For assessing the viability of the cells after labeling compared to control, supernatants from last 12h (fresh medium incubation) were collected in eppendorf tubes. A negative control of lysed cells was also included. After a mild centrifugation of the supernatants, 75L were mixed in a 96 well plate with 150L LDH reagents following the manufacturer protocol (Lactate Dehydrogenase Assay Kit. Sigma-Aldrich). After 20 minutes incubation, the plate was read in Synergy2 (Biotek) at 490nm. To calculate the viability rate respect to control, next equation was applied. 󰇟%󰇠        100 6.2.6. Intracellular iron determination An amount of 100.000 cells per condition was dissolved in 1 mL of 37% HCl (Hydrochloric acid fuming 37% extra pure, Merk) and then 4 mL of DIW was added to reach total volume of 5 mL. Iron concentration was determined by inductive coupled plasma optical emission spectroscopy (ICP-OES) (Varian Inc). The intensity of the emission line at 238.204 nm was measured for iron and compared to a standard solution. ICP-OES was performed at CACTUS-Lugo (University of Santiago de Compostela). 6.2.7. Prussian Blue staining Prussian Blue staining was performed to demonstrate the uptake of the MNPs by the cells. Labeled cells were planted and after 8h, were washed with PBS and incubated for 20 minutes with a mix of equal parts of aqueous solution of 20% 135 HCl and aqueous solution of 10% Potassium ferrocyanide Trihydrate (KFe(CN)6.3H2O, FW 422.2, Sigma-Aldrich). After incubation, cells were washed 3 times with PBS and photographs were taken using an inverted microscope (Olympus IX51). 6.2.8. Transmission electron microscopy (TEM) To assess the localization of the D-MNPs in the cell, TEM images of cells were taken. Fixation and postfixation of 500.000 cells were carried out in 2% glutaraldehyde in sodium cacodylate buffer and in 1% OsO4 in the same buffer. Inclusion was done in Spurr’s epoxy resin. Semithin sections (0.5 μm) were stained with Toluidine blue and ultrathin sections (100 nm) were stained with uranyl acetate and lead citrate. 6.2.9. Magnetic resonance imaging of labeled cells Agar phantom was made following the procedure described elsewhere.[227] 100.000 cells per condition were scanned in a Bruker Biospec 9.4 T small animal MR scanner (Bruker, Biospin, Ettlingen, Germany) equipped with actively shielded gradients (400 mT/m). A quadrature radiofrequency transmit-receive resonator was used for data acquisition. T2-weighted images were acquired using a MSME sequence of 10.26 ms echo time, 3 s repetition time, 16 echoes, 14 slices, 1 average, FOV of 7.5 cm x 7.5cm and matrix size of 256 x 256. T2*-weighted images were acquired using a MGE sequence of 4.44 ms echo time, 6.75 ms echo spacing, 1.8 s repetition time, 16 echoes, 14 slices, 2 averages, FOV of 7.5cm x 7.5cm and a matrix size of 256 x 136 256. Post-processing was performed using ImageJ software (Rasband, W. NIH). 6.2.10. Statistical analysis The statistical analysis of the data was performed using GraphPad Prism 5.01 (GraphPad Software, Inc). For descriptive analyses, means and standard deviations were calculated for the normally distributed continuous variables. For confirmatory analyses one-way ANOVA was performed to calculate overall p-values for the comparison of the different groups for each variable. The global significance level for all statistical test procedures was chosen as =0.05. 137 6.3. Results Thirteen conditions were studied for each type of cell. Every group and the nomenclature assigned are presented below: <Polymer>-MNPs <Concentration> <Incubation time> o Control o D-MNPs 500 24h: Cells labeled with 500 g/mL D- MNPs with1.5 g/mL of PLL incubated for 24 h. o D-MNPs 500 6 h: Cells labeled with 500 g/mL D-MNPs with1.5 g/mL of PLL incubated for 6 h. o D-MNPs 250 24 h: Cells labeled with 250 g/mL D- MNPs with1.5 g/mL of PLL incubated for 24 h. o D-MNPs 250 6h: Cells labeled with 250 g/mL D-MNPs with1.5 g/mL of PLL incubated for 6 h. o P-MNPs 35 24h: Cells labeled with 35 g/mL P-MNPs incubated for 24 h. o P-MNPs 35 6h: Cells labeled with 35 g/mL P-MNPs incubated for 6 h. o P-MNPs 15 24h: Cells labeled with 15 g/mL P-MNPs incubated for 24 h. o P-MNPs 15 6h: Cells labeled with 15 g/mL P-MNPs incubated for 6 h. o T-MNPs 35 24h: Cells labeled with 35 g/mL P-MNPs incubated for 24 h. o T-MNPs 35 6h: Cells labeled with 35 g/mL P-MNPs incubated for 6 h. o T-MNPs 15 24h: Cells labeled with 15 g/mL P-MNPs incubated for 24 h. o T-MNPs 15 6h: 15 g/mL P-MNPs incubated for 6 h. 138 6.3.1. Cellular internalization of the MNPs The amount of internalized iron oxide nanoparticles was determined by ICP and results are shown in Table 1 and Figure 39. All labeling conditions showed variable amount of iron internalization depending on MNPs concentration in the medium and also on the incubation time. MSCs C17.2 Mean [pg] SD Mean [pg] SD Control 0.4 0.1 0.4 0.0 D-MNPs 500 24h 15 2 9.3 1.6 D-MNPs 250 24h 8.7 0.9 8.4 0.1 D-MNPs 500 6h 7.7 0.9 4.8 0.7 D-MNPs 250 6h 5.9 0.9 5.1 0.3 P-MNPs 35 24h 36.8 1.4 28 7 P-MNPs 15 24h 16.2 0.4 9.7 0.8 P-MNPs 35 6h 39 3 33 3 P-MNPs 15 6h 8.9 0.5 12.7 0.8 T-MNPs 35 24h 37 3 39 1 T-MNPs 15 24h 9.9 1.5 14 2 T-MNPs 35 6h 36 4 16.7 1.5 T-MNPs 15 6h 12.4 1.4 8 3 Table 1: Iron content per cell [pg] Figure 39: Iron internalization [pg/cell] for all labeling conditions. 139 6.3.2. Magnetic resonance imaging of labeled cells To assess the contrast generated by MNPs internalization, MRI of all labeling conditions was performed. T2 and T2* maps were calculated from T2 and T2* weighted images respectively, and R2 and R2* relaxivities were obtained for each condition using ImageJ. R2* values are shown in Table 2 and plotted in Figure 40. Moreover, the contrast generated by the labeled and non-labeled cells can be observed in the first echo of T2* weighted images in Figure 41 and Figure 42 for MSCs and C17.2 respectively. MSCs C17.2 Mean [s -1 ] SD Mean [s -1 ] SD Control 32 5 31 5 D-MNPs 500 24h 69 23 60 3 D-MNPs 250 24h 57 17 59 5 D-MNPs 500 6h 62 21 41 7 D-MNPs 250 6h 47 14 40 12 P-MNPs 35 24h 69 24 63 24 P-MNPs 15 24h 60 21 66 5 P-MNPs 35 6h 71 32 53 14 P-MNPs 15 6h 78 29 58 12 T-MNPs 35 24h 53 26 57 31 T-MNPs 15 24h 69 10 67 11 T-MNPs 35 6h 72 16 62 8 T-MNPs 15 6h 72 22 48 8 Table 2: R 2 * of different label conditions determined by MRI in an agar phantom with concentrations of 500 cell/L. 146 Figure 48: (A) MSC labeled with D-MNPs 500 24h, (B) MSCs labeled with P- MNPs 35 24h, (C) MSCs labeled with T-MNPs 35 24h, (D) C17.2 labeled with D-MNPs 500 24h, (E) C17.2 labeled with P-MNPs 35 24h, (F) C17.2 labeled with T-MNPs 35 24h 147 Figure 49: MNPs internalized. (A) PLL coated D-MNPs internalized in an endosome in the cellular cytoplasm (1.5 μg/mL PLL and 500 μg/mL D- MNPs incubated for 24h), (B) P-MNPs internalized in an endosome in the cellular cytoplasm (35 μg/mL of P-MNPs incubated for 24h). (Scale bar 500 nm) 6.3.4. Effect of the MNPs internalization in the cells Proliferation rate experimental data is referred to control group (%). Experimental values are shown in Table 4 and plotted in Figure 50 (A) and (B). For the highest concentrations of MNPs there is a reduction in proliferation rate respect to control for both cell types. 148 MSCs C17.2 Mean [%] SD Mean [%] SD Control 100 18 100 4 D-MNPs 500 24 h 67 19 80 15 D-MNPs 250 24h 84 22 66 20 D-MNPs 500 6h 86 10 84 22 D-MNPs 250 6h 102 9 76 10 P-MNPs 35 24h 46 9 62 4 P-MNPs 15 24h 51 11 91 26 P-MNPs 35 6h 61 16 66 9 P-MNPs 15 6h 87 2 67 16 T-MNPs 35 24h 44 4 50 19 T-MNPs 15 24h 73 5 55 10 T-MNPs 35 6h 75 6 65 7 T-MNPs 15 6h 62 17 79 3 Table 4: Proliferation rate [%] respect to control. Moreover, LDH viability assay showed that there are not significant differences between control group and any labeling condition for MSCs and well as for C17.2 (Table 5 and Figure 50 (C) and (D)). MSCs C17.2 Mean [%] SD Mean [%] SD Control 100 2 100 4 D-MNPs 500 24 h 96 2 102 4 D-MNPs 250 24h 95 2 102 3 D-MNPs 500 6h 97 3 100 3 D-MNPs 250 6h 98 1 98 5 P-MNPs 35 24h 97 1 97 4 P-MNPs 15 24h 95 1 100 3 P-MNPs 35 6h 98 2 98 3 P-MNPs 15 6h 100 2 98 3 T-MNPs 35 24h 96 2 100 3 T-MNPs 15 24h 97 2 100 3 T-MNPs 35 6h 100 2 98 3 T-MNPs 15 6h 99 2 100 3 Table 5: Viability [%] respect to control. 149 Figure 50: (A) Proliferation rate (respect to control %) for MSCs after labeling, (B) Proliferation rate (respect to control %) for C17.2 after labeling, (C) Viability rate (respect to control %) for MSCs after labeling and (D) Viability rate (respect to control %) for MSCs after labeling. (* p<0.05; ** p<0.01) 6.4. Discussion In this chapter we have successfully validated the use of our synthesized dextran-, pluronic F127- and tetronic 908- coated superparamagnetic nanoparticles for in vitro applications. Previously we had reported the dependence on transfect agents as PLL for D-MNPs cellular internalization, and here we have proven that no transfection agent is needed for P-MNPs and T-MNPs cell labeling. However, besides transfect agent influence findings, other parameters such as incubation time, MNPs 150 concentration or cell type were also studied in order to find out the best labeling conditions. But, the best scenario requires not only the best MRI signal, which actually is the main goal as MRI contrast agents; but must also maintain the well-being of the cells, as they will be used as treatment agents in a potential therapy for stroke. One of the most direct parameters that can be influenced by the presence of MNPs is cell proliferation. With regard to the proliferation rate of labeled cells, PLL coated D-MNPs do not significantly affect their proliferative capacity compared to control for both types studied, MSCs and C17.2. Viability is slightly reduced for these conditions, but it only has statistic significance for D-MNPs 250 24h in MSCs (p<0.05). On the other hand, P-MNPs and T-MNPs have a similar behavior between them. For both cell types, the highest concentration (35 g/mL) and the longest incubation time (24h), reduce the proliferation up to 45- 60%, for P-MNPs (p<0.05) and also for T-MNPs (p<0.01). In addition, the lowest concentration (15 g/mL) of P-MNPs reduces the proliferation of MSCs to 51% (p<0.05) and also T-MNPs for C17.2 to 55% (p<0.01). Twelve hours after labeling, LDH viability assay shows a reduction in all conditions being statistic significant just for MSCs labeled with P-MNPs 250g/mL for 24h. As it has been mentioned before we can attribute this fact to an artifact (higher concentration does not show a significant effect) and not as a viability reduction. In general, the proliferation after labeling is reduced and depends on the incubation time as well as in the concentration of MNPs in medium. For MSCs it is possible to observe a trend, proliferation rate decreases with the increase of incubation time and MNPs concentration for all coated MNPs. In the case of C17.2, same trend can be observed for T-MNPs, meanwhile standard deviations of D-MNPs and P-MNPs are too high to 151 show this effect. In regard to viability 12h after labeling, there are no significant differences between groups in terms of LDH release. In order to perform Prussian Blue staining, harvested cells were replanted again. With this staining it is possible to observe that all studied conditions showed MNPs, which can be observed through optical microscopy in blue color Furthermore, the fact of replant the cells also indicates that the behavior of labeled cells is the same as controls, they get attached to the well and enabling Prussian Blue staining. Other way to evaluate the wellbeing of labeled cells and also demonstrate the MNPS presence is based on their morphology in TEM micrographs. There were no differences in morphology observed by TEM and all coated MNPs are encapsulated in vesicles along the cytoplasm. As it was demonstrated in the previous chapter, we can assess that the labeling protocol used is highly efficient to tag cells with MNPS, and no MNPS were observed attached to the cellular membrane. Particularly, the endosomes of cells labeled with P-MNPs and T-MNPs showed big clustering of MNPs where individual particles (cores) cannot be easily observed. On the other hand, endosomes containing D- MNPs are more regular in shape and present a uniform dotted pattern, where MNPs cores can be observed. The MNP internalization is variable depending on the MNPs type. We have already reported the dependency on PLL for D-MNPs internalization, however for this study the concentration has been kept constant (1.5g/mL). However, it is possible to observe the dependency on concentration and incubation time for both types of cells labeled with PLL coated D-MNPs. 152 The dependency on concentration is more remarkable for P-MNPs and T-MNPs, where the amount of MNPs during the incubation process is decisive. The more concentration, the more internalization value with one exception, T-MNPs 35g/mL 6h, which showed similar values to 15 g/mL. Nevertheless, it is important to remark that MNP internalization of P-MNPs and T-MNPs (for 35 g/mL) are 2, 3 or even 4 times higher than D-MNPs (for 500 g/mL). This could be explained based on two factors, on one hand the nature of the coating and, on the other hand on terms of MNPs stability in cell medium. Considering the small size of all these MNPs, it is not possible to observe clustering or aggregation after MNPs dispersion in cell medium by optical microscopy, but P-MNPs and T-MNPs could be precipitating onto the cells facilitating the uptake. TEM images indicate that Pluronic and Tetronic coated MNPs could be micro-aggregated due to the irregular morphology of the endosomes observed. This idea would be also supported by ICP data, which revealed a strong dependence on the MNPs concentration in medium for this kind