Transcriptomic Response of Brain Tissue to Focused Ultrasound-Mediated Blood-Brain Barrier Disruption Depends Strongly on Anesthesia

Focused ultrasound (FUS) mediated blood brain barrier disruption (BBBD) is a promising strategy for the targeted delivery of systemically-administered therapeutics to the central nervous system (CNS). Pre-clinical investigations of BBBD have been performed on different anesthetic backgrounds; however, the potential influence of the choice of anesthetic on the molecular response to BBBD is unknown, despite its potential to critically affect interpretation of experimental therapeutic outcomes. Here, using bulk RNA sequencing approaches, we comprehensively examined the transcriptomic response of both normal brain tissue and brain tissue exposed to FUS-induced BBBD in mice anesthetized with either isoflurane with medical air (Iso) or ketamine/dexmedetomidine (KD). In normal murine brain tissue, Iso alone elicited minimal differential gene expression (DGE) and repressed pathways associated with neuronal signaling. KD alone, however, led to massive DGE and enrichment of pathways associated with protein synthesis. In brain tissue exposed to BBBD (1 MHz, 0.5 Hz pulse repetition frequency, 0.4 MPa peak-negative pressure), we systematically evaluated the relative effects of anesthesia, microbubbles, and FUS on the transcriptome. Of particular interest, we observed that gene sets associated with sterile inflammatory responses and cell-cell junctional activity were induced by BBBD, regardless of the choice of anesthesia. Meanwhile, gene sets associated with metabolism, platelet activity, tissue repair, and signaling pathways, were differentially affected by BBBD, with a strong dependence on the anesthetic. We conclude that the underlying transcriptomic response to FUS-mediated BBBD may be powerfully influenced by anesthesia. These findings raise considerations for the translation of FUS-BBBD delivery approaches that impact, in particular, metabolism, tissue repair, and intracellular signaling.


107
Characterization of FUS-Induced BBBD and Passive Cavitation Analysis 108 Mice were anesthetized with either isoflurane in medical air (Iso) or 109 ketamine/dexmedetomidine (KD) and treated with Magnetic Resonance-guided Focused 110 Ultrasound (MRgFUS) targeted to the right or left striatum. To assess the extent and localization 111 of BBBD, contrast-enhanced images were collected before and after treatment ( Figure 1A). The 112 magnitude of signal enhancement was significantly greater in mice anesthetized with Iso 113 compared to KD with respect to fold difference ( Figure 1B) in mean grayscale intensity in treated 114 vs untreated hemispheres. To evaluate differences in oscillatory activity of circulating MB in 115 response to FUS, we analyzed acoustic emissions data obtained from a listening hydrophone 116 embedded in the therapeutic transducer. Steady oscillation of MB, called stable cavitation, imparts 117 the mechanical forces on vessel walls needed to disrupt the BBB and produces concomitant 118 peaks at harmonics (2f, 3f, 4f, f = operating frequency of the treatment transducer). Meanwhile, 119 unstable oscillation and violent collapse of MB, called inertial cavitation, can produce concomitant 120 broadband signal (in-between harmonics) in the Fourier domain. No significant differences in 121 stable cavitation (as measured by 2 nd , 3 rd , 4 th harmonics) or inertial cavitation (broadband 122 emission up to 10 MHz) were found between Iso and KD ( Figure 1C).

124
Transcriptomic Variation is Driven Primarily by KD and Secondarily by FUS BBBD 125 anesthetic alone, and mice treated with each anesthetic and microbubbles were also sequenced 128 6 h after treatment. After read alignment and QC, principal components analysis (PCA) was 129 performed on transformed transcript counts from each sample to assess global differences 130 between treatment conditions (Figure 2A). Interestingly, the first principal component segregated 131 samples by whether they received KD, with Iso-treated mice clustering more closely to the naïve 132 controls. FUS-treated mice formed a distinct cluster only in the KD treated mice. Similar results 133 were obtained when hierarchical clustering was performed on inter-sample Euclidian distances 134 computed between samples based on their transcript counts ( Figure 2B). With the exception of 135 one sample, the first branch point of the dendrogram separated samples by KD status, while the 136 second and third branch points distinguished samples by FUS treatment. 137  21 contrasts of the 7 conditions tested. Each row represents a numerator condition and each column represents a denominator condition. (D) Magnitude of significantly repressed (left) and enriched pathways (right) for all 21 contrasts of the 7 conditions tested. Each row represents a numerator condition and each column represents a denominator condition. For all genes and pathways, significance is defined as p-adjusted < 0.05.

