The number of proton therapy centers worldwide are increasing steadily, with more than two million cancer patients treated so far. Despite this development, pending questions on proton radiobiology still call for basic and translational preclinical research. Open issues are the on-going discussion on an energy-dependent varying proton RBE (relative biological effectiveness), a better characterization of normal tissue side effects and combination treatments with drugs originally developed for photon therapy. At the same time, novel possibilities arise, such as radioimmunotherapy, and new proton therapy schemata, such as FLASH irradiation and proton mini-beams. The study of those aspects demands for radiobiological models at different stages along the translational chain, allowing the investigation of mechanisms from the molecular level to whole organisms. Focusing on the challenges and specifics of proton research, this review summarizes the different available models, ranging from in vitro systems to animal studies of increasing complexity as well as complementing in silico approaches.

Background and purpose: After proton therapy of brain tumors, several studies have reported late image changes in follow-up magnetic resonance imaging, which result from blood–brain barrier (BBB) disruption. Astrocytes play a central role in the formation and maintenance of the BBB. To study the late response to partial-brain proton irradiation, preclinical mouse data were utilized to investigate the spatial distribution and dose dependence of reactive astrocytes. Material and methods: Previously, C57BL/6JRj mice were irradiated with protons targeting the right hippocampal region with single prescription doses of 45–85 Gy. After six months, mice were sacrificed and the excised brains axially cut into 3 µm thick slices and stained for glial fibrillary acidic protein (GFAP) to target astrocytes. Here, a workflow to segment the GFAP-positive area on slice images was established. The fraction of GFAP-positive area (GFAP+ fraction) was evaluated in the high-dose region in the right hemisphere and in the mirrored region in the left hemisphere. Dose distributions were simulated on pre-irradiation cone-beam computed tomography and co-registered to the histological slices. Results: For all irradiated mice, the GFAP+ fraction in the right hemisphere was significantly increased compared to the left hemisphere and to a sham-irradiated mouse with a highly symmetric GFAP distribution. The GFAP+ fraction in the right hemisphere increased approximately linearly with prescription dose. For comparable doses, the cerebral cortex showed lower GFAP+ fractions than the midbrain. Interpretation: GFAP upregulation correlated with dose level and distribution. In combination with other markers and timepoints, these findings contribute to a comprehensive understanding of cellular response.

Alzheimer’s disease has a long preclinical phase, during which no overt signs of the manifest disease are present, but subtle, usually non-specific changes are already detectable. Emerging early biomarkers underscore the importance of this phase for preventive measures including lifestyle interventions. As a reductionistic model for lifestyle factors, we used a novel enrichment paradigm in which App ^NL-G-F knock-in mice were continuously tracked until 7 months of age. Despite minimal plaque burden and no memory impairment at that age, there were early and progressive deficits in social parameters — such as following behavior, social interaction, and exploration – suggesting preclinical behavioral vulnerability. Altered correlations between adult neurogenesis and social parameters linked neural plasticity to preclinical behavior. Plasma profiling at 3 months identified early systemic shifts in markers of inflammation and apoptosis that predicted later cortical pathology. We found increased microglia coverage in more socially active animals. More actively exploring controls, but not App ^NL-G-F mice, exhibited more ramified and less amoeboid microglia, suggesting that AD pathology impairs immune surveillance at a very early stage. Single-cell RNA sequencing of hippocampal microglia revealed that enrichment dampened interferon-responsive microglia, which typically increase as amyloidosis advances. A shifted immune response was also measured by reduced transcripts related to antigen processing and presentation and by increased chemokine signaling. Our study demonstrates that the preclinical phase of AD is not silent, but even in a reductionistic knock-in model characterized by early interwoven preclinical changes in multiple domains, including brain plasticity, behavioral trajectories, sociality and immunity.

Preclinical studies demonstrated ultra-high dose-rate radiation treatment ability to effectively control local tumour growth while reducing damage to healthy tissues. This methodological pilot study was designed and conducted to establish an irradiation setup and to generate data for future studies investigating long-term side effects of mouse brain irradiation with proton FLASH (FLASH-PT) and conventional proton therapy (CONV-PT). Fourteen C57BL/6JRj female mice were partial brain irradiated with a single dose of 22.5 Gy using a proton spread-out Bragg peak (SOBP). Mice were randomised into FLASH-PT (268 Gy/s, n = 6), CONV-PT (0.27 Gy/s, n = 6), and unirradiated sham groups (n = 2). In addition to skin reaction monitoring as short-term endpoint, regular contrast-enhanced magnetic resonance imaging (MRI) was performed during the six-month follow-up. Brains were formalin-fixed and paraffin-embedded and stained to assess activation of microglia. The analysis of skin reactions demonstrates a trend in tissue-sparing effect for the FLASH group. In the brain MRI, no contrast enhancement was observed after irradiation. Although no group–level differences in microglia activation were observed between FLASH-PT and CONV-PT, both treatments showed increased density of activated microglia, which was distinct in specific regions within the periventricular area. The established irradiation setup allowed for a clinically relevant approach to preclinical FLASH-PT research. Partially irradiated brain showed no microglia-related FLASH effect. Further studies are required to fully comprehend dose–effect curves and timing for the FLASH effect in the brain.