Quantification of hypoxic regions distant from occlusions in cerebral penetrating arteriole trees

The microvasculature plays a key role in oxygen transport in the mammalian brain. Despite the close coupling between cerebral vascular geometry and local oxygen demand, recent experiments have reported that microvascular occlusions can lead to unexpected distant tissue hypoxia and infarction. To better understand the spatial correlation between the hypoxic regions and the occlusion sites, we used both in vivo experiments and in silico simulations to investigate the effects of occlusions in cerebral penetrating arteriole trees on tissue hypoxia. In a rat model of microembolisation, 25 µm microspheres were injected through the carotid artery to occlude penetrating arterioles. In representative models of human cortical columns, the penetrating arterioles were occluded by simulating the transport of microspheres of the same size and the oxygen transport was simulated using a Green’s function method. The locations of microspheres and hypoxic regions were segmented, and two novel distance analyses were implemented to study their spatial correlation. The distant hypoxic regions were found to be present in both experiments and simulations, and mainly due to the hypoperfusion in the region downstream of the occlusion site. Furthermore, a reasonable agreement for the spatial correlation between hypoxic regions and occlusion sites is shown between experiments and simulations, which indicates the good applicability of in silico models in understanding the response of cerebral blood flow and oxygen transport to microemboli. Author summary The brain function depends on the continuous oxygen supply through the bloodstream inside the microvasculature. Occlusions in the microvascular network will disturb the oxygen delivery in the brain and result in hypoxic tissues that can lead to infarction and cognitive dysfunction. To aid in understanding the formation of hypoxic tissues caused by micro-occlusions in the penetrating arteriole trees, we use rodent experiments and simulations of human vascular networks to study the spatial correlations between the hypoxic regions and the occlusion locations. Our results suggest that hypoxic regions can form distally from the occlusion site, which agrees with the previous observations in the rat brain. These distant hypoxic regions are primarily due to the lack of blood flow in the brain tissues downstream of the occlusion. Moreover, a reasonable agreement of the spatial relationship is found between the experiments and the simulations, which indicates the applicability of in silico models to study the effects of microemboli on the brain tissue.

. Microspheres were lodged in the left 142 hemisphere, leaving the right hemisphere as a control, as depicted in Fig. 1(c). 146 µm thick) were made between 1-3.5 mm bregma coordinates; (c) Ten consecutive coronal brain sections were selected for 147 the brain damage reconstruction. Brain tissue damage (in red) was mainly confined on the intervention hemisphere; (d) 148 Maximum intensity projection (MIP) of a 50 µm thick coronal brain section of the intervention hemisphere with hypoxic 149 regions (white). Tilescan image was made using confocal imaging, Scale bar 1000 µm; (e) The same brain section as in (d)  159 smaller than 10 mmHg and lectin is a dye which stains perfused blood vessels intravenously.

