Synopsis

Keywords: Diffusion/other diffusion imaging techniques, Permeability

Non-disruption alterations to the blood-brain barrier (BBB) can be difficult to detect and therefore require highly sensitive tools for reliable measurement. Here, we apply a BBB filter exchange imaging (BBB-FEXI) technique to assess the rat brain in response to mild Streptococcus pneumoniae lung infection. We observe a significant 78 ± 39 % increase in BBB water permeability during infection. Higher water exchange measures were associated with higher levels of vascular inflammation, while BBB tight junction proteins remained unchanged. The expression of aquaporin-4 water channel was 38% higher in infected animals, which may drive the increase in water exchange during infection.

INTRODUCTION

Blood-brain barrier (BBB) dysfunction occurs in many brain diseases, and there is increasing evidence that it is an early process in neurodegeneration [1], which may be exacerbated by peripheral infection [2]. Imaging studies have been able to detect BBB dysfunction caused by widespread inflammation from lupus [3, 4], but subtle, non-disruptive BBB alterations due to peripheral infection, can be difficult to identify, and therefore require highly sensitive tools for reliable measurement. Filter-exchange imaging (FEXI) is a promising technique that can be adapted to probe water exchange across BBB (here termed “BBB-FEXI”), as a potentially more sensitive measure of BBB alteration. BBB-FEXI has been successfully implemented in the human brain [5, 6] and attempts have been made to target brain tumours [7], but it has yet to be applied to examine subtle pathology. We have developed a preclinical BBB-FEXI approach to investigate the effects of Streptococcus pneumoniae lung infection on BBB function. We obtain measures of BBB water exchange in the rat brain before and during infection and validate against ex-vivo BBB markers: BBB tight junction proteins, aqupaporin-4 water channel protein and vascular inflammation marker, von Willibrand factor (VWF). This work demonstrates the potential of BBB-FEXI as a non-invasive tool to detect subtle BBB alterations in response to lung infection and helps to further our understanding of the impact of peripheral infection on the BBB.

METHODS

To assess the effects of peripheral infection on BBB water permeability, F334 rats (n = 14) were scanned, with a Bruker Avance III console interfaced with an Agilent 7T 16-cm bore magnet using our preclinical BBB-FEXI sequence, before infection (baseline) and again on day 8-9 of an ascending Streptococcus pneumoniae infection challenge. Imaging parameters: filter b-values (b_f) = 0, 250 s/mm²; detection b = 0, 250 s/mm²; mixing times (t_m) = 0.025, 0.05, 0.1, 0.2 and 0.3 s; TR = 5000 s; matrix size = 64 x 64; FOV = 32 x 32 mm²; single slice, resolution = 0.5 x 0.5 x 4.0 mm²; repetitions = 10, with a spin-echo EPI readout. Whole brain BBB water exchange (kin) estimates were obtained from ADC’(tm) measurements fitted with a novel crusher compensated exchange rate (CCXR) model.

VWF levels, a peripheral marker for vascular injury and inflammation were estimated from blood plasma samples taken on the day after MRI from a subset of rats (n = 9). BBB tight junction proteins (claudin-3, claudin-5, occludin and ZO1) and aquaporin-4 (AQP4) water channel protein were assessed by immunohistochemistry staining from tissue taken on the day after MRI in the same infected rats (n = 9) and an additional cohort of non-infected F344 rats as control (n = 7). Immunohistochemistry images were taken in the frontal cortex, hippocampus and posterior cingulate and temporal cortices. The mean percentage area of the tight junction proteins was estimated and averaged across all brain regions. Vessel profiles were manually drawn on lectin images in ImageJ (Fiji), and the area-under curve (AUC) of AQP4 and lectin line profiles were taken and averaged across all brain regions. Results presented as mean ± s.e.m for all studies.

RESULTS

Infection led to a significant 78 ± 39 % increase in the BBB water exchange rate (baseline k_in = 2.57 ± 0.31 s^-1; infection k_in = 3.67 ± 0.42 s^-1; p = 0.02; Figure 1c-d). A significant positive correlation was observed between kin during infection and plasma VWF levels; p = 0.02; r = 0.77 (Figure 1b). No significant changes were measured in any of the tight junction protein markers when comparing non-infected with infected animals; claudin-3 (8.9 ± 1.8 % vs 11.0 ± 0.5 %), claudin-5 (47.7 ± 6.0 % vs 45.0 ± 5.1 %), occludin (13.2 ± 2.3 % vs 19.2 ± 0.5 %) and ZO-1 (21.0 ± 4.1 % vs 17.8 ± 7.4 %), respectively (Figure 2b). Finally, we observed a significant 38% higher AUC in AQP4 intensity profiles of infected animals (20.5 ± 1.0 arbitrary units) relative to non-infected animals (14.9 ± 2.1 arbitrary units); p = 0.02 (Figure 3c). AUC of lectin profiles remained consistent between the infected and non-infected groups (Figure 3d).

DISCUSSION

BBB-FEXI detected a significant increase in k_in in the rat brain in response to peripheral lung infection (Figure 1d), which demonstrates its sensitivity as a non-invasive tool to measure subtle BBB water permeability alterations such as those induced by systemic inflammation. We found the levels of plasma VWF had a strong positive correlation with kin during infection (Figure 1b), suggesting that peripheral infection could promote BBB water exchange via a non-disruptive vascular inflammatory response, since BBB tight junction protein markers remained unchanged (Figure 2b). We also found that the expression of AQP4 was 38% higher in infected animals when compared with non-infected controls (Figure 3c), which could plausibly drive the increases in water exchange observed during infection.

CONCLUSION

The BBB-FEXI technique is sensitive to BBB alterations caused by lung infection and is a promising tool for detecting and monitoring early BBB dysfunction and for better understanding the impact of peripheral infection in brain diseases.

Acknowledgements

This work is funded by the EPSRC: grant code EP/S031510/1. The purchase of the breeding pair and the establishment of the Fisher-344 rat strain was jointly supported by the European Union’s Seventh Framework Programme (FP7/2007-2013; grant agreement n° HEALTH-F2-2011-278850; INMiND) and Alzheimer Research UK network funds at the UoM. SL has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 804746). We would like to thank the staff at The Biological Service Facility University of Manchester for their help maintaining animal welfare and environmental enrichment during these studies.