Abstract Ciprofloxacin was incorporated to plain bioresorbable self-reinforced polylactide/glycolyde 80/20 screws and tacks (ciprofloxacin releasing SR-PLGA). These implants were compared to otherwise similar conventional fixation devices. The effect of the ciprofloxacin addition on the pull-out force of screws and tacks was evaluated in human cadaver cranial bones. SR-PLGA tacks applied to cranial bone with a special applicator gun had a similar holding power as screws. Addition of the antibiotic compromised the strength of the screws so that ciprofloxacin-containing PLGA screws had lower pull-out strength than corresponding plain PLGA screws. Scanning electron microscopy showed that the fibrillar strip-like microstructure of plain SR-PLGA screws turned into a coarse uni-axial platelet-like pattern as a result of ciprofloxacin addition. It is concluded that this type of 4 mm long and 1.5 mm diameter ciprofloxacin-containing screws can only be used in non-load-bearing or slightly load-bearing applications. Tissue reactions elicited by plain bioresorbable self-reinforced polylactide/glycolide (SR-PLGA) 80/20 screws were compared to similar but ciprofloxacin-releasing SR-PLGA fixation devices in rabbit cranial bone. Plain and ciprofloxacin-PLGA 80/20 screws elicited only mild inflammatory reactions upon implantation in rabbit cranial bone, but they did not interfere with osteoblast activity in up to 72 week long follow-up. Release of the antibiotic from ciprofloxacin-PLGA screws was gradual and the drug concentration in bone tissues was still higher at 8 weeks than the minimal inhibitory concentration (MIC) of ciprofloxacin for S. aureus (0.1–1.0 μg/g). Ciprofloxacin-releasing SR-PLGA screws can find clinical usage in the prevention of implant-related infections in osteofixation in craniomaxillofacial bones in non-load-bearing or slightly load-bearing applications. Larger 6 mm long and 2 mm diameter ciprofloxacin-releasing tacks had a similar holding power to cranial bone as conventional tacks. Tacks can be recommended for clinical use as the application procedure saves time and costs.

Neste estudo analisamos o envolvimento de E3 ligases sob três aspectos da plasticidade muscular esquelética: a perda de massa decorrente do desuso, a manutenção de fibras tipo I e II e a regeneração do tecido muscular. O primeiro objetivo foi determinar se leucina atenuaria a perda de massa provocada pela atrofia. Nossos dados mostraram que este aminoácido evitou a perda de massa, sobretudo através da inibição da expressão de E3 ligases. A seguir, analisamos se a ausência de duas E3 ligases, MuRF1 e MuRF2, poderia alterar a proporção de miofibras tipo I e tipo II. Detectamos que na ausência destas E3 ligases, a identidade de fibras tipo II era perdida, além destas fibras estarem protegidas da atrofia sem MuRF1. Por fim, analisamos o papel de MuRF1 e MuRF2 durante a regeneração. Os resultados mostraram que estas E3 ligases em conjunto são cruciais para a fisiologia das células-satélites e portanto para a regeneração do tecido. Esta tese mostrou que determinadas E3 ligases exercem um papel crucial para a plasticidade muscular. / In the present thesis we analyzed the involvement of E3 ligases under three aspects of skeletal muscle plasticity: mass loss resulting from disuse, maintenance of fiber type I and II and regeneration of muscle tissue. Our first aim was to determine whether leucine, was able to attenuate the mass loss caused by wasting. Our results showed that this amino acid prevented the mass loss, mainly by inhibiting the expression of E3 ligases. The second aim was to determine whether the absence of two E3 ligases, MuRF1 and MuRF2, could alter the proportion of type I and type II fibers. We found that in the absence of these E3 ligases, the identity of type II fibers was lost, and these fibers were protected against atrophy in the absence of MuRF1. Lastly, we analyze the role of MuRF1 and MuRF2 in muscle tissue regeneration. The results showed that these E3 ligases together are crucial to satellite-cells physiology and consequently an adequate tissue regeneration. This thesis show that certain E3 ligases can play a crucial role in muscle plasticity.