of particles, suggesting that higher uptake would occur because more particles precipitate onto the cell. Clusters of P-MNPs and T-MNPs would be big enough to lay onto the cells and facilitate the uptake, and small enough to be washed after labeling and grant full MNPs internalization, as it was observed in TEM micrographs. Iron internalization can be related to MRI relaxation values. Relaxation rates R2* of labeled cells did not reflect the differences in terms of MNPs internalization. This fact can be explained due to the high concentration of labeled cells (500 cells/L) for the phantom. In T2* weighted images, the signal decays so fast that the adjustment to an exponential decay is based on 2-4 experimental points (others are zero). In order to perform a better exponential decay adjustment 153 it would be necessary to reduce the cell concentration (therefore the iron concentration in the phantom) or the echo time of the MRI sequence (MGE). Overall, in this chapter we have evaluated our synthesized MNPs, which enable efficient and non-invasive imaging of cells in MRI. D-MNPs, P-MNPs and T-MNPs demonstrated their use for cell labeling at different concentrations and incubation times without a short-term harmful effect for the most of the studied conditions. However, a long-term cellular wellbeing evaluation would be necessary to assess their further use for in vivo cell tracking studies. In this way, as mesenchymal stem cells have demonstrated their potential use in cerebrovascular diseases, also in stroke, we will consider MSC for further in vivo studies of cell tracking. Moreover, in this chapter we had validated our 3 synthesized coated MNPs as suitable MRI contrast agents for these cells. However, among all possibilities, PLL- coated D-MNPs offer several advantages in terms of cell labeling compared to P-MNPs and T-MNPs for further in vivo studies. MNP uptake for D-MNPs is concentration/incubation time dependent and stable in terms of internalization, as it has been shown here and in the previous chapter. It is possible to detect cells labeled with this kind of particles in MRI with relatively low amount of internalized nanoparticles (about 10pg/cell) and from 100g/mL to 500g/mL D-MNPs concentrations, there is not significant influence on proliferation and toxicity. The cytoplasmic distribution in endosomes of regular shape is also crucial because after proliferation, cells will split the internalized MNPs, being daughter cells also detectable by MRI. Another important point to take into account is the ease on the in situ synthesis of D-MNPs, because further in vivo studies will require a high amount of them. 154 6.5. Conclusions In the present chapter, we have evaluated the in vitro short-term biocompatibility of D-MNPs, P-MNPs and T-MNPs by labeling MSCs and C17.2 cells. In general terms, the internalization behavior was the same for both types of cells. Cellular uptake is conditioned by concentration and incubation time for PLL coated D-MNPs meanwhile P-MNPs and T-MNPs internalization depends mostly on the MNPs concentration in cell medium. MNPs influence on the cellular wellbeing is dependent on the internalized iron, reducing proliferation and viability for cells with more than 28 pg/cell and once again, we have demonstrated that the labeling protocol used is highly efficient for full MNPs internalization, providing reliable MRI values for labeled cells. Based on these results, for further in vitro and in vivo studies we will choose dextran-coated MNPs due to its ease for synthesis, stable cell labeling, reduced cellular toxicity, high MRI signal of cells labeled and nanoparticle endosomal distribution observed by TEM in the cellular cytoplasm. However, we have studied short-term biocompatibility, a long-term in vitro study of proliferation, viability and MRI signal evolution is needed for further in vivo studies. 155 7. Long-term study of mesenchymal stem cells labeled with PLL- Dextran coated superparamagnetic nanoparticles 7.1. Hypothesis So far, we have developed 3 different coated superparamagnetic nanoparticles, and we have already demonstrated their short-term biocompatibility for cell tracking. However, a long-term evaluation of labeled cells and how the proliferation rate influences the MRI signal are needed prior to in vivo cell tracking studies. Additionally, the study of magnetic field influence on the cells after magnetic vectorization is needed. In this way and considering previous results, the main goal of this chapter is the long-term biocompatibility study and the evaluation of the MRI signal clearance with the proliferation of rat mesenchymal stem cells labeled with PLL-Dextran coated superparamagnetic nanoparticles for further cell tracking and magnetic vectorization in vivo. 258 Figure 100: Representative fluorescence optical microscopy of rat lung slices where MSCs labeled with CFSE can be observed after jugular delivery. 9.3.2.2. Intra-arterial administration route So far, we have studied the fate of labeled cells injected intraparenchymally and intravenously, and in this section we will track labeled cells which have been injected intraarterially. The timing of administration was the same as intravenously, i.e. 4h after the onset of the stroke in the case of ischemic animals, and MRI was performed 4h after the injection for ischemic and non-ischemic animals. Thus, following intraarterial injection, Figure 101 shows MR T2* weighted images 4h after intra-arterial delivery of (A) MSCs labeled 259 exclusively with CFSE in a healthy animal, (B) MSCs labeled with CFSE and D-MNPs in a healthy animal and (C) MSCs labeled with CFSE and D-MNPs in an animal 8h after the onset of the cerebral ischemia. Differences between groups can be easily observed. Ischemic animal after intra-arterial delivery (C) presents more hypointensities and therefore more MSCs in the brain than the healthy animal (B) after being injected with the same number of cells. This observation could be explained by the disruption of the BBB and the tissue damage after the ischemic process, which could be retaining more cells in the brain. However, the healthy animal (B) with an intact BBB still presents labeled MSCs along the brain; the distribution pattern is more spread than ischemic brain, nevertheless cells can still be detected 4h after the injection. Figure 101 (A) does not display hypointensities, because injected cells were labeled exclusively with CFSE. Histochemistry evaluation of the tissue will elucidate MSCs distribution along these brains to compare with MRI. After MRI scanning, animals were perfused and brain, heart and lungs were processed for further histological analysis. From one extra animal, kidneys, spleen and liver were also processed in order to evaluate cell migration towards those organs. Microscopic results revealed that no cells were found in kidneys, spleen, liver, heart and lungs, however all brain slices showed CFSE labeled cells. MSCs were found distributed mostly along the right hemisphere of the brain, in agreement with MRI T2* weighted images cell distribution (Figure 102). 260 Figure 101: MR T 2 * weighted images of 3 brain slices 4h after intra-arterial delivery of (A) MSCs labeled with CFSE in a healthy animal, (B) MSCs labeled with CFSE and D-MNPs in a healthy animal and (C) MSCs labeled with CFSE and D-MNPs in an animal 8h after the onset of the cerebral ischemia. 261 Figure 102: Fluorescence optical microscopy of brain slices after a intraarterial delivery of MSCs (A) labeled exclusively with CFSE in a healthy animal, (B) labeled with CFSE and D-MNPs in a healthy animal and (C) labeled with CFSE and D-MNPs in a ischemic animal. From fluorescence optical microscopy images (Figure 102) we can observe that CFSE positive cells show tubular-like configuration for all injected animals. This result together with the fact that intra-arterial delivery of MSCs provoke multifocal ischemias (see previous chapter), could mean that injected MSCs are in the vessels, adopting the tubularlike structure observed, and at the same time occluding those vessels. 262 In order to elucidate if the injected cells are in the vessels or they have migrated into the brain parenchyma, we investigated by immunohistochemistry techniques the expression of CD31, which is used primarily to mark endothelial cells in histological tissue sections. The combination of vessels labeled with CD31 and MSCs labeled with CFSE will let us elucidate ex vivo where injected cells are. Besides CD31 and CFSE, cell nucleuses were stained with Hoechst dye (Figure 104 and Figure 105). Figure 103 is a representative brain region image of an animal injected with CFSE and D-MNPs labeled MSCs. Hoechst nucleuses and CD31 vessels staining are observed along all tissue section. After the co-localization (merge) no cells were detected out of vessels in this picture, but also in other regions of the brain. Moreover, in Figure 103, it is easy to note that the coincident vessel and CFSE labeled MSCs have a small and similar caliber. It will be interesting to observe what would happen if vessel diameter is wider than cell size to discriminate if the cells are adhered to the vessel wall or stacked there because of their similar size. 263 Figure 103: Injected MSCs labeled with CFSE are localized in the vessel. CFSE is co-localized with CD31 and Hoechst. (Scale bar 100μm). In Figure 104 a representative brain region where small and big vessels can be examined is shown. We can observe that the big vessel of the image do not present CFSE positive cells, however smaller vessels do display labeled MSCs. Small vessels with positive CFSE staining are zoomed in Figure 105 and once again, the diameter of the vessel is similar to cell dimensions for both cells. Thus, in all brain slices studied, all CFSE positive cells were found in the vessels of small caliber (Figure 104 and Figure 105) and no cells were observed for wider vessels. This finding of MSCs localized in the small vessels could explain why intra-arterial injection of MSCs provokes multifocal ischemias. However at this point we should take into account the observer effect. In science, the term “observer effect” refers to changes that the act of observation will make on a phenomenon being observed. This is often the result of 264 instruments that, by necessity, alter the state of what they measure in some manner. This effect in our experiments is reflected on the tissue processing protocol used for immunohistochemistry analysis. In order to eliminate the blood and therefore minimize background fluorescence, animals were perfunded with PBS and paraformaldehyde 4%. The perfusion is performed by injecting transcardially these substances, washing the animal blood. This process could alter the cell localization because the infusion parameters from the pump will be different from the heart pumping blood parameters, inducing variability on the cell positioning. However, inmunohistochemistry techniques require perfunded organs for an efficient marking and clear discrimination between structures. According to this, what we need is a technique which let us to study the distribution of the cells after being injected with the best possible resolution and without inducing changes in the tissue. The appropriate candidate to study cell positioning after injection minimizing tissue processing is the transmission electron microscopy. 265 Figure 104: Injected MSCs labeled with CFSE can be found in the small vessels. CFSE is co-localizatized with CD31 and Hoechst. (Scale bar 10μm). Figure 105: Magnification of Figure 39 for the two positive CFSE cells [1] and [2]. 266 9.3.3. Transmission electron microscopy (TEM) of brain slices after intra-arterial delivery of D- MNPs labeled MSCs Ultrastructural examination of tissues, cells and microorganisms plays a vital role in diagnostic pathology and biologic research. TEM is used to study the morphology of cells and their organelles,[267] which makes this technique appropriate for studying cell distribution once cells are injected. In order to study if the injected cells are in the vessels or dispersed in the brain parenchyma, injected MSCs were labeled with D-MNPs. In Section 3 we have assessed that MSCs labeled with D-MNPs in vitro can be visualized by TEM, which makes possible to distinguish between non-labeled MSCs and D-MNPs labeled MSCs because of their MNPs endosomes encapsulation. This detection ability together with the fact that the TEM fixation protocol used lets us to keep the brain intact after the animal sacrifice, make this technique suitable for elucidating where the injected cells are. However, results of this section may be taken carefully because so far only one animal was used for TEM examination. One healthy animal was selected for TEM brain examinations 4h after intra-arterial delivery of 1x106 D-MNPs labeled MSCs. Administration procedure was performed as previously reported, and 4h after cell delivery, MR T2* and T2 weighted images were acquired to assess MRI cell distribution. In previous chapters we have shown that most of ischemic animals which have been injected with one million cells died within 24h and healthy animals presented large multifocal ischemias. However, the study of the stem cell administration in a healthy animal was more interesting 267 due to its undamaged vessel structure and intact blood brain barrier which could let us understand the interaction between MSCs and the brain structures under a nonpathologic condition. Besides, we have injected 1x106 labeled cells to make easier the cell detection in the animal brain. Thus 4h after the intra-arterial injection, MRI scanning was performed and T2* weighted images were acquired. All right hemisphere and several regions of the left one displayed a hypointense dotted pattern as it can be observed in Figure 106. After MRI, the animal was sacrificed (not perfunded) and several sections (blue squares in Figure 106) of approximately 3 mm x 3 mm x 3 mm were extracted and immediately immersed in fresh glutaraldehyde for tissue fixation. Electron micrographs of the cortex in a healthy rat injected with 1x106 MSCs labeled with D-MNPs revealed the presence of the injected cells in the brain parenchyma, a representative cell can be observed in Figure 107. This micrograph shows one isolated MSC which had travelled from the vessels to the parenchyma and is in direct contact with the neuropil, where somas and cellular processes can be observed.