138
Overview of Differential Gene Expression and Gene Set Enrichment Analyses 139 To evaluate relative transcriptomic differences between conditions, differential gene 140 expression contrasts were computed for all 21 unique combinations of the 7 conditions evaluated 141 ( Figure 2C). KD alone produced the most profound effect on the transcriptome, with over 3000 142 genes significantly differentially regulated (p-adjusted < 0.05) compared to naïve brain. 143 Regardless of the anesthetic background, FUS and MB produced moderate (on the order of 144 hundreds of differentially expressed genes) and negligible (< 9 differentially expressed genes) 145 effects on gene expression respectively. Iso alone had a marginal effect on the transcriptome, 146 only significantly changing the expression of 26 genes. Next, we performed gene set enrichment 147 analysis (GSEA) to identify biological processes consistent with genes differentially expressed 148 within each contrast ( Figure 2D). GSEA was performed using the Gene Ontology (GO) Biological 149 Pathways database, wherein each "GO" term represents a collection of genes associated with a 150 particular biological phenomenon. Surprisingly, Iso alone affected more biological pathways than 151 KD, despite KD affecting considerably more genes. The addition of MB changed relatively few 152 biological pathways. FUS had the strongest effect on biological pathways on both anesthetic 153 backgrounds, inducing more pathways than it repressed.

Anesthetics Differentially Affect the Transcriptome of Normal Brain Tissue 156
The relative transcriptional impact of Iso and KD on the mouse striatum was marked, with 157 Iso significantly changing expression of 26 genes compared to the 3,291 significantly changed by 158 KD (Figure 3A). Iso alone induced a traditional anesthetic transcriptional program of repression 159 of neuronal activity ( Figure 3B). KD, however, had a minimal effect on these pathways, instead 160 enriching for steps of protein synthesis and targeting ( Figure 3C). These trends persisted upon 161 addition of MB or FUS. To assess the effect of anesthesia on neuroinflammation, we examined 162 GO processes related to inflammation differentially changed by Iso or KD alone ( Figure 3D) differentially expressed genes ( Figure 4A) and differentially regulated pathways ( Figure 4B), 180 controlling for changes due to anesthesia + MB alone. While more genes were differentially 181 regulated by KD-FUS, more gene sets were significantly enriched/repressed by Iso-FUS. 182 Interestingly, despite minimal intersection of transcript identities between the two BBBD 183 conditions, 41% of the pathways significantly induced by KD-FUS were also significantly induced 184 by Iso-FUS. Second, we identified 6 categories of biological pathways consistently changed by 185 Iso-FUS, KD-FUS, or both ( Figure 4C). Regardless of the anesthetic background, FUS led to 186 enrichment of genes involved in endothelial cell activity, including pathways associated with cell-187 cell adhesion and angiogenesis. Iso-FUS induced these pathways more significantly, and 188 additionally led to the expression of genes associated with leukocyte adhesion. Similarly, both 189 FUS conditions led to activation of many inflammation pathways, with the breadth and depth of 190 these responses substantially enhanced in the Iso-FUS condition. Notably, the MHC class I and 191 MHC class II antigen processing and presentation pathways were only upregulated when 192 comparing KD-FUS treated mice to naïve controls. We found the most significant divergence 193 between Iso-FUS and KD-FUS when comparing metabolic pathways. Iso-FUS led to repression 194 of broad and specific metabolic programs while several of these were enriched by  Consistent with significant inflammation and endothelial activation, platelet activity was enhanced 196 by Iso-FUS, while these pathways were relatively unchanged by KD-FUS. Gene sets associated 197 with tissue repair were enriched by FUS under both anesthetics and those associated with 198 neurogenesis were additionally upregulated by KD-FUS only. Signaling pathways engaged by 199 FUS treatment independent of anesthesia included VEGFR signaling, Wnt signaling, and the NF-  these transcripts were significantly upregulated by KD-FUS across multiple contrasts, only Cdh5 213 was significantly upregulated by FUS under Iso. Notably, when compared to naïve controls alone, 214 KD alone significantly downregulated Flt1 and KD + MB led to a trending decrease (p-adj = 0.06).