Neste estudo analisamos o envolvimento de E3 ligases sob três aspectos da plasticidade muscular esquelética: a perda de massa decorrente do desuso, a manutenção de fibras tipo I e II e a regeneração do tecido muscular. O primeiro objetivo foi determinar se leucina atenuaria a perda de massa provocada pela atrofia. Nossos dados mostraram que este aminoácido evitou a perda de massa, sobretudo através da inibição da expressão de E3 ligases. A seguir, analisamos se a ausência de duas E3 ligases, MuRF1 e MuRF2, poderia alterar a proporção de miofibras tipo I e tipo II. Detectamos que na ausência destas E3 ligases, a identidade de fibras tipo II era perdida, além destas fibras estarem protegidas da atrofia sem MuRF1. Por fim, analisamos o papel de MuRF1 e MuRF2 durante a regeneração. Os resultados mostraram que estas E3 ligases em conjunto são cruciais para a fisiologia das células-satélites e portanto para a regeneração do tecido. Esta tese mostrou que determinadas E3 ligases exercem um papel crucial para a plasticidade muscular. / In the present thesis we analyzed the involvement of E3 ligases under three aspects of skeletal muscle plasticity: mass loss resulting from disuse, maintenance of fiber type I and II and regeneration of muscle tissue. Our first aim was to determine whether leucine, was able to attenuate the mass loss caused by wasting. Our results showed that this amino acid prevented the mass loss, mainly by inhibiting the expression of E3 ligases. The second aim was to determine whether the absence of two E3 ligases, MuRF1 and MuRF2, could alter the proportion of type I and type II fibers. We found that in the absence of these E3 ligases, the identity of type II fibers was lost, and these fibers were protected against atrophy in the absence of MuRF1. Lastly, we analyze the role of MuRF1 and MuRF2 in muscle tissue regeneration. The results showed that these E3 ligases together are crucial to satellite-cells physiology and consequently an adequate tissue regeneration. This thesis show that certain E3 ligases can play a crucial role in muscle plasticity.

Protonentherapie ist eine wichtige Behandlungsmodalität in der Radioonkologie. Aufgrund einer vorteilhaften Dosisverteilung im bestrahlten Volumen kann diese Bestrahlungsmethode das tumorumgebende Normalgewebe schützen. Dadurch können Nebenwirkungen in bestimmten Patientenpopulationen, zum Beispiel Kindern oder Patienten mit Gehirntumoren, verringert werden. Trotzdem können nach Protonenbestrahlung von Gehirntumorpatienten Normalgewebsschäden auftreten. Gründe dafür können der notwendige klinische Sicherheitssaum im Normalgewebe, der Einfluss der relativen biologischen Wirksamkeit RBE sowie eine erhöhte Strahlensensitivität bestimmter Gehirnregionen sein. Um diese Aspekte zu beleuchten, werden geeignete präklinische Modelle für die Normalgewebsreaktion im Gehirn nach Protonenbestrahlung benötigt. Darüber hinaus kann eine Risikostratifizierung der Patienten durch die Vorhersage von Nebenwirkungswahrscheinlichkeiten oder der Tumorantwort den Behandlungserfolg erhöhen. Auch hier können präklinische Modelle helfen, um neue prädiktive Biomarker zu finden und um die zugrunde liegenden Mechanismen strahleninduzierter Gehirnschäden besser zu verstehen. Das Ziel dieser Dissertation war die Etablierung und Charakterisierung von adäquaten präklinischen Modellen für die Untersuchung von strahleninduzierten Normalgewebsschäden im Gehirn. Diese Modelle bilden die Grundlage für zukünftige Studien zur Untersuchung von RBE Effekten, der spezifische Strahlensensitivität einzelner Gehirnregionen und neuer Biomarker. Die getesteten Modellsysteme waren in vitro Kulturen von adulten organotypischen Gehirnschnitten, Tumorschnittkultur sowie in vivo Bestrahlung von Gehirnsubvolumina, jeweils mit dem Modellorganismus Maus. Die Etablierung eines Bestrahlungssetups in der experimentellen Protonenanlage und dessen dosimetrische Charakterisierung waren von großer Bedeutung für die Durchführung der biologischen Experimente. Ein weiteres Hauptziel war die Definition klinisch relevanter Endpunkte für frühe und späte Nebenwirkungen. Die Gewebsschnitte wurden durch Messungen des Zellüberlebens und der Entzündungsreaktion, sowie mittels in situ Analyse von Zellmorphologie und DNA Schäden untersucht. Als ergänzendes Modell wurde die Tumorschnittkultur etabliert