[267] Not only the background color of the injected MSC is different from the surrounding, the presence of dark endosomes let us its identification as MSC (yellow arrows in Figure 107). Zoom of one endosome is shown in Figure 108, where it is possible to note the dark dotted pattern inside the wrapping endosome membrane. Distribution, size and morphology are similar to those observed in vitro in Chapters 5 and 6. Close to this endosome, 4 mitochondria can be also identified by their cristae, demonstrating the well-being of this mesenchymal stem cell 4h after being injected. 274 Figure 109: Electron micrograph of rat brain cortex 4h after intra-arterial delivery of D-MNPs labeled MSCs. It can be observed a neuron nucleus (Ne), surrounded by neuropil (NP). Magenta arrows point out a vessel with an expansion (blue arrow) due to MSCs presence. 275 Figure 110: Electron micrograph of rat brain cortex 4h after intra-arterial delivery of D-MNPs labeled MSCs, zoom of Figure 109. Red blood corpuscles (RB), Platelets nucleus (P), mesenchymal stem cells (A, B, C) and yellow arrows point out D-MNPs. 276 Figure 111: Electron micrograph of rat brain cortex 4h after intra-arterial delivery of D-MNPs labeled MSCs, zoom of Figure 110. Endosomes of “A” MSC containing D-MNPs (yellow arrows), boundary of cellular membrane of MSC “A” denoted with magenta asterisks, Opened endosomes of a dead MSC “B” (cyan arrows). 277 Figure 112: Electron micrograph of rat brain cortex 4h after intra-arterial delivery of D-MNPs labeled MSCs, zoom of Figure 109. A vessel surrounded by neuropil (NePi) is observed. Fibrin 278 9.4. Conclusions Cell tracking of labeled MSCs following different administration routes is essential for a further understanding of the interaction between injected and host cells. In previous chapters, we had demonstrated the efficient labeling of MSCs with D-MNPs for in vivo MRI cell tracking. In this chapter we have assessed the feasibility of tagging MSCs with D-MNPs and also with CFSE, which let us not only in vivo cell tracking, ex vivo MSCs identification as well. In order to assess the efficiency of this double labeling, intraparenchymal injections for MSCs delivery were performed. Ex vivo histology demonstrated that injected cells were still labeled with D-MNPs 4h after the injection and no particles were found out of the cells assessed by CFSE and Prussian blue staining, therefore MRI in vivo signal corresponds to labeled MSCs and not to free particles. Intraparenchymal injection is a feasible administration route, however it is not the preferred for stem cell delivery due to the high invasiveness of the technique. Less invasive routes for stem cell delivery are intravenous and intraarterial injection. We have studied the fate after the injection of labeled cells through jugular vein and through the internal carotid artery and we have found important differences between them. On one hand, we have assessed by MRI and/or histochemistry techniques that cells injected through jugular vein were found in the lungs and no cells were detected in the brain, for healthy and ischemic animals. On the other hand, following the intraarterial route we have demonstrated that, for ischemic or non-ischemic animals, injected MSCs are found in the brain and no cells can be detected in other organs, opening a big window for further stem cell therapy administration. 279 Because of the blood brain barrier integrity in healthy animals and its disruption in ischemic ones, we also studied the localization of the injected cells by immunohistochemistry techniques after an intra-arterial administration. By CD31 staining and CFSE labeling it was possible to assess that injected cells are found in small vessels along the brain while large ones do not express CFSE positive cells. This selective intra-vessel localization could explain the multifocal ischemias observed in ischemic and non-ischemic intra-arterial injected rats, however because it is necessary to perfuse de animal to perform immunohistochemistry, other technique for corroborating this result was studied. Preliminary results of electron microscopy have shown that MSCs labeled with D-MNPs are able to migrate into brain parenchyma in a healthy animal after an intra-arterial delivery of one million cells. However, TEM images of same animal also revealed vessel occlusions with MSCs and fibrin reaction associated which could explain previous results of multifocal ischemias. In conclusion, for a non invasive brain localization of injected MSCs, intra-arterial injection is the most efficient administration route. On the other hand, MSCs delivered following intra-jugular injection results in cell entrapment in the lungs, and no cells reach the brain. 280 281 Section VI Neuroreparation mediated by mesenchymal stem cell therapy 282 283 10. Neuroreparation study in an animal model of ischemic stroke using mesenchymal stem cells labeled with Dextran-coated superparamagnetic nanoparticles 10.1. Hypothesis In previous chapters we have assessed the fate and the kinetics of labeled mesenchymal stem cells after an intraarterial (ia) and intravenous (iv) administration routes. In this chapter, MSCs were ia and iv delivered during the acute phase in a cerebral ischemia model to investigate whether the cell distribution, and therefore administration route may improve the therapeutic outcome. MRI studies, behavioral tests, histological examinations and blood serum determinations were performed to evaluate the therapeutic benefits of the treatment. 290 for body support. Forelimb contacts while rearing are scored with a total of 10 contacts recorded for each animal. The number of impaired and non-impaired forelimb contacts is calculated as a percentage of total contacts. Cylinder test was performed prior to tMCAO and 7 and 14 days after the ischemia. 10.2.7. Blood serum determinations: Vascular endothelial growth factor (VEGF) and interleunkin-6 (IL-6) Venous blood was collected from the tail vein prior to tMCAO, 4h, 8h, 24h, 1day, 3 days, 7 days and 14 days after the onset of the ischemia. Blood serum was obtained and stored at -80ºC for further VEGF and IL-6 determinations. ELISA assays were performed using Quantikine ELISA Rat VEGF kit (R&D Systems, Inc.) and Quantikine ELISA Rat IL-6 kit (R&D Systems, Inc.) respectively. 10.2.8. Immunohistochemistry processing Rats were anesthetized under sevofluorane (8%), perfused transcardiacally with 4% paraformaldehyde (PFA) in 0.1 M PBS, the brains were removed, postfixed (4% PFA, 4°C, 24 h), cryopreserved (Sucrose 35%), included in OCT and stored at -80ºC. 15 μm thick axial sections were cut by a cryostat and dried in an oven at 38ºC for 4h. Slices were rehydrated in PBS, exposed to Citrate buffer (0.01 M, pH 6.0) (Dako) for 20 min at 99ºC and washed with PBS 1x. 291 10.2.8.1. Immunohistochemistry processing for Ki-67 and CD31 staining dyes Slides were incubated first with Rabbit polyclonal antibody to CD31 (Abcam) (1:25), Mouse monoclonal antibody to ki- 67 (Dako) (1:30), normal horse serum (15%), normal goat serum (15%), Triton X-100 (0.2%) and PBS overnight. The slides were then incubated at room temperature for 1h with biotinylated horse anti-Mouse IgG antibody (Vector) (1:200), DyLight 488 Goat anti-Rabbit IgG antibody (Vector) (1:100), Triton (0.2%) and PBS. After 3 washes with PBST, slides were incubated with dylight 594 streptavidin (Vector) (1:500), Triton (0.2%) and PBS. Then, 10 μL of Hoechst (Invitrogen) in 60 mL of PBS 1x were added to the slides, incubated for 15 min and washed afterwards with PBS. Slides were mounted with fluorescence mounting medium. Photographs were taken using a Leica DMI 6000 B and processed using LAS AF V.1.0.0 software (Leica Microsystems). 10.2.8.2. Immunohistochemistry processing for Ki-67 and doublecortin (DCX) staining dyes Slides were incubated first with Rabbit polyclonal antibody to DCX (1:25), Mouse monoclonal antibody to ki-67 (Dako) (1:30), normal horse serum (15%), normal goat serum (15%), Triton X-100 (0.2%) and PBS overnight. The slides were then incubated at room temperature for 1h with biotinylated horse anti-Mouse IgG antibody (Vector) (1:200), DyLight 488 Goat anti-Rabbit IgG antibody (Vector) (1:100), Triton (0.2%) and PBS. After 3 washes with PBST, slides were incubated with dylight 594 streptavidin (Vector) (1:500), Triton (0.2%) and PBS. Then, 10 μL of Hoechst (Invitrogen) in 60 mL of PBS 1x were added to the slides, incubated for 15 min and washed afterwards with PBS. Slides were mounted with 292 fluorescence mounting medium. Photographs were taken using a Leica DMI 6000 B and processed using LAS AF V.1.0.0 software (Leica Microsystems) 10.2.9. Statistical analysis The statistical analyses of the data were performed using GraphPad Prism 5.01 (GraphPad Software, Inc). For descriptive analyses, means and standard deviations were calculated for the normally distributed continuous variables. For confirmatory analyses one-way ANOVA was performed to calculate overall p-values for the comparison of the different groups for each variable. The global significance level for all statistical test procedures was chosen as =0.05. 10.2.10. Timeline of the study In order to evaluate the therapeutic properties of different administration routes for of MSCs in an animal model of ischemic stroke, 4 groups were studied: o Control 1 (n=6): tMCAO o Control 2 (n=6): tMCAO which were injected i.a. with 300 μL of PBS 4h after the onset of the ischemia. o MSCs ia (n=6): tMCAO which were injected i.a. with 0.25x106 of MSCs labeled with D-MNPs dispersed in 300 μL of PBS 4h after the onset of the ischemia. o MSCs iv (n=6): tMCAO which were injected i.v. with 1x106 of MSCs labeled with D-MNPs dispersed in 300 μL of PBS 4h after the onset of the ischemia. Experimental procedures diagram can be observed in Figure 113. 293 Figure 113: Timeline evaluation of the therapeutic benefits of the different routes for MSCs administration in an animal model of ischemic stroke. Cylinder test was performed prior to tMCAO to assess the locomotor symmetry of the animals and on the days 7th and 14th to assess the symmetry/asymmetry due to the stem cell administration. The infarct size and brain edema follow-up were performed during the occlusion by a MR-DWI sequence and by a T2-weighted images on the 1st, 3rd, 7th and 14th day after the ischemia induction. Moreover, T2*- weighted images were acquired to track the injected cells and to evaluate hemorrhagic lesions 8h, 1 day, 3 days, 7 days and 14 days after the ischemia. We have also included MR angiography to evaluate the occlusion during the tMCAO and to study morphologic changes on the vessels during the following-up of the animals. Venous blood samples were collected during the occlusion, 4h, 8h, 1 day, 3 days, 7 days and 14 days after the onset of the 294 insult for further VEGF and IL-6 determinations. And finally, on the 14th day, animals were transcardiacally perfused and their brains processed as described previously for further histology analysis. 10.3. Results 10.3.1. Cell tracking after intra-arterial administration D-MNPs labeled MSCs were easily detected by MR T2* weighted images 4h, 24h and 3 days after the injection.. Next timepoint is 7 days, and it was not possible to observe labeled cells in all animals scanned (Figure 114). 295 Figure 114: MR T2* weighted images of one ischemic animal injected with 0.25x106 D-MNPs labeled MSCs at different timepoints. 296 10.3.2. Effects on ischemic lesion volume Infarct volume sizes were measured on ADC maps, calculated from MR DWI, acquired during the occlusion (t=0), and on T2 maps, obtained from T2 weighted images at days 1, 3, 7 and 14 after the induction of the ischemia. In Table 11 and Figure 115, infarct volume sizes are expressed in percentage respect the hemisphere volume corrected by edema factor. Control 1 Control 2 MSCs ia MSCs iv Mean SD Mean SD Mean SD Mean SD 0 (tMCAO) 32 6 33 6 35 2 35 7 Day 1 25 5 30 10 32 3 27 7 Day 3 30 5 37 11 36 3 32 10 Day 7 20 8 22 11 22 3 18 11 Day 14 15 5 19 11 19 4 15 10 Table 11: Infarct volumes (% hemisphere) measured from MR ADC and T 2 maps for all groups. Figure 115: Infarct volumes (% hemisphere) measured from MR ADC and T 2 maps for all groups. 297 No infarct volume reduction compared to control was observed 14 days after the onset of the ischemia and mesenchymal stem cell administration in any group (Figure 116). Infarct volumes for intra-arterial administration of PBS or MSCs are slightly increased respect to control or intravenous administration of MSCs, however results are not statistically significant. 298 Figure 116: Representative MRI of each group at different time points. tMCAO corresponds to ADC maps and 24h, 3days, 7days and 14days to T 2 weighted images. 