215
We next compared the LEA overlap on the Immune System Process gene set (GO:0002682), a 216 broad collection of 1709 genes associated with the immune system ( Figure 4E). Iso-FUS and 217 KD-FUS enriched 512 and 304 of these respectively, with 103 genes enriched by both. IL-1α was 218 found in both LEAs and significantly upregulated across multiple contrasts while IL-1β was only 219 found in the Iso-FUS LEA and indeed only significantly upregulated in Iso-only FUS contrasts. 220 TNFα was found in both LEAs to be significantly upregulated by FUS under both anesthetics 221 when compared to naïve controls, and trending upward in other FUS contrasts. To narrow the 222 scope of immune system-related LEA overlaps, we repeated this analysis on the Chemokine 223 Activity gene set (GO:0008009) which only contains 34 genes ( Figure 4F). Iso-FUS and KD-FUS 224 enriched 16 and 12 chemokines respectively, 7 of which were shared. Iso-FUS induced the 225 strongest Ccl2 upregulation regardless of the control condition. KD alone induced a comparable 226 upregulation of Ccl2 with no additional effect due to FUS. Cxcl16 however was more strongly 227 induced with KD-FUS than Iso-FUS when controlling for anesthetic. Ccl3 was upregulated by FUS 228 under both anesthetics as well as KD alone. In summary, while FUS promotes phenotypes such 229 as cell junction organization, inflammation, and chemokine activity independent of anesthetic, the 230 nature of the transcripts mediating these effects are often anesthesia-dependent. 231

232
Anesthetics Differentially Affect Transcripts Associated with BBB Structure and Function 233 We next evaluated the effects of anesthesia, MB, and FUS on transcripts known to be 234 associated with the BBB [32]. Iso-FUS upregulated transcripts mediating leukocyte adhesion, 235 including E-selectin, P-selectin, and Icam1 ( Figure 5A). Icam1 was also upregulated by KD alone 236 when compared to sham and by KD-FUS when compared to KD or KD + MB. With respect to 237 BBB tight junction transcripts, FUS upregulated Cldn5 and Emp1 independent of anesthetic 238 ( Figure 5B). KD alone led to downregulation of Ocln and Tjp1. We next evaluated the effect of 239 our experimental conditions on BBB transporter transcripts and observed heterogeneous effects 240 ( Figure 5C). In general, KD led to significantly more DGE in this category than Iso, with very few 241 transcripts changing their expression due to FUS or MB on either anesthetic background. This 242 trend was even more extreme when evaluating BBB transcripts involved in transcytosis and other 243 miscellaneous functions ( Figure 5D); KD was the only variable significantly changing the 244 expression of transcripts in this class.

246 Tissue Damage Elicited by FUS BBBD is Minimal and Not Affected by Anesthetic, 247
Given the anesthesia-dependence of BBBD and FUS-induced gene expression, we next 248 tested whether anesthesia significantly affected the extent of damage in the brain parenchyma 249 after treatment with the same FUS pressure. To address this, we performed histological analysis 250 of murine brains treated with combinations of Iso, KD, and FUS (Figures 6A-D). Brains treated 251 with 0.8 MPa (twice the acoustic pressure of our standard BBBD protocol) were used as positive 252 controls for damage. We scored multiple transverse sections from each condition for RBC 253 extravasation and vacuolation ( Figure 6E). With the exception of the 0. selection for preclinical BBBD studies and will likely find utility when comparing clinical outcomes 273 to pre-clinical results for FUS mediated BBBD drug and gene delivery approaches.