und ähnliche Endpunkte analysiert. Adulte Gehirnschnitte stellten sich als ungeeignet für präklinische Experimente in der Radioonkologie heraus. Die Messungen von Zelltod und Entzündungswerten zeigten eine starke Zellreaktion auf die Inkulturnahme, aber keine auf die Protonenbestrahlung. In der Histologie wurden gestörte Zellmorphologie, reduzierte Vitalität und eingeschränkte Reparaturfähigkeit von DNA Schäden beobachtet. Daher sollten für strahlenbiologische Experimente andere 3D Zellkulturmodelle in Betracht gezogen werden, wie zum Beispiel Organoide oder durch Tissue Engineering hergestellte Kulturen. Durch die Publikation der Daten leistet diese Dissertation einen wichtigen Beitrag zur aktuellen Forschung, da so künftig die limitierten Ressourcen, die für strahlenbiologische Experimente mit Protonen zur Verfügung stehen, auf relevantere Modelle verwendet werden können. Die Bestrahlung von Gehirnsubvolumina in Mäusen wurde mit dem Ziel etabliert, klinisch vergleichbare Felder zu erreichen. Das gewählte Zielvolumen war der rechte Hippocampus; der Protonenstrahl sollte in der Mitte des Gehirns stoppen. Im Rahmen des Projekts wurde ein Arbeitsablauf für präzise und reproduzierbare Bestrahlung entwickelt. Zur Verifizierung wurde der induzierte DNA Schaden ausgewertet und anschließend mit Monte-Carlos Dosissimulationen korreliert. Die Maushirnbestrahlung lieferte wertvolle Ergebnisse für frühe Zeitpunkte (d.h. innerhalb 24 h nach Bestrahlung). Im Verlauf des Projekts wurde ein Algorithmus erstellt, der schnell und zuverlässig die räumliche Verteilung des DNA Schadens in Relation zur Gesamtzellzahl analysiert. Diese Auswertung zeigte, wie bei der Bestrahlungsplanung vorgesehen, ein Stoppen des Protonenstrahls im Gehirn. Eine anschließende Korrelation der Schadensverteilung mit der applizierten Dosis weist nach, dass das Modell einen wichtigen Beitrag zur Untersuchung des RBE leisten kann. In einer darauf folgenden Studie wurde der Dosis-Zeitverlauf der beobachteten Strahlenreaktion des Normalgewebes genauer beleuchtet. Dafür wurden Untersuchungen des Allgemeinzustands der Versuchstiere, regelmäßige Magnetresonanztomografie (MRI) Messungen über einen Zeitraum von sechs Monaten, sowie abschließende Histologie korreliert. Die Volumenzunahme des Kontrastmittelaustritts, die den Zusammenbruch der Blut-Hirn-Schranke anzeigt, wurde konturiert; aus diesen Daten entstand ein prädiktives Dosis-Volumen Modell. Die Pilotstudie konnte eine dosisabhängige Strahlenreaktion nachweisen, die sich im Zusammenbruch der Blut-Hirn-Schranke, einer Hautreaktion mit vorrübergehender Alopezie, Gewichtsabnahme und zelluläre Veränderung äußerte. Das von den MRI Messungen abgeleitete Modell konnte zuverlässig das Eintreten der Nebenwirkungen, den Krankheitsverlauf, sowie die geschätzte Überlebensdauer der Mäuse vorhersagen. Zusätzlich konnte ein Zusammenhang zwischen den MRI Bildänderungen und den pathologischen Gewebsveränderungen beobachtet werden. Durch die außerordentlich homogene Strahlenreaktion der Tiere können aus den vorliegenden Daten künftig zuverlässig geeignete Dosen für spezifische experimentelle Endpunkte bestimmt werden. Zusammenfassend wurden in dieser Arbeit zwei präklinische Modelle für die Protonengehirnbestrahlung etabliert, nämlich organotypische Gewebsschnitte als 3D Zellkulturmodell sowie in vivo Bestrahlung von Gehirnsubvolumina in Mäusen. Während Zellkulturexperimente die Erwartungen nicht erfüllen konnten, stellen sich die Tierexperimente als hervorragendes Modell für translationale Radioonkologie heraus, welches zusätzlich für andere Strahlenqualitäten eingesetzt werden kann. Darauf basierend können aktuelle und zukünftige Studien die Ursachen von strahleninduzierten Normalgewebsschäden im Gehirn beleuchten, RBE Effekte untersuchen und neue prädiktive Biomarker erforschen.:Contents Abstract i Zusammenfassung v Publications ix List of Figures xiii List of Acronyms and Abbreviations xiv 1 Introduction 3 2 Background 5 2.1 Proton therapy for brain cancer treatment 5 2.1.1 Fundamentals of radiobiology 5 2.1.2 Proton therapy 6 2.1.3 Tumors of the central nervous system 8 2.2 Radiation effects on brain cells 8 2.2.1 Neurons and myelin 9 2.2.2 Blood-brain barrier 9 2.2.3 Astrocytes 10 2.2.4 Microglia 10 2.3 Principles of histology 11 2.3.1 Hematoxylin & eosin staining 12 2.3.2 Immunohistochemistry 