299 10.3.3. Effects on edema formation Brain edema volume sizes were measured on T2 maps, obtained from T2 weighted images at days 1, 3, 7 and 14 after the induction of the ischemia. Edema volumes were calculated as the increase of ipsilateral hemispheric volume respect to contralateral hemispheric volume expressed in percentage (Table 12 and Figure 117). Control 1 Control 2 MSCs ia MSCs iv Mean SD Mean SD Mean SD Mean SD Day 1 7 3 9 5 12 3 7 3 Day 3 9 3 11 5 13 2 10 5 Day 7 2 3 1 2 2 1 1 2 Day 14 -4 1 -4 2 -4 1 -4 2 Table 12: Edema volume formation (% of the increase of ipsilateral hemispheric volume respect to contralateral hemispheric volume). Figure 117: Edema volume (% of the increase of ipsilateral hemispheric volume respect to contralateral hemispheric volume). No edema volume reduction was observed after MSCs intra-arterial or intravenous administration in the acute 306 CD31 combined with Ki-67 did not show an increase in the angiogenesis process in treated animals compared to control 14 days after cerebral ischemia. Furthermore, VEGF and IL-6 were analyzed in blood serum samples. VEGF is an angiogenic and neuroprotective factor that increase axonal outgrowth, block neuronal apoptosis and increase neurogenesis, thus promoting neuroprotection from stroke.[156] In this work we have previously demonstrated that D-MNPs labeled MSCs secrete same levels of VEGF than non-labeled cells. Thus, at this point the study of VEGF levels after MSCs administration was crucial for further understanding of the stem cells therapeutic mechanism. However, blood serum samples showed lower levels than the detectable by ELISA technique. Other studies have determined VEGF levels by measuring brain sections by ELISA instead by blood serum.[276] Besides VEGF, IL-6 was determined in blood serum samples because its demonstrated relationship with pro- and antiimmflamatory effects following focal ischemia. Serum levels measurements of IL-6 were lower than the resolution limit of the ELISA kit. As happened before with VEGF determination, brain sections would report values for measuring using ELISA assays. Collectively, our data indicates that intravenous administration of 1x106 MSCs slightly improve functional outcome after cerebral ischemia, and that the intra-arterial or intravenous delivery of mesenchymal stem cells promotes neurogenesis measured 14 days after the injection. 307 10.5. Conclusions In this study we have compared intra-arterial and intravenous administration of MSCs for the therapy of focal cerebral ischemia. Only cells injected i.a. reached the brain and were detectable for 3 days after the delivery, guaranteeing the cellular proximity to the injury site. Under our experimental conditions and by using dextrancoated superparamagnetic nanoparticles, we have assessed that the local effect of MSCs injected following i.a. route did not improve the outcome after a cerebral ischemia compared to a systemic administration following the i.v. route. 308 309 Section VII Conclusions 310 311 General conclusions 1. Three types of superparamagnetic nanoparticles coated with different polymers, dextran, pluronic F127 and tetronic 908, have been developed and characterized for further biological and biomedical applications as MRI contrast agents. 2. The optimization procedure and in vitro validation of our synthesized superparamagnetic nanoparticles have been performed in mesenchymal stem cells and neural stem cells assessing the biocompatibility of these nanoparticles for cell labeling. 3. We have assessed in vitro the feasibility of label and detect mesenchymal stem cells after proliferation when labeled with dextrancoated superparamagnetic nanoparticles without harmful effects. In addition, it is possible to vectorize these labeled cells with magnetic fields in vitro. 4. We have demonstrated that MSCs labeled with D-MNPs can be visualized in vivo by T2 and T2* weighted images in MRI, and the hypointensities observed correspond to labeled cells and not to free superparamagnetic nanoparticles. 5. A model of intra-arterial administration of stem cells with low associated risks and high brain delivery efficiency has been developed for further cerebral ischemia treatment studies. 6. The doses of mesenchymal stem cells for intra-arterial administration as a potential treatment for ischemic stroke are determined by the presence of multifocal ischemias, being 2.5x105 cells the highest administrable dose. 312 7. We have assessed that the brain targeting with labeled mesenchymal stem cells has not been increased by the use of the magnetic vectorization in vivo. 8. We have observed that after an intravenous administration of labeled mesenchymal stem cells, no cells were found in the brain while cells were detected in the lungs. 9. Dextran-coated superparamagnetic nanoparticles labeling has also allowed the visualization of injected cells for electron microscopy ex vivo analysis. Our first results suggested that some labeled MSCs remained in the vessels and some crossed to the brain parenchyma 4h after an intra-arterial administration. 10. Superparamagnetic nanoparticles are a valuable tool for distinguishing the fate between injected cells following intra-arterial and intravenous administration in an animal model of stroke. 11. Under our experimental conditions, we have not observed that the intra-arterial administration of mesenchymal stem cells presents more advantages than an intravenous delivery in terms of safety and outcome after cerebral ischemia. 313 Conclusiones generales 1. Hemos desarrollado y caracterizado 3 tipos de nanopartículas superparamagnéticas recubiertas de diferentes polímeros, dextran, pluronic F127 y tetronic 908 para uso biomédico como agentes de contraste de resonancia magnética nuclear. 2. Estas nanopartículas se han probado en cultivos celulares de células madre mesenquimales y de células madre neurales, demostrando su alta biocompatibilidad y sus posibles aplicaciones para seguimiento celular. 3. Las células madre mesenquimales marcadas con nanopartículas recubiertas de dextran han sido examinadas tras diversas divisiones celulares asegurando que sus propiedades no se ven afectadas por el marcaje. Además, la vectorización magnética de estas células no sólo es posible sino que no afecta a su viabilidad en estudios in vitro. 4. Diversos experimentos in vivo, han puesto de manifiesto que es posible la visualización de las células marcadas tras ser inyectadas, y que las hipointensidades observadas por resonancia magnética se corresponden a células marcadas y no a nanopartículas libres. 5. Hemos desarrollado un modelo de administración intraarterial de células en animales con isquemia cerebral que minimiza los riesgos asociados a la inyección y que asegura la presencia de células en el cerebro. 6. Las dosis de células madre mesenquimales administradas intra-arterialmente como futura opción terapéutica en la isquemia cerebral, están limitadas por la posibilidad de inducción de isquemias 314 multifocales, siendo 2.5x105 la dosis máxima administrable encontrada. 7. Los experimentos in vivo de vectorización magnética han evidenciado que no es posible incrementar el número de células madre mesenquimales marcadas por la acción de campos magnéticos bajo nuestras condiciones experimentales. 8. Las células madre mesenquimales marcadas y administradas a través de la vena yugular han sido detectadas en los pulmones, sin que se haya observado migración de las mismas al cerebro del animal. 9. El marcaje con nanopartículas superparamagnéticas recubiertas de dextran ha permitido la discriminación entre células inyectadas y células del cerebro del animal mediante técnicas de microscopía electrónica. Nuestros resultados sugieren que las células madre mesenquimales en algunos casos han permanecido en el interior de los vasos, mientras que en otros, han cruzado al parénquima cerebral de un animal sano 4h tras ser administradas. 10. Hemos demostrado que las nanopartículas superparamagnéticas desarrolladas en este trabajo son la herramienta idónea para estudiar la migración celular a nivel cerebral con células administradas tanto i.a. como i.v. 11. Bajo nuestras condiciones experimentales, no hemos observado que la administración intra-arterial suponga una mejora en términos de riesgo beneficio respecto la administración intravenosa de células madre mesenquimales. 315 Section VIII Appendix 322 Figure 125: MR T 2 weighted image of one brain slice of a Wistar rat 7 days after the onset of the cerebral ischemia. Black and white ring inhomogeneities were observed at the last part of the internal carotid artery. T 2 * weighted image and MR angiography of this animal is shown in Figure 126 and Figure 127 respectively. 323 Figure 126: MR T 2 * weighted image of one brain slice of a Wistar rat 7 days after the onset of the cerebral ischemia. White density surrounded by a black ring was observed at the last part of the internal carotid artery. Hemorrhagic event was discarded due to the white density observed. MR angiography of this animal is shown in Figure 127. 324 Figure 127: (A) Brain slice of MR angiography of one brain slice of a Wistar rat 7 days after the onset of the cerebral ischemia where it is possible to note that anterior cerebral artery is expanded. (B) Z-Projection of MR angiography where the aneurism can be observed. 325 Figure 128: MR T 2 * weighted image of one brain slice of a Wistar rat 8h after the onset of the cerebral ischemia and 4h after the intra-arterial administration of 10 6 Mesenchymal stem cells labeled with dextran coated superparamagnetic nanoparticles. 326 Figure 129: TEM micrograph of a brain capillary wall. (From the left to the right) Black shape corresponds to a red blood cell inside the vessel. The wall of the vessel is composed by endothelial cells, basal membrane, smooth muscle cells, basal membrane and astrocytic processes. 327 Figure 130: Two mesenchymal stem cells can be observed inside a capillary transversally observed by TEM micrograph. Their membranes allow its differentiation, and black hypointensities point out the nature of these cells. 328 329 Section IX 330 331 Summary Cerebrovascular diseases are the second cause of death and the first cause of disability in developed countries. However, despite their social and economical importance, therapies for these diseases are quite limited. The most common etiology of stroke, ischemic stroke, consists in the reduction of blood flow in a brain area after the occlusion of a cerebral artery. This process triggers a series of events at molecular and cellular levels that leads to energy depletion in the affected area, yielding on a quick process of cell death by necrosis. This area is called the infarct core. The surrounding zone, usually called the ischemic penumbra but probably better defined as the peri-infarct area, it is constituted by hypo-perfused tissue. Here, blood flow is reduced to maintain electric activity, but cerebral tissue is still viable. This area is subjected to a wave of deleterious metabolic processes that propagate from the ischemic core to the neighboring tissue, including excitotoxicity, spreading depression, oxidative stress, and inflammatory response, all of them leading to the expansion of the ischemic core into the peri-infarct region, and subsequent worsening of the clinical outcome. So far, pharmacological thrombolysis or mechanical reperfusion are the strategies that report higher benefits in the acute phase for the patients, in terms of neurological outcome, being the recombinant tissue plasminogen activator (rt-PA) the most common thrombolytic agent. However, only 3-7% of stroke patients are currently treated by these procedures due to the narrow therapeutic window (4.5h after the onset of the symptoms) and due to the risks associated to this pharmacological therapy. One of the most important associated risks is the high rate of 338 them transfect agents (TA) as poly-l-lysine (PLL) are the most used. Preliminary experiments using dextran-, Pluronic F127- and Tetronic 908- coated MNPs have shown that only D- MNPs needed PLL to tag cells, while P-MNPs and T-MNPs were internalized without the use of TA. Experimental data revealed that 1.5 μg/mL of PLL combined with D-MNPs ensured a full internalization of MNPs without harmful effects for the cells. In order to study the biocompatibility and elucidate which was the best superparamagnetic nanoparticle for cell tagging, D-MNPs combined with PLL, P-MNPs and T-MNPs were evaluated in vitro by using rat MSCs and a mice multipotent neural progenitor cell line C17.2. Several parameters as incubation time and MNPs concentration were also studied. The observed uptake mechanism for D-MNPs was different from P-MNPs and T-MNPs; the first one was conditioned by the concentration and incubation time, while for P-MNPs and T-MNPs the internalization depended mostly on the MNP concentration in the cell medium. MNPs influence on the cellular wellbeing depended on the internalized iron; thus proliferation and viability was reduced for those cells with more than 28 pg of iron/cell. Moreover, the combination of MRI and TEM techniques showed that MNPs were fully internalized, and MR T2 and T2* weighted images showed that labeled cells could be properly detected in MRI, providing promising results for further in vivo applications. Overall, considering the MNPs distribution in the cell, amount of internalized iron, proliferation and viability, D-MNPs were pointed as the best candidates for cell labeling. However, the biocompatible study of D-MNPs with different types of cells was performed just after labeling. Prior to go into in vivo studies of cell tracking, a long term evaluation of 339 labeled cells and how the cellular proliferation rate influenced the MR signal was evaluated. Because of the multi-potential abilities of MSCs and with the aim of a future stem cell therapy in an animal model of ischemic stroke, the long term biocompatibility study and the evaluation of the MRI signal