274
As shown in Figure 1,  with respect to CNS gene expression (Figures 2A-B). The most striking of these was KD, 298 inducing DGE (p-adjusted <.05) of 3291 genes when compared to naïve controls ( Figure 3A).

299
Whether this profound change in gene expression is attributable to ketamine, dexmedetomidine, 300 or both is unclear. Microarray studies of developing rat brain have shown a similar magnitude of 301 acute differential gene expression from ketamine alone [44]. More specifically, investigators 302 reported 819 differentially expressed genes with fold change >1.4, p-adj < 0.05 compared to the 303 1182 meeting these criteria in our study at an identical timepoint. Though ketamine's mechanism 304 of action is still unclear, recent studies into its rapid anti-depressant action suggest ketamine 305 indirectly suppresses eukaryotic elongation factor 2 kinase (eEf2K), leading to increased protein 306 translation [45]. This mechanism is in agreement with our pathway level findings ( Figure 3C).

307
Though fewer transcriptomic level studies exist for dexmedetomidine, it is known to acutely 308 augment transcriptional programs associated with inflammation and circadian rhythm [46,47]. In 309 stark contrast to KD, we found Iso had a negligible impact on gene expression, only significantly 310 altering the expression of 26 genes. This finding is in close agreement with existing acute 311 transcriptomic studies of inhalable anesthetics in rats, which report between 0 and 20 differentially 312 expressed genes [48,49]. Interestingly, despite weak changes in expression magnitude, Iso 313 changed regulation of significantly more pathways than KD ( Figure 2D). We thus hypothesize 314 that, while Iso influences more targeted transcriptional programs, the combination of ketamine 315 and dexmedetomidine elicits wide-ranging, complex transcription thereby preventing GSEA from 316 detecting discrete pathway enrichment.

317
We observed increases in inflammatory signatures elicited by both anesthetics (Figure  318  3D). Of the few genes upregulated by Iso alone, a surprising number were immune-associated.

319
Some examples include upregulation of T-cell associated markers Ly6a and Ctla2a, upregulation 320 of adhesion markers Pecam1 and CD93, and downregulation of Nfkbia, the protein product of 321 which inhibits NF-κB. Indeed, activation of NF-κB has been proposed as a mechanism by which 322 volatile anesthetics elicit neuroinflammation [50,51]. Several rodent studies have demonstrated 323 volatile anesthetics can also acutely induce expression of IL-6, IL-1β, and activated caspase-3 324 [52][53][54][55]. It is worth noting that under conditions of CNS stress, including ischemia or LPS 325 exposure, volatile anesthetics have been shown to attenuate inflammation, suggesting that these 326 drugs may contribute to maintaining homeostasis in the brain, rather than being strictly pro-or 327 anti-inflammatory [56-59]. KD also induced signatures associated with inflammation, though to 328 a lesser extent and with a less clear mechanism than Iso. At the chemokine level, for example, 329 we found KD significantly upregulated Ccl17, Ccl2, Ccl3, and Ccl6 with minor but significant 330 downregulation of Cxcl12 and Cx3cl1. These mixed effects may be caused by contrasting 331 neuroinflammatory effects produced by ketamine and dexmedetomidine. Ketamine has been 332 shown to be acutely inflammatory in naïve mice, increasing levels of IL-6, IL-1β, and of these mechanisms is implicated, and can be affected by choice of anesthetic ( Figure 4C). between FUS application and SI. Enrichment of junctional assembly pathways, VEGF signaling, 358 and angiogenesis supports FUS-induced activation of endothelial cells, leading to both 359 recruitment of leukocytes and barrier repair, especially under Iso. Of note, we observed significant 360 upregulation of claudin-5 transcript, whose tight junction protein product is essential to BBB 361 integrity, in both FUS groups. This may indicate initiation of transcriptional programs to repair the 362 disrupted barrier (Figure 5B). In contrast, a microarray study of brain microvessels did not detect 363 significant differences in claudin-5 post-FUS [25]. This discrepancy could be due to differences in 364 species (i.e. mouse vs. rat), the source of the analyzed tissue in the brain, anesthesia protocol, 365 and several focused ultrasound and microbubble parameters. Downregulation of multiple 366 metabolic pathways in Iso-FUS contrasts further suggests Iso may prime the BBB for more 367 significant alteration than KD. 368 Despite such differential responses at the transcriptional level, FUS applied under both 369 anesthetics led to little to no generation of petechiae by H&E (Figure 6). With respect to 370 coagulation signatures by RNA-seq, only Iso-FUS led to increased platelet activity despite no 371 significant difference in RBC extravasation compared to KD-FUS ( Figure 4C) signatures of inflammation, endothelial activation, coagulation, and metabolic alteration supports 383 its use over Iso for pathologies where further CNS stress is undesirable.