165 2.1.3 3D reconstruction of intervention hemisphere
166 For the spatial analysis between microspheres and brain damage, a 500 μm thick volume of 167 the intervention hemisphere was reconstructed. Ten consecutive coronal brain sections (50 168 μm thick) devoid of tearing were selected from the forebrain between 1 and 3.5 mm of the 169 bregma as shown in Fig. 1(b)-(c). Tilescan z-stack images (resolution: x,y: 3.033 μm and z: 5 170 μm) of the intervention hemisphere were acquired using a confocal laser scanning 171 microscope SP8 (Leica Microsystems, Wetzlar, Germany) with a 10x objective. To facilitate 172 alignment, the z-stack images were converted to maximum intensity projection (MIP) 173 images as depicted in Fig. 1(d). To this end, the ImageJ software (Rasband, W.S., ImageJ, U. 267 where α is the blood oxygen solubility, is the blood PO 2 , is the oxygen binding 268 capacity per unit volume of red blood cells and is the oxygen saturation of haemoglobin. 278 where , and are the oxygen diffusion coefficient, the oxygen solubility and PO 2 in 279 brain tissue. Metabolic rate of oxygen ( ) and tissue PO 2 are assumed to follow a Michaelis-      433 434 Next, we tested whether there was a correlation between microspheres and hypoxic pixels.  Fig. 7(a) but using a different capillary network (Fig. S2), and the other in a cortical 467 column using the same capillary geometry but connected to a different arteriole tree (Fig.   468 S3). The change in capillary geometry or arteriole geometry leads to a root-mean-square 469 difference of 9.7 or 15.9 mmHg for PO 2 in each tissue voxel in the column, respectively.
470 However, these local variations were found to only have negligible effects on the overall PO 2 471 distribution in the column (Fig. S4). 500 The pixel-based Gx function has a consistent sigmoidal shape in each animal in experiments 501 ( Fig. 9(a)) and in each cortical column in simulations ( Fig. 9(b)). This results in small standard 502 deviations between each animal and between each cortical column as shown in Fig. 9(c). In 503 addition, there is a close match for the pixel-based Gx function between experiments and 504 simulations in that they both reach 50% at around 300 m and reach about 90% at 800 m.
505 However, the experimental curve is found to start to increase at shorter distances than in 506 the simulations, which indicates that there are more hypoxic regions in the vicinity of 507 occlusion sites in experiments than was found in the simulations. 508 509 These discrepancies are more clearly shown in the hypoxic intensity results. The 510 experimental hypoxic intensity curves decrease as the radius increases ( Fig. 9(d)), however, 511 most of the simulated hypoxic intensity curves increase at the start ( Fig. 9(e)). The reason 512 for this difference is that the experiments have more type A hypoxic intensity (30.9%), while 513 there is only one case out of 91 simulations that is type A (Fig. S5). This leads to some 514 significant differences between experiments and simulations ( Fig. 9(f)). The standard 515 deviation of the hypoxic intensity is also larger than that of the pixel-based Gx function in 516 both experiments and simulations. These results will be further discussed in the next 517 section.  555 overlap between ischaemia and hypoxia. In our previous work where we injected a mixture 556 of microsphere sizes and killed the animals after 1, 3 and 7 days we found that infarction 557 volume at day 7 was similar to that of day 1, suggesting that infarction develops within 24h 558 after microembolisation [17]. As a consequence, we missed in our data the cells which had 559 already undergone cell death. This can explain the ischaemic regions which were not 560 hypoxic. In cases where hypoxic regions did not overlap with ischaemia, we think that larger 561 ischaemic regions, which are formed due to multiple occlusions of the same or different 562 arterial trees, are likely responsible for these results. Hypoxic regions could span beyond the 563 analysed tissue (500 μm thick). As a result, we detect only the hypoxia in our analysed brain 564 tissue, while the ischaemic source is further away in the z-direction. Considering the 565 pathological process of infarct growth and the distal effects of large ischaemic regions, the 566 overlap between ischaemia and hypoxia in our experiments is reasonable.

567
568 Distal hypoxic regions. In the experimental data we found both local (<100 µm, 50.8% of all 569 cases) and distal (>100 µm, 49.2% of all cases) effects due to micro-occlusions. Contrary to 570 previous in vivo rodent studies where the occlusion site was highly correlated to brain tissue 571 damage [22-24], we found distal effects after occlusion of penetrating arteriole branches 572 using a rat model of microembolisation [41]. Our simulations suggest that a 25 µm 573 microsphere will occlude a branch of the penetrating arteriole, which will lead to 574 hypoperfusion downstream of the occlusion site in the corresponding cortical column (Fig.   575 8). The hypoperfusion will then result in distal tissue hypoxia from the occlusion site. This is 576 partially supported by the overlap between ischaemic and hypoxic regions in the 577 experiments. 590 591 To understand the discrepancy found between microembolisation and the photothrombotic 592 model we address here some differences between the two techniques. Firstly, in the current 593 experimental study, microspheres are lodged not only in the cortex but also in deep brain