key: cord-0335212-3b6q1ko5 authors: Liu, Li; Dodd, Stephen; Hunt, Ryan D.; Pothayee, Nikorn; Atanasijevic, Tatjana; Bouraoud, Nadia; Maric, Dragan; Moseman, E. Ashley; Gossa, Selamawit; McGavern, Dorian B.; Koretsky, Alan P. title: Early detection of cerebrovascular pathology and protective antiviral immunity by MRI date: 2021-10-22 journal: bioRxiv DOI: 10.1101/2021.10.21.465276 sha: 650496411abe408550b5accba3ce96c51e3d9a95 doc_id: 335212 cord_uid: 3b6q1ko5 Central nervous system (CNS) infections are a major cause of human morbidity and mortality worldwide. Even patients that survive CNS infections can have lasting neurological dysfunction resulting from immune and pathogen induced pathology. Developing approaches to noninvasively track pathology and immunity in the infected CNS is crucial for patient management and development of new therapeutics. Here, we develop novel MRI-based approaches to monitor virus-specific CD8+ T cells and their relationship to cerebrovascular pathology in the living brain. We studied a relevant murine model in which a neurotropic virus (vesicular stomatitis virus) was introduced intranasally and then entered the brain via olfactory sensory neurons – a route exploited by many pathogens in humans. Using T2*-weighted high-resolution MRI, we identified small cerebral microbleeds as the earliest form of pathology associated with viral entry into the brain. Mechanistically, these microbleeds occurred in the absence of peripheral immune cells and were associated with infection of vascular endothelial cells. We monitored the adaptive response to this infection by developing methods to iron label and track individual virus specific CD8+ T cells by MRI. Transferred antiviral T cells were detected in the brain within a day of infection and were able to reduce cerebral microbleeds. These data demonstrate the utility of MRI in detecting the earliest pathological events in the virally infected CNS as well as the therapeutic potential of antiviral T cells in mitigating this pathology. The central nervous system (CNS) is protected by several physical barriers, 40 however, a variety of neurotropic viruses from different families are able to infect the 41 CNS (Swanson & McGavern, 2015) . Viruses use multiple strategies to access the CNS. 42 Some can infect the brain through the blood brain barrier (BBB) directly, whereas others 43 enter the peripheral nervous system and use axonal transport along motor and olfactory In this study, we used intranasal infection of mice with vesicular stomatitis virus 73 (VSV) as a model (Sabin & Olitsky, 1937) to explore cerebrovascular pathology and The main goal of this study was to noninvasively define the temporal and spatial 79 relationship between cerebrovascular breakdown and CD8 T cell infiltration in the virus-80 infected brain. More specifically, our aims were to: (i) characterize early stage Detection of microbleeds in the VSV-infected turbinates and brain by MRI. 95 Upon intranasal VSV inoculation, immune cell accumulation in the turbinates and 96 olfactory bulbs (OB) has been reported (Chhatbar et al., 2018; Moseman et al., 2020) . To determine how VSV infection affects vessel integrity, high resolution T2* weighted 98 MRI was used to detect bleeding at different time points post-infection (Fig. 1) . Large 99 areas of bleeding were readily detected in the turbinates at day 4 ( Fig 1A; S2A) . The 100 amount of bleeding, as indicated by hypointense areas (red arrows), increased from day 101 4 to 11. Two major sites of bleeding were also detected by MRI in the OB ( On day 6, punctate hypointensities near the center of the OB were also detected by in-111 vivo MRI (Fig 1D) . The number of microbleeds in the OB increased from day 4 to 11. The increasing amount of hypointensities in the turbinates and OB were quantified as a 113 function of time ( Fig 1E) . OB viral titers