384
Our investigation has some limitations. First, RNA-sequencing only provides transcript-385 level information and several studies highlight that mRNA may not always correlate proportionally 386 to protein expression [80][81][82][83]. This risk is mitigated at the pathway level, where we present 387 significant alteration of large families of genes consistently up or downregulated by FUS and/or 388 anesthesia. We further assert that the high intragroup consistency along with the absolute 389 magnitude of differential gene or pathway level changes we present make noise an unlikely driver 390 of the diverse changes we observe. However we also note that because RNA-seq was performed 391 on bulk tissue, it is not easy to distinguish changes in transcription from changes in relative cell 392 numbers. Protein and phenotypic studies may provide additional insight into the consequences 393 of the results generated herein. Next, whether transcriptional changes in Iso-FUS mice are a 394 consequence of isoflurane's interaction with FUS or enhanced BBB permeability is unclear.

395
Finally, not all experiments were performed on the same FUS-system. Though transducer 396 frequencies and acoustic pressures were matched between systems, it is possible that 397 differences in transducer geometries produced confounders in experimental endpoints.

398
We present here a detailed account of how Iso and KD, the two most commonly used however, induced sweeping transcriptome changes alone, but blunted markers of SI while 406 promoting gene sets associated with tissue repair upon FUS application compared to Iso-FUS.

407
These results provide important context for previous preclinical FUS studies, and underscore 408 anesthesia as an important experimental variable to consider for future work. More research is 409 required to understand whether the findings described herein are maintained at the protein level 410 and how anesthesia-dependent responses to FUS evolve with varying FUS parameters, MB 411 characteristics, and time. Mice in groups designated "KD" received 50-70 mg/kg Ketamine and 0.25-0.5 mg/kg 424 Dexmedetomidine via intraperitoneal injection with no additional maintenance or reversal drug 425 given. Mice in groups designated as "Iso" or "Iso-MA" were placed in an induction chamber and 426 received isoflurane delivered to effect in concentrations of 2.5% in medical air using a vaporizer. 427 For isoflurane groups, anesthesia was maintained via nosecone for a total of 90 minutes.

429
MRgFUS mediated BBBD 430 Once anesthetized, a tail vein catheter was inserted to permit intravenous injections of 431 MBs and the MRI contrast agent. The heads of the mice were shaved and depilated, and the 432 animals were then placed in a supine position over a degassed water bath coupled to an MR-433 compatible small animal FUS system (RK-100; FUS Instruments, Toronto, Canada). The entire 434 system was then placed in a 3T MR scanner (Magnetom Prisma; Siemens Medical Solutions, 435 Malvern, Pennsylvania). A 3.5 cm diameter receive RF coil, designed and built in-house, was 436 placed around the head to maximize imaging SNR. Baseline three-dimensional T1-weighted MR 437 images were acquired at 0.3 mm resolution using a short-TR spoiled gradient-echo pulse 438 sequence and used to select 4 FUS target locations in and around the right or left striatum. 439 Mice received an injection of albumin-shelled MBs (1 x 10 5 MBs/g b.w.), formulated as 440 previously described [14,84,85]. Sonication began immediately after clearance of the catheter. 441 Sonications (4 spots in a 2x2 grid) were performed at 0.4 MPa peak-negative pressure (PNP) 442 using a 1.1 MHz single element focused transducer (FUS Instruments, Toronto, Canada) 443 operating in 10 ms bursts, 0.5 Hz pulse repetition frequency and 2 minutes total duration. 444 Immediately following the FUS treatment, mice received an intravenous injection of gadolinium-445 based contrast agent (0.05 ml of 105.8 mg/ml preparation; Multihance; Bracco Diagnostics), and 446 contrast-enhanced images were acquired to assess BBBD using the same T1-weighted pulse 447 sequence mentioned above.