clearance with the proliferation was performed in MSCs. Experimental results showed that labeled cells did not differ from control cells in terms of proliferation rate and viability after 5 days, as well as the maintenance of CD45-, CD90+ and CD73+ cellular phenotype. Angiogenesis capacity was not altered by the presence of internalized D-MNPs, as well as vascular endothelial growth factor (VEGF) release measurements did not exhibit differences between control and labeled cells. Under the labeling conditions studied and for cell tracking applications, MRI can easily detect labeled cells in a phantom 3 days after being tagged. From then, MRI signal is affected by the clearance of the internalized D-MNPs due to proliferation, being only possible to detect MNPs loaded cells by ICP techniques. In addition to MRI cell detection, dextran coated superparamagnetic nanoparticles could be combined with magnetic fields for enhancing targeted delivery of stem cells. In this work the in vitro efficacy of cellular magnetic vectorization in a liquid environment without harmful effects for future targeted delivery of stem cells has been demonstrated. However, in a non-liquid environment the use of external magnetic fields for labeled cell vectorization showed no cellular migration. Overall, the in vitro validation of synthesized superparamagnetic nanoparticles suggest that MSCs loaded with D-MNPs were a promising tool for cell tracking 340 and cellular magnetic vectorization in vivo without cellular harmful effects. In order to elucidate if labeled cells could be detected in vivo and to corroborate that MRI signal corresponded to labeled cells and not to free particles, the intraparenchymal injection was the best approach. Moreover, in addition to MNPs labeling, MSCs can be tagged also with CFSE, which let us not only in vivo cell tracking, ex vivo MSCs identification as well. In order to assess the efficiency of this double labeling, intraparenchymal injections for MSCs delivery were performed. Histology demonstrated that injected cells were still labeled with D-MNPs 4h after the injection and no particles were found outside of the cells assessed by CFSE and Prussian blue staining. Therefore, MRI in vivo signal corresponded to labeled MSCs and not to free particles. Animal models of cerebral ischemia are designed to generate reproducible infarcts with a minimum of surgical manipulation; however therapeutic approaches have to be also optimized for each disease. Thus, optimal conditions for applying stem cell therapies for cerebral ischemia are still under discussion, fundamental questions related to cell type, characterization and dosage, administration route, therapeutic timing, toxicity, or the relationship between biodistribution, fate and outcome are still on the bench. Administration routes of labeled cells was a crucial parameter to take into account not only for cell tracking, but for stem cell therapy as well. For intra-arterial delivery route, different combinations of animal models of ischemic stroke and intra-arterial administration routes of MSCs were studied. In order to elucidate which model was the most efficient and the 341 safest for an intra-arterial injection, MSCs labeled with D- MNPs were used. The optimization of the method for transient middle cerebral artery occlusion (tMCAO) and intra-arterial delivery of tagged cells was performed by using D-MNPs labeled MSCs. Thus, for an efficient intraarterial delivery, after tMCAO surgery the common carotid artery must be perfused and also during and after the cell administration (that is performed through the external carotid artery) so the heart pumping would help on carrying the delivered cells to the brain territory. These intra-arterial injected cells could be detected in the ischemic brain for more than 4h after injection, and their presence was not increased by the use of external magnetic fields, as magnetic vectorization. Moreover, the monitoring of cerebral blood flow (CBF) by transcranial laser Doppler was not enough to assess the safety of the intraarterial delivery, and healthy animals injected with 2.5x105 MSCs showed multifocal ischemic lesions 24h after the stem cell administration, evidencing that injected cells could be inducing focal ischemias. Tracking of administered cells is important to understand the relationship between possible therapeutic mechanisms and cell localizations. Intraparenchymal injection is a feasible administration route, however it is not the preferred for stem cell delivery due to the high invasiveness of the technique. Less invasive routes for stem cell delivery are intravenous and intra-arterial injection. We have studied the fate after the injection of labeled cells through jugular vein and through the internal carotid artery and we have found important differences between them. On one hand, we have assessed by MRI and/or histochemistry techniques that cells injected through jugular vein were found in the lungs and no cells were detected in the brain, both in 342 healthy and ischemic animals. On the other hand, following the intra-arterial route we have demonstrated that, for ischemic or non-ischemic animals, injected MSCs were found in the brain and no cells can be detected in other organs, opening a big window for further stem cell therapy administration. Because of the blood brain barrier integrity in healthy animals and its disruption in ischemic ones, we also studied the localization of the injected cells by immunohistochemistry techniques after an intra-arterial administration. By CD31 staining and CFSE labeling it was possible to assess that injected cells were found in small vessels along the brain while large ones do not express CFSE positive cells. This selective intra-vessel localization could explain the multifocal ischemias observed in ischemic and non-ischemic intra-arterial injected rats, however because it was necessary to perfuse de animal to perform immunohistochemistry, other technique for corroborating this result was studied. Preliminary results of electron microscopy have shown that MSCs labeled with D-MNPs were able to migrate into brain parenchyma in a healthy animal after an intra-arterial delivery of one million cells. However, TEM images of same animal also revealed vessel occlusions with MSCs and fibrin reaction associated which could explain previous results of multifocal ischemias. Nevertheless, more studies of intraarterial administration of D-MNPs labeled MSCs at different time points are needed for a better understanding of these mechanisms. Thus, intra-arterial injection was found as the most efficient administration route for brain targeting, and also MSCs 343 delivered following intra-jugular injection resulted in cell entrapment in the lungs, with no cells reaching the brain. Finally, MSCs were intra-arterially and intravenously delivered during the acute phase in a cerebral ischemia model to investigate whether the cell distribution, and therefore administration route may improve the therapeutic outcome. MRI studies, behavioral tests, histological examinations and blood serum determinations were performed to evaluate the therapeutic benefits of the treatment. Experimental results showed that only cells injected intra-arterially reached the brain and were detectable for 3 days after the delivery, guaranteeing the cellular proximity to the injury site. Under our experimental conditions and by using D-MNPs, we have assessed that the local effect of MSCs injected following intra-arterial route did not improve the outcome after a cerebral ischemia compared to a systemic administration following the intravenous route. In summary, we have synthesized biocompatible superparamagnetic nanoparticles for in vivo cell tracking. Superparamagnetic nanoparticles-tagged mesenchymal stem cells can be monitored in vivo by MRI without harmful effects, providing information of the cellular fate after different routes of administration. Moreover, the nanoparticle labeling allowed us the study of therapeutic effects in an animal model of ischemic stroke based on the localization of the cells after delivery. 344 345 Resumen Las enfermedades cerebrovsculares son la segunda causa de muerte y la primera de discapacidad en países desarrollados. Sin embardo, a pesar de su importancia socio-económica, las terapias disponibles son limitadas. La etiología más frecuente del ictus, el ictus isquémico, consiste en la reducción localizada del flujo cerebral tras la oclusión de una arteria cerebral. Este proceso desencadena una serie de eventos a nivel molecular y celular que culminan con un fallo energético en el área afectada, seguida de un proceso de muerte celular por necrosis. Esta área recibe el nombre de core isquémico. La región circundante, llamada penumbra isquémica pero normalmente definida como región peri-infarto, está constituida por tejido hipoperfundido. En esa región el flujo cerebral se encuentra reducido para mantener la actividad eléctrica, pero el tejido es todavía viable si se establece el flujo cerebral. Esta área se encuentra además sometida a una serie de procesos metabólicos que se propagan desde el core isquémico al tejido circundante, incluyendo procesos de exitotoxicidad, estrés oxidativo, respuesta inflamatoria, condicionando el pronóstico clínico. La trombolisis farmacológica o la reperfusión mecánica son las estrategias que reportan mejores beneficios a los pacientes, en términos de pronóstico neurológico, siendo el activador tisular del plasminógeno recombinante (rt-PA) el más común de todos ellos. Sin embargo, solo entre el 3 y el 7% de los pacientes con un ictus pueden beneficiarse de esta terapia debido principalmente la estrecha ventana terapéutica de aplicación (4.5h tras el comienzo de los síntomas) y también debido a los riesgos asociados a esta 346 terapia farmacológica. Uno de los riesgos más relevantes es el alto grado de transformación hemorrágica inducida por la reperfusión de las arterias. Así, el mantenimiento de la barrera hematoencefálica, la homeostasis de los fluidos, y el bloqueo de los radicales libres tras el tratamiento de reperfusión constituyen algunas de las principales estrategias neuroprotectoras durante la fase aguda del ictus. La neuroprotección es un término que incluye a todas las terapias dirigidas a reducir la muerte celular tras un proceso isquémico sin ejercer ningún tipo de influencia en la reperfusión del tejido durante la fase aguda del ictus. Hasta ahora, algunos fármacos han mostrado resultados prometedores en estudios experimentales, pero una vez trasladados a ensayos clínicos no han demostrado su eficacia. Por otro lado, las estrategias neurorreparadoras cuentan con una ventana terapéutica mayor que la fase aguda del ictus. Estas estrategias incluyen la restauración de la función cerebral, ya sea bien por regeneración del tejido cerebral dañado (neurorregeneración) o bien por la creación de nuevas conexiones neuronales o sinapsis (plasticidad cerebral). De esta forma, las terapias de neurorreparación no están exclusivamente dirigidas a neuronas, ya que la restauración de la unidad neurovascular incluye también procedimientos enfocados a la potenciación del a angiogénesis y sinaptogénesis. Los tratamientos neurorreparativos incluyen pues la formación de nuevas neuronas, sinapsis, y vasos, ya sea por la aplicación de nuevas terapias farmacológicas o bien por tratamiento con células madre. 347 La terapia celular ha surgido como una estrategia prometedora frente a tratamientos farmacológicos convencionales debido a sus múltiples y potenciales mecanismos de acción, entre los cuales se han destacado la integración de las células administradas en los tejidos, procesos de inmunomodulación o secreción de factores de crecimiento. Sin embargo, otros aspectos como el tipo celular, dosis, ventana terapéutica, toxicidad o la relación entre biodistribución y pronóstico deben ser todavía estudiadas en profundidad. En cuanto a la estirpe celular, varios tipos celulares han demostrado efectos beneficiosos en estudios preclínicos de isquemia cerebral, incluyendo células madre embrionarias, células madre neurales, células madre pluripotentes inducidas y células madre mesenquimales (MSCs) entre otras. En numerosos estudios también se han evaluado diferentes tiempos y rutas de administración con prometedores resultados. Las MSCs han sido descritas como son unas de las candidatas favoritas para terapias regenerativas no sólo en el campo de la isquemia cerebral. Esto se debe no sólo a que son multipotentes, sino también a sus capacidades inmunomoduladoras y de secreción de factores de crecimiento. En los últimos años se han descrito varios mecanismos de acción de esas células en el cerebro isquémico, como su diferenciación a células de estirpe neural, inducción de la neurogénesis, angiogénesis, sinaptogénesis, activación de procesos endógenos, regulación del flujo sanguíneo cerebral, y otros mecanismos neuroprotectores como la reducción de la apoptosis, de la inflamación o incremento de la supervivencia astrocitaria, entre otros. Además, algunos estudios preclínicos en roedores han encontrado recuperación funcional y reducción del volumen de infarto siguiendo diversas rutas de administración de éstas células. 