13 2.3.3 Bioimage analysis 13 2.4 Techniques in medical imaging 14 2.4.1 Projectional radiography 14 2.4.2 Computed tomography 14 2.4.3 Magnetic resonance imaging 15 2.5 Preclinical models for radiation injury 17 2.5.1 Technical requirements 17 2.5.2 In vitro models 17 2.5.3 Small animal models 18 3 Applying Tissue Slice Culture in Cancer Research – Insights from Preclinical Proton Radiotherapy 19 3.1 Aim of the study 19 3.2 Conclusion 19 3.3 Author’s contribution 19 3.4 Publication 21 4 High-precision image-guided proton irradiation of mouse brain sub-volumes 41 4.1 Aim of the study 41 4.2 Conclusion 41 4.3 Author’s contribution 41 4.4 Publication 43 5 Late side effects in normal mouse brain tissue after proton irradiation 51 5.1 Aim of the study 51 5.2 Conclusion 51 5.3 Author’s contribution 52 5.4 Publication 53 6 Discussion 71 6.1 Establishment of preclinical models for radiooncology 71 6.1.1 3D cell culture 71 6.1.2 In vivo irradiation of brain subvolumes 73 6.2 Current applications of the mouse model 75 6.2.1 Ongoing data analysis 75 6.2.2 Innovating on-site imaging 76 6.2.3 RBE investigations 77 6.3 Future studies of radiation-induced brain tissue toxicities 79 Acknowledgement XV Supplementary Material XVII 1 Applying Tissue Slice Culture in Cancer Research – Insights from Preclinical Proton Radiotherapy XVII 2 High-precision image-guided proton irradiation of mouse brain sub-volumes XXVI 3 Late side effects in normal mouse brain tissue after proton irradiation XXXI / Proton therapy is an important modality in radiation oncology. Due to a favorable dose distribution in the irradiated volume, this treatment allows to spare tumor-surrounding normal tissue. Although this protection can lead to reduced side effects in certain patient populations, such as brain tumor or pediatric patients, normal tissue toxicities can occur to some extend. This could be due to clinical safety margins around the tumor that lead to dose deposition in the normal tissue. The underlying causes might also be related to relative biological effectiveness (RBE) variations or elevated radiosensitivity of certain brain regions. To address these issues, suitable preclinical models for normal brain tissue reaction after proton therapy are needed. In addition, patient stratification to predict the tumor response or the probability of side effects will contribute to increased treatment effectiveness. Preclinical models can improve the process of finding new predictive biomarkers and help to understand underlying mechanisms of radiation-induced brain injury. The aim of this thesis was to establish and characterize suitable preclinical models of brain tissue irradiation effects and set the base for future studies designed to reveal RBE effects, brain region specific radiation sensitivities, and novel biomarkers. The tested model systems were in vitro organotypic brain slice culture (OBSC) and in vivo irradiation of brain subvolumes, both on mouse brain tissue. Setup establishment at the experimental proton beam line and subsequent dosimetry built the foundation for conducting the biological experiments. Additionally, one main goal was defining clinically relevant endpoints for both short- and long-term effects. For OBSC, assays for cell death and inflammation, as well as in situ analysis of cell morphology and DNA damage induction were tested. As comparative model to OBSC, tumor slice culture was established and the results were also used for proton investigation. Adult OBSC turned out as inadequate model for preclinical experiments in radiation oncology. The assays measuring cell death and inflammation indicated a severe reaction during the first days in culture, but no response to irradiation. Histology revealed deficient cell morphology, reduced vitality and impaired DNA damage repair. In conclusion, other 3D cell culture models, such as organoids or tissue engineered constructs, should be considered for radiobiological experiments with protons. By publishing the observations, this thesis contributes to conserving the limited resources of proton radiobiology for more meaningful models. A methodology for irradiation of mouse brain subvolumes was established with a focus on creating fields comparable to clinical practice. The chosen target was the right hippocampus and the goal was to stop the proton beam