449
Passive Cavitation Detection 450 Acoustic emissions were detected with a 2.5 mm wideband unfocused hydrophone 451 mounted in the center of the transducer. Acoustic signal was captured using a scope card 452 (ATS460, Alazar, Pointe-Claire, Canada) and processed using an in-house built MATLAB 453 (MathWorks) algorithm. Acoustic emissions at the fundamental frequency, harmonics (2f, 3f, 4f), 454 sub harmonic (0.5f), and ultra-harmonics (1.5f, 2.5f, 3.5f) were assessed by first taking the root 455 mean square of the peak spectral amplitude (Vrms) in each frequency band after applying a 200 456 Hz bandwidth filter, and then summing the product of Vrms and individual sonication duration over 457 the entire treatment period. Broadband emissions were assessed by summing the product of 458 Vrms 120 s, 10-ms bursts, 0.5-Hz burst rate) was targeted to the right striatum. The 6-dB acoustic 483 beamwidths along the axial and transverse directions are 15 mm and 4 mm, respectively. The 484 waveform pulsing was driven by a waveform generator (AFG310; Tektronix) and amplified using 485 a 55-dB RF power amplifier (ENI 3100LA; Electronic Navigation Industries). 486 Once anesthetized, a tail-vein catheter was inserted to permit i.v. injections of MBs and 487 Evans Blue. The heads of the mice were shaved and depilated, and the animals were then 488 positioned prone in a stereotactic frame (Stoelting). The mouse heads were ultrasonically coupled 489 to the FUS transducer with ultrasound gel and degassed water and positioned such that the 490 ultrasound focus was localized to the right striatum. Mice received an i.v. injection of the MBs (1 491 x 10 5 MBs/g b.w.) and Evans Blue, followed by 0.1 mL of 2% heparinized saline to clear the 492 catheter. Sonication began immediately after clearance of the catheter. In contrast to the MR-493 guided experiments, which targeted four spots, only one location was targeted in these studies 494 due to the increased focal region of the transducer (4 mm in the transverse direction, relative to 495 1 mm for the transducer in the MR-compatible system).

497
Histological Processing and Analysis 498 60 minutes after Evans Blue injection, mice were euthanized via an overdose of 499 pentobarbital sodium and phenytoin sodium. A macroscopic image was taken immediately after 500 whole brain harvest. Brains were then placed in 10% NBF, embedded in paraffin, and sectioned 501 400 µm apart. H&E stained sections were imaged with 4x and 20x objectives on an Axioskop light 502 microscope (Zeiss, Germany) equipped with a PROGRES GRYPAX microscope camera 503 (Jenoptik, Germany). 10 20x images from the region of the right striatum with maximal Evans 504 Blue extravasation were taken per section and 2 -6 sections were imaged per brain. A researcher 505 blinded to treatment condition assigned a score of 0 (none), 1 (mild), 2 (moderate), or 3 (severe) 506 to each 20x image for RBC extravasation and vacuolation using a custom MATLAB (MathWorks) 507 script. 508 509 Acknowledgements.  References 531 532 [1] W.M. Pardridge, The blood-brain barrier: bottleneck in brain drug development., 533 NeuroRx.