354 proporcione información de la evolución de la señal de RMN con la proliferación celular. Considerando una futura terapia celular en un modelo animal de isquemia cerebral, y debido a sus múltiples capacidades regenerativas, el estudio in vitro de biocompatibilidad celular y de evolución de señal de RMN a largo plazo se llevó a cabo en MSCs. Su seguimiento a largo plazo (5 días tras la internalización de las MNPs) fue condicionado por la señal en RMN ya que el fin último de las D-MNPs es seguimiento celular por RMN. Los resultados experimentales mostraron iguales valores de proliferación y viabilidad de las células marcadas con D- MNPs y las no marcadas tras 5 días de seguimiento, al igual que se mantuvo inalterada la expresión de marcadores de MSCs como CD45-, CD90+ and CD73+. La capacidad angiogénica no se encontró alterada en aquellas células que estaban marcadas, al igual que la secreción de vascular endothelial growth factor (VEGF). Considerando las condiciones de marcaje estudiadas y para aplicaciones de seguimiento celular, fue posible detectar por RMN en un phantom células que han sido marcadas hasta 3 días tras el marcaje. A partir del tercer día no fue posible detectar por RMN estas células ya que la cantidad de MNPs/célula se ha reducido debido a la proliferación celular, aunque los valores de ICP demostraron que todavía había nanopartículas internalizadas. Además de la detección por RMN, las D-MNPs combinadas con campos magnéticos externos podrían incrementar la presencia de MSCs en ciertas regiones de interés, como podría ser el cerebro en patología isquémica cerebral. Dos aproximaciones de vectorización magnética de MSCs marcadas con D-MNPs fueron estudiadas en este trabajo; 355 por un lado se estudió la vectorización en un entorno líquido, demostrando que esta aproximación es no sólo posible sino que no afecta a la viabilidad celular, y por otro lado, la vectorización magnética fue también evaluada en un entorno tridimensional, usando Matrigel como soporte. En este último caso no se observó migración celular. En general, la validación in vitro llevada a cabo sugirió que las MSCs marcadas con D-MNPs son una herramienta prometedora en el campo del seguimiento celular por RMN y para la vectorización magnética in vivo sin daños asociados a las células marcadas. Para corroborar que las células administradas pueden ser detectadas in vivo, y que la señal de RMN se corresponde a células marcadas y no a nanopartículas libres, se llevaron a cabo inyecciones intraparenquimales en rata sana. Las células administradas se marcaron con D-MNPs y también con CFSE, una sonda fluorescente que permitirá la identificación histológica de las células inyectadas. Así, los estudios histológicos con tinción de Prussian Blue y CFSE demostraron que las nanopartículas seguían internalizadas en las células, y que no se observaron nanopartículas libres, con lo que la señal de RMN observada in vivo se correspondió exclusivamente a células marcadas. Para llevar a cabo estudios fiables preclínicos de ictus isquémico se necesita un modelo animal de isquemia cerebral que proporcione lesiones isquémicas reproducibles y cuya cirugía sea mínimamente invasiva. Sin embargo también es necesaria la optimización de las estrategias terapéuticas para cada patología. Así, las condiciones óptimas para terapia celular en la isquemia cerebral todavía no han sido establecidas. La ruta de administración es un parámetro crucial, no sólo para 356 estudiar la biodistribución celular, sino también para el estudio de los mecanismos involucrados en la terapia celular. La administración intra-arterial está íntimamente relacionada con el modelo animal de ictus isquémico. La optimización del modelo de isquemia cerebral y de inyección celular se llevó a cabo atendiendo a la seguridad y a la eficiencia de la administración y mediante el uso de MSCs marcadas con D-MNPs. Así, y tras varias configuraciones de las cirugías se observó que tras la oclusión transitoria de la arteria cerebral media (tMCAO), es necesario reperfundir la arteria común carótida, y que ésta también ha de permanecer abierta durante la administración de las MSCs (que se hace a través de la arteria carótida externa), para que la sangre procedente del corazón colabore en el transporte estas células al cerebro. Las células administradas intra-arterialmente pudieron ser detectadas en el cerebro isquémico durante más de 4h tras la inyección. Además, la vectorización magnetica llevada a cabo con un imán de 1T demostró no incrementar la presencia de MSCs marcadas con D-MNPs en el cerebro, desde el punto de vista macroscópico de imagen por RMN. Lo que sí se observó fue que la administración intra-arterial de 2.5x105 MSCs en animales sanos cuyo flujo cerebral fue monitorizado con laser doppler transcraneal produjo isquemias multifocales 24h tras la administración, evidenciando que las mediciones por laser Doppler no son suficientes para asegurar el mantenimiento del flujo cerebral. El seguimiento celular es crucial para estudiar la relación entre localización celular y posibles mecanismos 357 terapéuticos. Las inyecciones intraparenquimales han sido utilizadas al comienzo de la aplicación de la terapia celular, sin embargo su carácter altamente invasivo hace que poco a poco haya caído en desuso. La administración intra-arterial e intravenosa son técnicas menos invasivas y por tanto más atractivas para terapia celular, y en este trabajo se ha estudiado la biodistribución de MSCs marcadas con D-MNPs siguiendo estas vías por RMN y/o técnicas histoquímicas. Así, se observó que las células administradas en animales sanos e isquémicos a través de la vena yugular se encontraban en los pulmones y no en el cerebro, mientras que tras una administración intra-arterial sólo se detectaron células en el cerebro y no en otros órganos. Estos resultados han abierto una vía de estudio que relacione la función terapéutica frente a localización de las células madre administradas. Además, se ha estudiado histológicamente mediante el marcaje de vasos con CD31 y el marcaje de CFSE de las MSCs administradas, la localización de las células inyectadas intra-arterialmente debido a las diferencias en integridad de la barrera hematoencefálica de animales sanos e isquémicos. Estos estudios mostraron que las células administradas se encontraban en pequeños vasos, mientras que vasos de mayor calibre no mostraron células. Esta presencia selectiva de pequeños vasos podría explicar las isquemias multifocales observadas en animales sanos e isquémicos tras la administración celular, sin embargo dado que para llevar a cabo estos estudios histológicos fue preciso perfundir el animal, sería adecuado utilizar otra técnica que confirmase estos resultados. Los primeros resultados de microscopía electrónica muestran que las MSCs marcadas con D-MNPs pueden migrar al parénquima cerebral en un animal sano tras una 358 administración intra-arterial de 106 de células. Sin embargo, imágenes de TEM del mismo animal en otra región cerebral evidenciaron una oclusión en un capilar producida por las MSCs administradas y donde también se observó una reacción de fibrina, lo cual explicaría así las isquemias multifocales observadas en este trabajo con otras técnicas. Estos resultados de TEM no son concluyentes, pues se necesitaría evaluar la migración en más animales y estudiar el perfil temporal de migración y/o oclusión tras la administración. Finalmente, se estudió las capacidades terapéuticas de las MSCs administradas intravenosas e intra-arteriales, y por tanto, la influencia de la distribución celular en la mejora tras la isquemia cerebral en un modelo animal. Para determinar los efectos beneficiosos de esta terapia se llevaron a cabo estudios de RMN, test de comportamiento, análisis histológico, y determinaciones en suero sanguíneo. Los resultados experimentales de seguimiento celular mostraron que sólo las células inyectadas intraarterialmente fueron detectadas en el cerebro del animal durante los 3 días siguientes a su administración. Además, bajo las condiciones experimentales utilizadas en este estudio y mediante el marcaje con D-MNPs, los animales tratados con MSCs administradas intra-arterialmente no mejoraron significativamente respecto de los tratados con MSCs administradas intravenosamente. En resumen, hemos sintetizado nanopartículas superparamagneticas biocompatibles para seguimiento celular in vivo. Las células madre mesenquimales marcadas con estas nanopartículas superparamagneticas pueden ser monitorizadas in vivo por resonancia magnética nuclear sin afectar a su viabilidad, proporcionando información de la biodistribución celular tras ser administradas siguiendo 359 distintas vías de administración. Además, el marcaje celular nos ha permitido el estudio de los efectos terapéuticos en un modelo animal de ictus isquémico basándonos en la localización de las células tras la administración. 360 361 Section X Bibliography 362 1. (1988) The World Health Organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. WHO MONICA Project Principal Investigators. J Clin Epidemiol 41: 105-114. 2. Rodríguez-Yáñez M, Fernández Maiztegui C, Pérez- Concha T, Castillo J, Zarranz J (2008) Enfermedades vasculares cerebrales. Neurología. Cuarta Edición ed. Madrid, España: Elsevier España. pp. 337-411. 3. Arboix A, Díaz J, Pérez-Sempere A, Álvarez-Sabín J (2002) Ictus: Tipos etiológicos y criterios diagnósticos. Rev Neurol 17: 3-12. 4. Arias-Rivas S, Vivancos-Mora J, Castillo J (2012) Epidemiología de los subtipos de ictus en pacientes hospitalizados atendidos por neurólogos: resultados del registro EPICES (I). Rev Neurol 54: 385-393. 5. Del Zoppo GJ, Saver JL, Jauch EC, Adams HP, Jr., American Heart Association Stroke C (2009) Expansion of the time window for treatment of acute ischemic stroke with intravenous tissue plasminogen activator: a science advisory from the American Heart Association/American Stroke Association. Stroke 40: 2945-2948. 6. Adams HP, Jr., Bendixen BH, Kappelle LJ, Biller J, Love BB, et al. (1993) Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 24: 35- 41. 7. Feinberg WM, Albers GW, Barnett HJ, Biller J, Caplan LR, et al. (1994) Guidelines for the management of transient ischemic attacks. From the Ad Hoc Committee on Guidelines for the Management of Transient Ischemic Attacks of the Stroke Council of the American Heart Association. Circulation 89: 2950-2965. 363 8. Castillo J (2000) Fisiopatología de la isquemia cerebral. Rev Neurol 30: 459-464. 9. Back T (1998) Pathophysiology of the ischemic penumbra--revision of a concept. Cell Mol Neurobiol 18: 621-638. 10. Astrup J, Symon L, Branston NM, Lassen NA (1977) Cortical evoked potential and extracellular K+ and H+ at critical levels of brain ischemia. Stroke 8: 51-57. 11. Hansen AJ (1985) Effect of anoxia on ion distribution in the brain. Physiol Rev 65: 101-148. 12. Blank WF, Jr., Kirshner HS (1977) The kinetics of extracellular potassium changes during hypoxia and anoxia in the cat cerebral cortex. Brain Res 123: 113-124. 13. Choi DW (1987) Ionic dependence of glutamate neurotoxicity. J Neurosci 7: 369-379. 14. Choi DW, Rothman SM (1990) The role of glutamate neurotoxicity in hypoxic-ischemic neuronal death. Annu Rev Neurosci 13: 171-182. 15. White BC, Sullivan JM, DeGracia DJ, O'Neil BJ, Neumar RW, et al. (2000) Brain ischemia and reperfusion: molecular mechanisms of neuronal injury. J Neurol Sci 179: 1-33. 16. Choi DW (1992) Excitotoxic cell death. J Neurobiol 23: 1261-1276. 17. Schiene K, Bruehl C, Zilles K, Qu M, Hagemann G, et al. (1996) Neuronal hyperexcitability and reduction of GABAA- receptor expression in the surround of cerebral photothrombosis. J Cereb Blood Flow Metab 16: 906-914. 18. Banasiak KJ, Xia Y, Haddad GG (2000) Mechanisms underlying hypoxia-induced neuronal apoptosis. Prog Neurobiol 62: 215-249. 370 72. Teng H, Zhang ZG, Wang L, Zhang RL, Zhang L, et al. (2008) Coupling of angiogenesis and neurogenesis in cultured endothelial cells and neural progenitor cells after stroke. J Cereb Blood Flow Metab 28: 764-771. 73. Chopp M, Zhang ZG, Jiang Q (2007) Neurogenesis, angiogenesis, and MRI indices of functional recovery from stroke. Stroke 38: 827-831. 74. Jiang Q, Zhang ZG, Ding GL, Zhang L, Ewing JR, et al. (2005) Investigation of neural progenitor cell induced angiogenesis after embolic stroke in rat using MRI. Neuroimage 28: 698-707. 75. Parr AM, Tator CH, Keating A (2007) Bone marrowderived mesenchymal stromal cells for the repair of central nervous system injury. Bone Marrow Transplant 40: 609-619. 76. Hurty A (2009) The other brain cells. In: Spector R, editor. Stanford Medicine Magazine. 77. Gage FH, Ray J, Fisher LJ (1995) Isolation, characterization, and use of stem cells from the CNS. Annu Rev Neurosci 18: 159-192. 78. Sobrino T, Campos F, Castillo J (2015) The Role of Endothelial Progenitor Cells in Stroke. In: Zhao L-R, Zhang JH, editors. Cellular Therapy for Stroke and CNS Injuries: Springer International Publishing. pp. 109-123. 79. Takagi Y, Nishimura M, Morizane A, Takahashi J, Nozaki K, et al. (2005) Survival and differentiation of neural progenitor cells derived from embryonic stem cells and transplanted into ischemic brain. J Neurosurg 103: 304-310. 80. Buhnemann C, Scholz A, Bernreuther C, Malik CY, Braun H, et al. (2006) Neuronal differentiation of transplanted embryonic stem cell-derived precursors in stroke lesions of adult rats. Brain 129: 3238-3248. 371 81. Shen LH, Li Y, Chen J, Cui Y, Zhang C, et al. (2007) Oneyear follow-up after bone marrow stromal cell treatment in middle-aged female rats with stroke. Stroke 38: 2150-2156. 82. Chen J, Li Y, Wang L, Zhang Z, Lu D, et al. (2001) Therapeutic benefit of intravenous administration of bone marrow stromal cells after cerebral ischemia in rats. Stroke 32: 1005-1011. 83. Mehler MF, Rozental R, Dougherty M, Spray DC, Kessler JA (1993) Cytokine regulation of neuronal differentiation of hippocampal progenitor cells. Nature 362: 62-65. 84. Wei X, Du Z, Zhao L, Feng D, Wei G, et al. (2009) IFATS collection: The conditioned media of adipose stromal cells protect against hypoxia-ischemia-induced brain damage in neonatal rats. Stem Cells 27: 478-488. 