peaked on day 6 before dropping to 114 undetectable levels by day 11 (Fig 1F) . ONL, GL and GCL were the major sites of bleeding (Fig 1G, H; S2D , E). These are also 117 the sites where VSV entered the brain and replicated. On day 4, IHC showed 118 microbleeds and VSV in the ONL, GL and external plexiform layer (EPL) of the OB (Fig 119 1G, Fig S2D) . On day 6, microbleeds and VSV were shown at the mitral cell layer 120 (MCL) (Fig 1 H1; S2E) and center of the OB (Fig 1 H2) . Most of the virus was localized 121 in ONL, GL and the center of the OB on day when viral titer was the highest (Fig 1H) . Thus, microbleeds largely co-localized with sites of VSV replication in the brain. Fig S3A, B) . Punctate hypointensities were detected from 126 frontal brain to midbrain beginning at day 4. In most of the infected mice, microbleeds 127 were detected in the orbital frontal cortex area and somatomotor areas of the cortex 128 around day 4. On day 6, microbleeds were detected in the caudoputamen (CP) and 129 thalamus (TH). From day 8 to 11, microbleeds were observed in the midbrain (MB). Through IHC study, microbleeds and VSV were observed at the forebrain on day 6 and 131 in the midbrain on day 11 (Fig 2D; S3C , D). The increasing amount of hypointensities in 132 the brain (not including the OB) were quantified as a function of time ( Fig 2C) . For the 133 OB and other brain regions, sites of bleeding were enumerated without accounting for 134 size. This was done to assess the number of new bleeding regions that developed. Foci of bleeding (defined as spots/mm 3 ) was ~100-fold greater in the OB than the brain. For these calculations, the size of mouse OB was assumed to be ~2% of the brain by treated with isotype control antibodies was on day 4, which was the same as untreated 160 9 VSV-infected mice ( Fig 1B) . Blocking immune cell infiltration also increased bleeding at 161 the center of the OB on day 3 as detected by in-vivo MRI (Fig 3D) , and a large area of 162 hemorrhage was detected in the thalamus on day 6 ( Fig 3E, F ; S4D-F). 163 As quantified in Fig 3G, the volume of hypointensity in the turbinates increased 164 from 2.1 ± 0.8 mm 3 to 5.0 ± 1.0 mm 3 on day 6 in anti-LFA-1/VLA-4 treated mice, and the 165 number of bleeding spots increased in the brain by ~2-fold. In addition, blockade of 166 immune infiltration elevated viral titers in the OB by ~2-3 fold on day 6 ( Fig 3H) . Thus, 167 VSV induces hemorrhage in the brain without involvement of the peripheral immune 185 viral clearance. 186 Inhibiting peripheral immune cells led to more brain bleeding. We therefore 187 tested if adoptive transfer of virus-specific CD8 T cells could be used to reduce viral 188 loads and inhibit brain bleeding. 5 x 10 5 virus-specific mTomato CD8+ OT-I T cells Fig 5A) , the volume of bleeding decreased from 12.2 ± 1.3 mm 3 to 1.6 196 ± 1.1 mm 3 on day 11 ( Fig 5E) . In the OB (Fig 5B) , the number of hypointensity spots 197 decreased from 665 ± 127 to 114 ± 23 on day 11 ( Fig 5F) . In the brain (Fig 5D) , the 198 number of hypointensity spots decreased from 329 ± 72 to 83 ± 19 on day 11 ( Fig 5G) . greater than the actual size of the material (Lauterbur, 1996) . This amplification is 291 critical for MRI detection of vascular effects during brain activation (BOLD functional 292 MRI), for sensitive detection of bleeding by MRI, and for many pre-clinical studies that 293 have labeled and imaged cells with MRI. Sensitivity down ~1 pg iron per voxel can be 294 achieved at high resolution. This is why we followed in-vivo MRI with ex-vivo MRI of 295 fixed brains because the latter provides significantly higher resolution to guide histology. There is the potential to detect ~10 deoxygenated RBCs, which facilitates high 297 sensitivity visualization of vessel bleeds where only a small amount of blood has 298 entered the brain. MRI was shown to be more