in the middle of the brain. The project included a workflow for this precise irradiation in a robust and reproducible manner. Evaluation of the induced DNA damage and its correlation to Monte Carlo dose simulations were used for verification. Irradiation of mouse brain subvolumes yielded valuable results for early (i.e. within 24 h after irradiation) time points. An evaluation algorithm was designed for fast and robust analysis of spatial DNA damage distribution in relation to the total cell count. This ratio showed that the beam stopped in the brain tissue, in accordance to the treatment planning. Furthermore, the DNA damage could be reliably correlated with the dose simulation, which proves the value of the presented model for future RBE studies. In a follow-up experiment, the dose-time relationship of induced normal tissue reactions was analysed. For this, scoring of the animals' health status was combined with regular MRI measurements over the course of up to 6 months, and final histopathology. The volume increase of contrast agent leakage - representing breakdown of the blood brain barrier (BBB) - was contoured and the data was used to create a dose-volume response model. This pilot study on long-term radiation effects revealed dose-dependent normal tissue toxicities, including breakdown of the BBB, a skin reaction with temporary alopecia, weight reduction and changes on the cellular level. The model derived from MRI data reliably predicts onset of side effects, volume of brain damage as well as the expected animal survival. In addition, MRI image changes could be correlated to underlying tissue alterations by histopathology. Due to the uniform radiation response of the animals this data set enables to determine endpoint-specific dose values in future experiments. In conclusion, two preclinical models for proton brain irradiation were established, namely OBSC as 3D cell culture model and in vivo irradiation of mouse brain subvolumes. While the former could not yield the anticipated results, the latter emerged as excellent model for translational radiooncology, which can also be applied for experiments with other radiation types. Ongoing and future studies will focus on revealing the causes of normal brain tissue toxicities, studying RBE effects, and investigating new predictive biomarkers.:Contents Abstract i Zusammenfassung v Publications ix List of Figures xiii List of Acronyms and Abbreviations xiv 1 Introduction 3 2 Background 5 2.1 Proton therapy for brain cancer treatment 5 2.1.1 Fundamentals of radiobiology 5 2.1.2 Proton therapy 6 2.1.3 Tumors of the central nervous system 8 2.2 Radiation effects on brain cells 8 2.2.1 Neurons and myelin 9 2.2.2 Blood-brain barrier 9 2.2.3 Astrocytes 10 2.2.4 Microglia 10 2.3 Principles of histology 11 2.3.1 Hematoxylin & eosin staining 12 2.3.2 Immunohistochemistry 13 2.3.3 Bioimage analysis 13 2.4 Techniques in medical imaging 14 2.4.1 Projectional radiography 14 2.4.2 Computed tomography 14 2.4.3 Magnetic resonance imaging 15 2.5 Preclinical models for radiation injury 17 2.5.1 Technical requirements 17 2.5.2 In vitro models 17 2.5.3 Small animal models 18 3 Applying Tissue Slice Culture in Cancer Research – Insights from Preclinical Proton Radiotherapy 19 3.1 Aim of the study 19 3.2 Conclusion 19 3.3 Author’s contribution 19 3.4 Publication 21 4 High-precision image-guided proton irradiation of mouse brain sub-volumes 41 4.1 Aim of the study 41 4.2 Conclusion 41 4.3 Author’s contribution 41 4.4 Publication 43 5 Late side effects in normal mouse brain tissue after proton irradiation 51 5.1 Aim of the study 51 5.2 Conclusion 51 5.3 Author’s contribution 52 5.4 Publication 53 6 Discussion 71 6.1 Establishment of preclinical models for radiooncology 71 6.1.1 3D cell culture 71 6.1.2 In vivo irradiation of brain subvolumes 73 6.2 Current applications of the mouse model 75 6.2.1 Ongoing data analysis 75 6.2.2 Innovating on-site imaging 76 6.2.3 RBE investigations 77 6.3 Future studies of radiation-induced brain tissue toxicities 79 Acknowledgement XV Supplementary Material XVII 1 Applying Tissue Slice Culture in Cancer Research – Insights from Preclinical Proton Radiotherapy XVII 2 High-precision image-guided proton irradiation of mouse brain sub-volumes XXVI 3 Late side effects in normal mouse brain tissue after proton irradiation XXXI