85. Borlongan CV, Kaneko Y, Maki M, Yu SJ, Ali M, et al. (2010) Menstrual blood cells display stem cell-like phenotypic markers and exert neuroprotection following transplantation in experimental stroke. Stem Cells Dev 19: 439-452. 86. Martino G, Pluchino S (2006) The therapeutic potential of neural stem cells. Nat Rev Neurosci 7: 395-406. 87. Sheikh AM, Nagai A, Wakabayashi K, Narantuya D, Kobayashi S, et al. (2011) Mesenchymal stem cell transplantation modulates neuroinflammation in focal cerebral ischemia: contribution of fractalkine and IL-5. Neurobiol Dis 41: 717-724. 88. Bacigaluppi M, Pluchino S, Peruzzotti-Jametti L, Kilic E, Kilic U, et al. (2009) Delayed post-ischaemic neuroprotection following systemic neural stem cell transplantation involves multiple mechanisms. Brain 132: 2239-2251. 372 89. Hirko AC, Dallasen R, Jomura S, Xu Y (2008) Modulation of inflammatory responses after global ischemia by transplanted umbilical cord matrix stem cells. Stem Cells 26: 2893-2901. 90. Arvidsson A, Collin T, Kirik D, Kokaia Z, Lindvall O (2002) Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat Med 8: 963-970. 91. Mine Y, Tatarishvili J, Oki K, Monni E, Kokaia Z, et al. (2013) Grafted human neural stem cells enhance several steps of endogenous neurogenesis and improve behavioral recovery after middle cerebral artery occlusion in rats. Neurobiol Dis 52: 191-203. 92. Goldman SA, Chen Z (2011) Perivascular instruction of cell genesis and fate in the adult brain. Nat Neurosci 14: 1382-1389. 93. Chopp M, Li Y, Zhang ZG (2009) Mechanisms underlying improved recovery of neurological function after stroke in the rodent after treatment with neurorestorative cell-based therapies. Stroke 40: S143-145. 94. Lee SH, Lumelsky N, Studer L, Auerbach JM, McKay RD (2000) Efficient generation of midbrain and hindbrain neurons from mouse embryonic stem cells. Nat Biotechnol 18: 675-679. 95. Nagai N, Kawao N, Okada K, Okumoto K, Teramura T, et al. (2010) Systemic transplantation of embryonic stem cells accelerates brain lesion decrease and angiogenesis. Neuroreport 21: 575-579. 96. Tae-Hoon L, Yoon-Seok L (2012) Transplantation of mouse embryonic stem cell after middle cerebral artery occlusion. Acta Cir Bras 27: 333-339. 97. Seyed Jafari SS, Ali Aghaei A, Asadi-Shekaari M, Nematollahi-Mahani SN, Sheibani V (2011) Investigating the 373 effects of adult neural stem cell transplantation by lumbar puncture in transient cerebral ischemia. Neurosci Lett 495: 1-5. 98. Guzman R, De Los Angeles A, Cheshier S, Choi R, Hoang S, et al. (2008) Intracarotid injection of fluorescence activated cell-sorted CD49d-positive neural stem cells improves targeted cell delivery and behavior after stroke in a mouse stroke model. Stroke 39: 1300-1306. 99. Chen SJ, Chang CM, Tsai SK, Chang YL, Chou SJ, et al. (2010) Functional improvement of focal cerebral ischemia injury by subdural transplantation of induced pluripotent stem cells with fibrin glue. Stem Cells Dev 19: 1757-1767. 100. Wang J, Chao F, Han F, Zhang G, Xi Q, et al. (2013) PET demonstrates functional recovery after transplantation of induced pluripotent stem cells in a rat model of cerebral ischemic injury. J Nucl Med 54: 785-792. 101. Chen J, Sanberg PR, Li Y, Wang L, Lu M, et al. (2001) Intravenous administration of human umbilical cord blood reduces behavioral deficits after stroke in rats. Stroke 32: 2682-2688. 102. Kozlowska H, Jablonka J, Janowski M, Jurga M, Kossut M, et al. (2007) Transplantation of a novel human cord blood-derived neural-like stem cell line in a rat model of cortical infarct. Stem Cells Dev 16: 481-488. 103. Li Y, Chen J, Wang L, Lu M, Chopp M (2001) Treatment of stroke in rat with intracarotid administration of marrow stromal cells. Neurology 56: 1666-1672. 104. Kurozumi K, Nakamura K, Tamiya T, Kawano Y, Ishii K, et al. (2005) Mesenchymal stem cells that produce neurotrophic factors reduce ischemic damage in the rat middle cerebral artery occlusion model. Mol Ther 11: 96-104. 374 105. Oki K, Tatarishvili J, Wood J, Koch P, Wattananit S, et al. (2012) Human-induced pluripotent stem cells form functional neurons and improve recovery after grafting in stroke-damaged brain. Stem Cells 30: 1120-1133. 106. Robinton DA, Daley GQ (2012) The promise of induced pluripotent stem cells in research and therapy. Nature 481: 295-305. 107. Shinozuka K, Dailey T, Tajiri N, Ishikawa H, Kaneko Y, et al. (2013) Stem Cell Transplantation for Neuroprotection in Stroke. Brain Sci 3: 239-261. 108. Lee JM, Jung J, Lee HJ, Jeong SJ, Cho KJ, et al. (2012) Comparison of immunomodulatory effects of placenta mesenchymal stem cells with bone marrow and adipose mesenchymal stem cells. Int Immunopharmacol 13: 219-224. 109. Liu R, Yuan H, Yuan F, Yang SH (2012) Neuroprotection targeting ischemic penumbra and beyond for the treatment of ischemic stroke. Neurol Res 34: 331-337. 110. Fazekasova H, Lechler R, Langford K, Lombardi G (2011) Placenta-derived MSCs are partially immunogenic and less immunomodulatory than bone marrow-derived MSCs. J Tissue Eng Regen Med 5: 684-694. 111. Netto CA, Hodges H, Sinden JD, Le Peillet E, Kershaw T, et al. (1993) Effects of fetal hippocampal field grafts on ischaemic-induced deficits in spatial navigation in the water maze. Neuroscience 54: 69-92. 112. Darsalia V, Kallur T, Kokaia Z (2007) Survival, migration and neuronal differentiation of human fetal striatal and cortical neural stem cells grafted in stroke-damaged rat striatum. Eur J Neurosci 26: 605-614. 113. Hoehn M, Kustermann E, Blunk J, Wiedermann D, Trapp T, et al. (2002) Monitoring of implanted stem cell migration in vivo: a highly resolved in vivo magnetic resonance 375 imaging investigation of experimental stroke in rat. Proc Natl Acad Sci U S A 99: 16267-16272. 114. Fischer UM, Harting MT, Jimenez F, Monzon-Posadas WO, Xue H, et al. (2009) Pulmonary passage is a major obstacle for intravenous stem cell delivery: the pulmonary first-pass effect. Stem Cells Dev 18: 683-692. 115. Lee ST, Chu K, Jung KH, Kim SJ, Kim DH, et al. (2008) Anti-inflammatory mechanism of intravascular neural stem cell transplantation in haemorrhagic stroke. Brain 131: 616- 629. 116. Li L, Jiang Q, Ding G, Zhang L, Zhang ZG, et al. (2010) Effects of administration route on migration and distribution of neural progenitor cells transplanted into rats with focal cerebral ischemia, an MRI study. J Cereb Blood Flow Metab 30: 653-662. 117. Guo L, Ge J, Wang S, Zhou Y, Wang X, et al. (2013) A novel method for efficient delivery of stem cells to the ischemic brain. Stem Cell Res Ther 4: 116. 118. Keimpema E, Fokkens MR, Nagy Z, Agoston V, Luiten PG, et al. (2009) Early transient presence of implanted bone marrow stem cells reduces lesion size after cerebral ischaemia in adult rats. Neuropathol Appl Neurobiol 35: 89- 102. 119. Gutierrez-Fernandez M, Rodriguez-Frutos B, Alvarez- Grech J, Vallejo-Cremades MT, Exposito-Alcaide M, et al. (2011) Functional recovery after hematic administration of allogenic mesenchymal stem cells in acute ischemic stroke in rats. Neuroscience 175: 394-405. 120. Gorelik M, Orukari I, Wang J, Galpoththawela S, Kim H, et al. (2012) Use of MR cell tracking to evaluate targeting of glial precursor cells to inflammatory tissue by exploiting the very late antigen-4 docking receptor. Radiology 265: 175- 185. 376 121. Suarez-Monteagudo C, Hernandez-Ramirez P, Alvarez- Gonzalez L, Garcia-Maeso I, de la Cuetara-Bernal K, et al. (2009) Autologous bone marrow stem cell neurotransplantation in stroke patients. An open study. Restor Neurol Neurosci 27: 151-161. 122. Savitz SI, Dinsmore J, Wu J, Henderson GV, Stieg P, et al. (2005) Neurotransplantation of fetal porcine cells in patients with basal ganglia infarcts: a preliminary safety and feasibility study. Cerebrovasc Dis 20: 101-107. 123. Li WY, Choi YJ, Lee PH, Huh K, Kang YM, et al. (2008) Mesenchymal stem cells for ischemic stroke: changes in effects after ex vivo culturing. Cell Transplant 17: 1045-1059. 124. Savitz SI, Misra V, Kasam M, Juneja H, Cox CS, Jr., et al. (2011) Intravenous autologous bone marrow mononuclear cells for ischemic stroke. Ann Neurol 70: 59-69. 125. Honmou O, Houkin K, Matsunaga T, Niitsu Y, Ishiai S, et al. (2011) Intravenous administration of auto serumexpanded autologous mesenchymal stem cells in stroke. Brain 134: 1790-1807. 126. Bhasin A, Srivastava MV, Kumaran SS, Mohanty S, Bhatia R, et al. (2011) Autologous mesenchymal stem cells in chronic stroke. Cerebrovasc Dis Extra 1: 93-104. 127. Moniche F, Montaner J, Gonzalez-Marcos JR, Carmona M, Pinero P, et al. (2014) Intra-arterial bone marrow mononuclear cell (BM-MNC) transplantation correlates with GM-CSF, PDGF-BB and MMP-2 serum levels in stroke patients: Results from a clinical trial. Cell Transplant. 128. Friedrich MA, Martins MP, Araujo MD, Klamt C, Vedolin L, et al. (2012) Intra-arterial infusion of autologous bone marrow mononuclear cells in patients with moderate to severe middle cerebral artery acute ischemic stroke. Cell Transplant 21 Suppl 1: S13-21. 377 129. Battistella V, de Freitas GR, da Fonseca LM, Mercante D, Gutfilen B, et al. (2011) Safety of autologous bone marrow mononuclear cell transplantation in patients with nonacute ischemic stroke. Regen Med 6: 45-52. 130. Liu X, Ye R, Yan T, Yu SP, Wei L, et al. (2014) Cell based therapies for ischemic stroke: from basic science to bedside. Prog Neurobiol 115: 92-115. 131. Paul G, Anisimov SV (2013) The secretome of mesenchymal stem cells: potential implications for neuroregeneration. Biochimie 95: 2246-2256. 132. Boxall SA, Jones E (2012) Markers for characterization of bone marrow multipotential stromal cells. Stem Cells Int 2012: 975871. 133. Chamberlain G, Fox J, Ashton B, Middleton J (2007) Concise review: mesenchymal stem cells: their phenotype, differentiation capacity, immunological features, and potential for homing. Stem Cells 25: 2739-2749. 134. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, et al. (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284: 143-147. 135. Makino S, Fukuda K, Miyoshi S, Konishi F, Kodama H, et al. (1999) Cardiomyocytes can be generated from marrow stromal cells in vitro. J Clin Invest 103: 697-705. 136. Chung DJ, Choi CB, Lee SH, Kang EH, Lee JH, et al. (2009) Intraarterially delivered human umbilical cord bloodderived mesenchymal stem cells in canine cerebral ischemia. J Neurosci Res 87: 3554-3567. 137. Koh SH, Kim KS, Choi MR, Jung KH, Park KS, et al. (2008) Implantation of human umbilical cord-derived mesenchymal stem cells as a neuroprotective therapy for ischemic stroke in rats. Brain Res 1229: 233-248. 378 138. Wislet-Gendebien S, Hans G, Leprince P, Rigo JM, Moonen G, et al. (2005) Plasticity of cultured mesenchymal stem cells: switch from nestin-positive to excitable neuronlike phenotype. Stem Cells 23: 392-402. 139. Brazelton TR, Rossi FM, Keshet GI, Blau HM (2000) From marrow to brain: expression of neuronal phenotypes in adult mice. Science 290: 1775-1779. 140. Kopen GC, Prockop DJ, Phinney DG (1999) Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 96: 10711- 10716. 141. Mezey E, Chandross KJ, Harta G, Maki RA, McKercher SR (2000) Turning blood into brain: cells bearing neuronal antigens generated in vivo from bone marrow. Science 290: 1779-1782. 142. Chen J, Li Y, Katakowski M, Chen X, Wang L, et al. (2003) Intravenous bone marrow stromal cell therapy reduces apoptosis and promotes endogenous cell proliferation after stroke in female rat. J Neurosci Res 73: 778-786. 143. Li Y, Chen J, Chen XG, Wang L, Gautam SC, et al. (2002) Human marrow stromal cell therapy for stroke in rat: neurotrophins and functional recovery. Neurology 59: 514- 523. 144. Li Y, Chopp M, Chen J, Wang L, Gautam SC, et al. (2000) Intrastriatal transplantation of bone marrow nonhematopoietic cells improves functional recovery after stroke in adult mice. J Cereb Blood Flow Metab 20: 1311- 1319. 145. Shen LH, Li Y, Chen J, Zhang J, Vanguri P, et al. (2006) Intracarotid transplantation of bone marrow stromal cells 379 increases axon-myelin remodeling after stroke. Neuroscience 137: 393-399. 146. Borlongan CV, Lind JG, Dillon-Carter O, Yu G, Hadman M, et al. (2004) Bone marrow grafts restore cerebral blood flow and blood brain barrier in stroke rats. Brain Res 1010: 108-116. 147. Chen J, Li Y, Wang L, Lu M, Chopp M (2002) Caspase inhibition by Z-VAD increases the survival of grafted bone marrow cells and improves functional outcome after MCAo in rats. J Neurol Sci 199: 17-24. 148. Chopp M, Li Y (2002) Treatment of neural injury with marrow stromal cells. Lancet Neurol 1: 92-100. 149. Li X, Zuo P (2005) Effects of Abeta25-35 on neurogenesis in the adult mouse subventricular zone and dentate gyrus. Neurol Res 27: 218-222. 150. Wang C, Liu F, Liu YY, Zhao CH, You Y, et al. (2011) Identification and characterization of neuroblasts in the subventricular zone and rostral migratory stream of the adult human brain. Cell Res 21: 1534-1550. 151. Curtis MA, Kam M, Nannmark U, Anderson MF, Axell MZ, et al. (2007) Human neuroblasts migrate to the olfactory bulb via a lateral ventricular extension. Science 315: 1243- 1249. 152. Hou SW, Wang YQ, Xu M, Shen DH, Wang JJ, et al. (2008) Functional integration of newly generated neurons into striatum after cerebral ischemia in the adult rat brain. Stroke 39: 2837-2844. 153. Yamashita T, Ninomiya M, Hernandez Acosta P, Garcia- Verdugo JM, Sunabori T, et al. (2006) Subventricular zonederived neuroblasts migrate and differentiate into mature neurons in the post-stroke adult striatum. J Neurosci 26: 6627-6636. 386 206. Moghimi SM, Hunter AC (2000) Poloxamers and poloxamines in nanoparticle engineering and experimental medicine. Trends Biotechnol 18: 412-420. 207. Jain TK, Foy SP, Erokwu B, Dimitrijevic S, Flask CA, et al. (2009) Magnetic resonance imaging of multifunctional pluronic stabilized iron-oxide nanoparticles in tumor-bearing mice. Biomaterials 30: 6748-6756. 