sensitive than CT to detect bleeding. For Guided by MRI we demonstrated that neurotropic VSV can cause brain vessel 309 breakdown without the involvement of the peripheral immune system (Fig 3) . VSV spike In our study we also demonstrated that the peripheral immune response to VSV 320 infection is important for the maintenance of vascular integrity and pathogen clearance, 321 as blockade of CNS immune cell entry via administration of anti-LFA1/VLA4 antibodies 322 increased brain bleeding and viral titers (Fig 3) . Moreover, supplementing the immune We also observed that administering too many virus-specific CD8+ T cells (3 x 344 10 6 T cells) on day 6 post-infection can be pathogenic (Fig S9) . This higher dose of cells The purity of CD8 T cells was confirmed by flow cytometry (Fig S5) , using the Beckman Fluor 594 conjugated anti-mouse CD8a (2 g/mL; clone 53-6.7; BioLegend) was used. To study microglia and activation, rabbit anti-IBA-1 antibody (0.4 g/mL, FUJIFILM 472 Wako) and rat anti-CD68 antibody (5 g/mL, clone FA-11, Bio-Rad) was used. For Splenocytes isolation. Splenocytes were prepared by smashing with the textured end of a sterile 10-mL syringe followed by pipetting through a 40-micron filter into a 15-mL conical tube. The cells were collected by centrifugation at 300 g for 6 min and decanted. Cell pellets were resuspended in mouse ACK lysis buffer and waited for 10 min at room temperature followed by two washes with PBS (w/1% FBS). Upon transfer with unlabeled CD8 T cells, on day 6 post-infection, MRI could detect localized hypointensities (in-vivo) and punctate hypointensities (ex-vivo) at the edge of OB, near GL and EPL (red arrows in Fig S8A) . These hypointensities were caused by bleedings. Upon transfer with MPIO-labeled CD8 T cells, more hypointensities were detected, not only near the edge, but also in the center of the OB. These hypointensities were caused by microbleeds, plus the infiltration of MPIO-labeled T cells (red and green arrows in Fig S8B) . As quantified in Fig S8C and D , significantly more hypointensity spots from the mice transferred with MPIO-labeled T cells were detected than the mice transferred with unlabeled T cells in the OB and brain. The number of hypointensty spots in the OB transferred with unlabeled T cell and MPIOlabeled T cell were 72 ± 22 and 124 ± 20 on day 6, respectively ( Fig S8C) . When CD8 T cells were transferred on day 6, on day 11, the difference in number and intensity of the hypointensity spots between the two groups was even more significant. The combination of MRI contrast from MPIO-labeled T cells and bleeds increased sensitivity to microbleeds about 2 folds at each time point (Fig S8C, D) . Representative IHC image verified the infiltration of MPIO-labeled OT-1 CD8 T cells in the brain. Due to the similarity in contrast at this stage we cannot distinguish microbleeds from MPIO-labeled T cells. However, the total signal from MPIO-labeled T cells plus bleeds made sites of inflammation easily detectable by MRI, which might provide an approach for imaging neuroinflammation. transfer study during early-stage infection. CD8 T cells were transferred on day 0-2. Invivo MRI study was performed from day 1 to 6. The encephalitis peak time was day 6, which was labeled in red. Remarkable amount of CD45+ cells infiltration into the OB was reported since day 4, which was labeled in purple. (B) Paradigm for CD8 T cell transfer study during peak-stage infection. CD8 T cells were transferred on day 6. Invivo MRI study was performed from day 7 to 11. Multifield and inverse-contrast switching 603 of magnetocaloric high contrast ratio MRI labels The distinct temporal 606 origins of olfactory bulb interneuron subtypes Vesicular stomatitis virus with the rabies virus glycoprotein directs retrograde 610 transsynaptic transport among neurons in vivo. Frontiers in Neural Circuits On the pathogenic role of brain-615 sequestered alphabeta CD8+ T cells in experimental cerebral malaria Imaging of genetically 619 engineered T cells by PET using gold nanoparticles complexed to Copper-64 Receptor Signaling of Neurons and Astrocytes Regulates Microglia Activation 625 during Viral Encephalitis New Jersey serotype) 629 infection in the deer mouse (Peromyscus maniculatus) Local Type I IFN Receptor Signaling Protects against Virus 633 Spread within the Central Nervous System Immune Responses and Anti-inflammatory Strategies in a Clinically Relevant Model of Thromboembolic Ischemic Stroke with Reperfusion Magnetic resonance imaging assessment of 641 macrophage accumulation in mouse brain after experimental traumatic brain 642 injury Traumatic microbleeds suggest vascular 646 injury and predict disability in traumatic brain injury Imaging of central nervous 649 system viral diseases Cerebral Microbleeds: Imaging and Clinical Significance Therapeutic antiviral T cells 655 noncytopathically clear persistently infected microglia after conversion into 656 antigen-presenting cells Cell penetrating peptide functionalized perfluorocarbon nanoemulsions 660 for targeted cell labeling and enhanced fluorine-19 MRI detection Distribution of vesicular stomatitis virus proteins in the brains 664 of BALB/c mice following intranasal inoculation: an immunohistochemical 665 analysis Myelomonocytic cell 668 recruitment causes fatal CNS vascular injury during acute viral meningitis Protective and Pathological Immunity during 671 Central Nervous System Infections Microscopic NMR Imaging of the Magnetic Field around Magnetic Particles Proceedings of the Society for Magnetic Resonance in Medicine, 5th Annual 676 Meeting Microvascular 680 Injury in the Brains of Patients with Covid-19 Zika Virus Infects Neural 684 Progenitors in the Adult Mouse Brain and Alters Proliferation Long-distance neuronal migration in the adult 687 mammalian brain Neurotropic virus infections as the cause of 693 immediate and delayed neuropathology New advances in CNS immunity against viral 696 infection Poor human olfaction is a 19th-century myth Illuminating viral infections in the nervous 701 system Lipoprotein Receptor on Endothelium of Brain Capillaries T cell 708 engagement of cross-presenting microglia protects the brain from a nasal virus 709 infection Structural basis for the recognition of LDL-receptor family members by VSV 713 glycoprotein Outgrowth Endothelial Cells as a Cellular Carrier for Oncolytic Vesicular 717 Stomatitis Virus Expressing Interferon-O in Preclinical Models of Non Magnetic resonance imaging of odorant activity-722 dependent migration of neural precursor cells and olfactory bulb growth COVID-725 19-associated Acute Hemorrhagic Necrotizing Encephalopathy: Imaging 726 Features CD8+ T cells target cerebrovasculature in children with cerebral malaria Laminar specific detection 733 of APP induced neurodegeneration and recovery using MEMRI in an olfactory 734 based Alzheimer's disease mouse model Influence of Host Factors on Neuroinvasiveness of 737 Vesicular Stomatitis Virus : I. Effect of Age on the Invasion of the Brain by Virus 738 Instilled in the Nose MRI detection of single particles for cellular imaging Induce Fatal Brainstem Pathology during Cerebral Malaria via Luminal Antigen-754 Specific Engagement of Brain Vasculature Viral diseases of the central nervous 757 system Self-assembling nanocomplexes by combining ferumoxytol, heparin and 761 protamine for cell tracking by magnetic resonance imaging In situ labeling of immune cells with iron oxide particles: An approach to 765 detect organ rejection by cellular MRI Inhibition of LFA-1/ICAM-1 and VLA-4/VCAM-1 as a therapeutic 770 approach to inflammation and autoimmune diseases Ellipsoidal Microcavities: 773 Electromagnetic Properties, Fabrication, and Use as Multispectral MRI Agents