208. Alvarez-Lorenzo C, Sosnik A, Concheiro A (2011) PEOPPO block copolymers for passive micellar targeting and overcoming multidrug resistance in cancer therapy. Curr Drug Targets 12: 1112-1130. 209. Bae KH, Ha YJ, Kim C, Lee KR, Park TG (2008) Pluronic/chitosan shell cross-linked nanocapsules encapsulating magnetic nanoparticles. J Biomater Sci Polym Ed 19: 1571-1583. 210. Soenen SJ, De Smedt SC, Braeckmans K (2012) Limitations and caveats of magnetic cell labeling using transfection agent complexed iron oxide nanoparticles. Contrast Media Mol Imaging 7: 140-152. 211. Soenen SJ, Nuytten N, De Meyer SF, De Smedt SC, De Cuyper M (2010) High intracellular iron oxide nanoparticle concentrations affect cellular cytoskeleton and focal adhesion kinase-mediated signaling. Small 6: 832-842. 212. Crabbe A, Vandeputte C, Dresselaers T, Sacido AA, Verdugo JM, et al. (2010) Effects of MRI contrast agents on the stem cell phenotype. Cell Transplant 19: 919-936. 213. Soenen SJ, De Cuyper M (2010) Assessing iron oxide nanoparticle toxicity in vitro: current status and future prospects. Nanomedicine (Lond) 5: 1261-1275. 214. Soenen SJ, Himmelreich U, Nuytten N, De Cuyper M (2011) Cytotoxic effects of iron oxide nanoparticles and 387 implications for safety in cell labelling. Biomaterials 32: 195- 205. 215. Soenen SJ, Himmelreich U, Nuytten N, Pisanic TR, 2nd, Ferrari A, et al. (2010) Intracellular nanoparticle coating stability determines nanoparticle diagnostics efficacy and cell functionality. Small 6: 2136-2145. 216. Arbab AS, Yocum GT, Wilson LB, Parwana A, Jordan EK, et al. (2004) Comparison of transfection agents in forming complexes with ferumoxides, cell labeling efficiency, and cellular viability. Mol Imaging 3: 24-32. 217. Kalish H, Arbab AS, Miller BR, Lewis BK, Zywicke HA, et al. (2003) Combination of transfection agents and magnetic resonance contrast agents for cellular imaging: relationship between relaxivities, electrostatic forces, and chemical composition. Magn Reson Med 50: 275-282. 218. Son KK, Patel DH, Tkach D, Park A (2000) Cationic liposome and plasmid DNA complexes formed in serum-free medium under optimum transfection condition are negatively charged. Biochim Biophys Acta 1466: 11-15. 219. Wang S, Fang J, Zhang T, Wang B, Chen J, et al. (2011) Magnetic resonance imaging targeting of intracranial glioma xenografts by Resovist-labeled endothelial progenitor cells. J Neurooncol 105: 67-75. 220. Jasmin, Torres AL, Nunes HM, Passipieri JA, Jelicks LA, et al. (2011) Optimized labeling of bone marrow mesenchymal cells with superparamagnetic iron oxide nanoparticles and in vivo visualization by magnetic resonance imaging. J Nanobiotechnology 9: 4. 221. Riggio C, Calatayud MP, Hoskins C, Pinkernelle J, Sanz B, et al. (2012) Poly-l-lysine-coated magnetic nanoparticles as intracellular actuators for neural guidance. Int J Nanomedicine 7: 3155-3166. 388 222. Smirnov P, Gazeau F, Beloeil JC, Doan BT, Wilhelm C, et al. (2006) Single-cell detection by gradient echo 9.4 T MRI: a parametric study. Contrast Media Mol Imaging 1: 165-174. 223. Carenza E, Barcelo V, Morancho A, Levander L, Boada C, et al. (2013) In vitro angiogenic performance and in vivo brain targeting of magnetized endothelial progenitor cells for neurorepair therapies. Nanomedicine. 224. Nishida K, Tanaka N, Nakanishi K, Kamei N, Hamasaki T, et al. (2006) Magnetic targeting of bone marrow stromal cells into spinal cord: through cerebrospinal fluid. Neuroreport 17: 1269-1272. 225. Kyrtatos PG, Lehtolainen P, Junemann-Ramirez M, Garcia-Prieto A, Price AN, et al. (2009) Magnetic tagging increases delivery of circulating progenitors in vascular injury. JACC Cardiovasc Interv 2: 794-802. 226. Pardoe H, Chua-anusorn W, St. Pierre TG, Dobson J (2001) Structural and magnetic properties of nanoscale iron oxide particles synthesized in the presence of dextran or polyvinyl alcohol. Journal of Magnetism and Magnetic Materials 225: 41-46. 227. Trekker J, Leten C, Struys T, Lazenka VV, Argibay B, et al. (2014) Sensitive in vivo cell detection using size-optimized superparamagnetic nanoparticles. Biomaterials 35: 1627- 1635. 228. Petcharoen K, Sirivat A (2012) Synthesis and characterization of magnetite nanoparticles via the chemical co-precipitation method. Materials Science and Engineering: B 177: 421-427. 229. Kim W, Suh CY, Cho SW, Roh KM, Kwon H, et al. (2012) A new method for the identification and quantification of magnetite-maghemite mixture using conventional X-ray diffraction technique. Talanta 94: 348-352. 389 230. Wang YX (2011) Superparamagnetic iron oxide based MRI contrast agents: Current status of clinical application. Quant Imaging Med Surg 1: 35-40. 231. Park JH, von Maltzahn G, Zhang L, Schwartz MP, Ruoslahti E, et al. (2008) Magnetic Iron Oxide Nanoworms for Tumor Targeting and Imaging. Adv Mater 20: 1630-1635. 232. Simberg D, Park JH, Karmali PP, Zhang WM, Merkulov S, et al. (2009) Differential proteomics analysis of the surface heterogeneity of dextran iron oxide nanoparticles and the implications for their in vivo clearance. Biomaterials 30: 3926-3933. 233. Carmen Bautista M, Bomati-Miguel O, del Puerto Morales M, Serna CJ, Veintemillas-Verdaguer S (2005) Surface characterisation of dextran-coated iron oxide nanoparticles prepared by laser pyrolysis and coprecipitation. Journal of Magnetism and Magnetic Materials 293: 20-27. 234. Mahdavi M, Ahmad MB, Haron MJ, Namvar F, Nadi B, et al. (2013) Synthesis, surface modification and characterisation of biocompatible magnetic iron oxide nanoparticles for biomedical applications. Molecules 18: 7533-7548. 235. Gao Q, Liang Q, Yu F, Xu J, Zhao Q, et al. (2011) Synthesis and characterization of novel amphiphilic copolymer stearic acid-coupled F127 nanoparticles for nano-technology based drug delivery system. Colloids Surf B Biointerfaces 88: 741-748. 236. Kassis I, Vaknin-Dembinsky A, Bulte J, Karussis D (2010) Effects of supermagnetic iron oxide labeling on the major functional properties of human mesenchymal stem cells from multiple sclerosis patients. Int J Stem Cells 3: 144-153. 237. Yu MX, Chen WL, Zhou Q, Gao P (2010) Study on ASTC- a-1 cells labeled with superparamagnetic iron oxide and its 390 magnetic resonance imaging. Exp Biol Med (Maywood) 235: 1053-1061. 238. Roeder E, Henrionnet C, Goebel JC, Gambier N, Beuf O, et al. (2014) Dose-response of superparamagnetic iron oxide labeling on mesenchymal stem cells chondrogenic differentiation: a multi-scale in vitro study. PLoS One 9: e98451. 239. Struys T, Ketkar-Atre A, Gervois P, Leten C, Hilkens P, et al. (2013) Magnetic resonance imaging of human dental pulp stem cells in vitro and in vivo. Cell Transplant 22: 1813- 1829. 240. Nohroudi K, Arnhold S, Berhorn T, Addicks K, Hoehn M, et al. (2010) In vivo MRI stem cell tracking requires balancing of detection limit and cell viability. Cell Transplant 19: 431-441. 241. Soenen SJ, Brisson AR, Jonckheere E, Nuytten N, Tan S, et al. (2011) The labeling of cationic iron oxide nanoparticle-resistant hepatocellular carcinoma cells using targeted magnetoliposomes. Biomaterials 32: 1748-1758. 242. Kroll A, Pillukat MH, Hahn D, Schnekenburger J (2012) Interference of engineered nanoparticles with in vitro toxicity assays. Arch Toxicol 86: 1123-1136. 243. Holder AL, Goth-Goldstein R, Lucas D, Koshland CP (2012) Particle-induced artifacts in the MTT and LDH viability assays. Chem Res Toxicol 25: 1885-1892. 244. Schmidtke-Schrezenmeier G, Urban M, Musyanovych A, Mailander V, Rojewski M, et al. (2011) Labeling of mesenchymal stromal cells with iron oxide-poly(L-lactide) nanoparticles for magnetic resonance imaging: uptake, persistence, effects on cellular function and magnetic resonance imaging properties. Cytotherapy 13: 962-975. 391 245. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248- 254. 246. Jones E, McGonagle D (2008) Human bone marrow mesenchymal stem cells in vivo. Rheumatology 47: 126-131. 247. Ma S, Xie N, Li W, Yuan B, Shi Y, et al. (2014) Immunobiology of mesenchymal stem cells. Cell Death Differ 21: 216-225. 248. Song M, Kim YJ, Kim YH, Roh J, Kim SU, et al. (2010) Using a neodymium magnet to target delivery of ferumoxide-labeled human neural stem cells in a rat model of focal cerebral ischemia. Hum Gene Ther 21: 603-610. 249. Vanecek V, Zablotskii V, Forostyak S, Ruzicka J, Herynek V, et al. (2012) Highly efficient magnetic targeting of mesenchymal stem cells in spinal cord injury. Int J Nanomedicine 7: 3719-3730. 250. Mikhaylova M, Mori N, Wildes FB, Walczak P, Gimi B, et al. (2008) Hypoxia increases breast cancer cell-induced lymphatic endothelial cell migration. Neoplasia 10: 380-389. 251. Gelosa P, Ballerio R, Banfi C, Nobili E, Gianella A, et al. (2010) Terutroban, a thromboxane/prostaglandin endoperoxide receptor antagonist, increases survival in stroke-prone rats by preventing systemic inflammation and endothelial dysfunction: comparison with aspirin and rosuvastatin. J Pharmacol Exp Ther 334: 199-205. 252. Yukawa H, Noguchi H, Oishi K, Takagi S, Hamaguchi M, et al. (2009) Cell transplantation of adipose tissue-derived stem cells in combination with heparin attenuated acute liver failure in mice. Cell Transplant 18: 611-618. 253. Yukawa H, Watanabe M, Kaji N, Okamoto Y, Tokeshi M, et al. (2012) Monitoring transplanted adipose tissue-derived 392 stem cells combined with heparin in the liver by fluorescence imaging using quantum dots. Biomaterials 33: 2177-2186. 254. Stroke Therapy Academic Industry R (1999) Recommendations for standards regarding preclinical neuroprotective and restorative drug development. Stroke 30: 2752-2758. 255. Wang-Fischer Y (2008) Manual of Stroke Models in Rats: Taylor & Francis. 256. Devault K, Gremaud PA, Novak V, Olufsen MS, Vernieres G, et al. (2008) Blood Flow in the Circle of Willis: Modeling and Calibration. Multiscale Model Simul 7: 888- 909. 257. Coyle P, Jokelainen PT (1982) Dorsal cerebral arterial collaterals of the rat. Anat Rec 203: 397-404. 258. van Everdingen KJ, van der Grond J, Kappelle LJ, Ramos LM, Mali WP (1998) Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke 29: 1783-1790. 259. Yavagal DR, Lin B, Raval AP, Garza PS, Dong C, et al. (2014) Efficacy and dose-dependent safety of intra-arterial delivery of mesenchymal stem cells in a rodent stroke model. PLoS One 9: e93735. 260. Karlupia N, Manley NC, Prasad K, Schafer R, Steinberg GK (2014) Intraarterial transplantation of human umbilical cord blood mononuclear cells is more efficacious and safer compared with umbilical cord mesenchymal stromal cells in a rodent stroke model. Stem Cell Res Ther 5: 45. 261. Vasconcelos-dos-Santos A, Rosado-de-Castro PH, Lopes de Souza SA, da Costa Silva J, Ramos AB, et al. (2012) Intravenous and intra-arterial administration of bone marrow mononuclear cells after focal cerebral ischemia: Is 393 there a difference in biodistribution and efficacy? Stem Cell Res 9: 1-8. 262. Fagan SC, Lapchak PA, Liebeskind DS, Ishrat T, Ergul A (2013) Recommendations for preclinical research in hemorrhagic transformation. Transl Stroke Res 4: 322-327. 263. Wang L, Menendez P, Shojaei F, Li L, Mazurier F, et al. (2005) Generation of hematopoietic repopulating cells from human embryonic stem cells independent of ectopic HOXB4 expression. J Exp Med 201: 1603-1614. 264. Finley RB (1953) A New Subspecies of Wood Rat (Neotoma mexicana) from Colorado: University of Kansas Publications Museum of Natural History. 527-534 p. 265. Willing AE, Garbuzova-Davis S, Sanberg PR, Saporta S (2002) Routes of stem cell administration in the adult rodent. Methods Mol Biol 198: 357-374. 266. Schrepfer S, Deuse T, Reichenspurner H, Fischbein MP, Robbins RC, et al. (2007) Stem cell transplantation: the lung barrier. Transplant Proc 39: 573-576. 267. Villalobos V, Bonilla E, Castellano A, Novo E, Caspersen R, et al. (2009) Ultrastructural changes of the olfactory bulb in manganese-treated mice. Biocell 33: 187-197. 268. Iihoshi S, Honmou O, Houkin K, Hashi K, Kocsis JD (2004) A therapeutic window for intravenous administration of autologous bone marrow after cerebral ischemia in adult rats. Brain Res 1007: 1-9. 269. Zhang L, Li Y, Zhang C, Chopp M, Gosiewska A, et al. (2011) Delayed administration of human umbilical tissuederived cells improved neurological functional recovery in a rodent model of focal ischemia. Stroke 42: 1437-1444. 270. Zhang L, Yi L, Chopp M, Kramer BC, Romanko M, et al. (2012) Intravenous administration of human umbilical tissue- 394 derived cells improves neurological function in aged rats after embolic stroke. Cell Transplant. 271. Ramos-Cabrer P, Justicia C, Wiedermann D, Hoehn M (2010) Stem cell mediation of functional recovery after stroke in the rat. PLoS One 5: e12779. 272. Vu Q, Xie K, Eckert M, Zhao W, Cramer SC (2014) Metaanalysis of preclinical studies of mesenchymal stromal cells for ischemic stroke. Neurology 82: 1277-1286. 273. Suzuki J, Sasaki M, Harada K, Bando M, Kataoka Y, et al. (2013) Bilateral cortical hyperactivity detected by fMRI associates with improved motor function following intravenous infusion of mesenchymal stem cells in a rat stroke model. Brain Res 1497: 15-22. 274. Chopp M, Li Y (2008) Treatment of stroke and intracerebral hemorrhage with cellular and pharmacological restorative therapies. Acta Neurochir Suppl 105: 79-83. 275. Zhang ZG, Zhang L, Jiang Q, Zhang R, Davies K, et al. (2000) VEGF enhances angiogenesis and promotes bloodbrain barrier leakage in the ischemic brain. J Clin Invest 106: 829-838. 276. Chen J, Shehadah A, Pal A, Zacharek A, Cui X, et al. (2013) Neuroprotective effect of human placenta-derived cell treatment of stroke in rats. Cell Transplant 22: 871-879.