Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Fluids and Barriers of the CNS https://doi.org/10.1186/s12987-022-00380-6 RESEARCH Open Access Accelerated amyloid angiopathy and related vascular alterations in a mixed murine model of Alzheimer´s disease and type two diabetes Maria Vargas‑Soria1,2†, Juan Jose Ramos‑Rodriguez1,3†, Angel del Marco1,2, Carmen Hierro‑Bujalance1,2, Maria Jose Carranza‑Naval1,2,4, Maria Calvo‑Rodriguez5, Susanne J. van Veluw5, Alan W. Stitt6, Rafael Simó7,8, Brian J. Bacskai5, Carmen Infante‑Garcia1,2* and Monica Garcia‑Alloza1,2* Abstract Background: While aging is the main risk factor for Alzheimer´s disease (AD), emerging evidence suggests that metabolic alterations such as type 2 diabetes (T2D) are also major contributors. Indeed, several studies have described a close relationship between AD and T2D with clinical evidence showing that both diseases coexist. A hallmark patho‑ logical event in AD is amyloid-β (Aβ) deposition in the brain as either amyloid plaques or around leptomeningeal and cortical arterioles, thus constituting cerebral amyloid angiopathy (CAA). CAA is observed in 85–95% of autopsy cases with AD and it contributes to AD pathology by limiting perivascular drainage of Aβ. Methods: To further explore these alterations when AD and T2D coexist, we have used in vivo multiphoton micros‑ copy to analyze over time the Aβ deposition in the form of plaques and CAA in a relevant model of AD (APPswe/ PS1dE9) combined with T2D (db/db). We have simultaneously assessed the effects of high-fat diet-induced prediabe‑ tes in AD mice. Since both plaques and CAA are implicated in oxidative-stress mediated vascular damage in the brain, oxidation, MMP activity by ­DQ™ Gelatin, and vascular functionality. as well as in the activation of matrix metalloproteinases (MMP), we have also analyzed oxidative stress by Amplex Red Results: We found that prediabetes accelerates amyloid plaque and CAA deposition, suggesting that initial meta‑ bolic alterations may directly affect AD pathology. T2D significantly affects vascular pathology and CAA deposition, which is increased in AD-T2D mice, suggesting that T2D favors vascular accumulation of Aβ. Moreover, T2D synergisti‑ cally contributes to increase CAA mediated oxidative stress and MMP activation, affecting red blood cell velocity. Conclusions: Our data support the cross-talk between metabolic disease and Aβ deposition that affects vascular integrity, ultimately contributing to AD pathology and related functional changes in the brain microvasculature. Keywords: Alzheimer’s disease, Type 2 diabetes, Prediabetes, Multiphoton microscopy, Amyloid, Oxidative stress, Matrix metalloproteinases Background Alzheimer´s disease (AD) is the most common cause † of dementia. Classical neuropathological features Maria Vargas-Soria and Juan Jose Ramos-Rodriguez have contributed equally to this work include aberrant amyloid and tau deposition in the brain that leads to a progressive neuronal and synap- *Correspondence: carmen.infante@gm.uca.es; monica.garcia@uca.es tic loss [1]. Amyloid pathology involves amyloid-β (Aβ) 1 Division of Physiology. School of Medicine, University of Cadiz, Cadiz, Spain peptide, derived from proteolytic processing of the Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 2 of 13 transmembrane amyloid precursor protein. As a patho- disorders, to our knowledge the impact of prediabetes genic hallmark of AD, Aβ can form amyloid plaques and/ and overt T2D on vascular function, oxidative stress or or deposit around leptomeningeal and cortical arterioles, MMP alterations associated with amyloid pathology in as cerebral amyloid angiopathy (CAA) [2]. These vascular AD remains largely unexplored. Therefore, we have ana- deposits are uncommon in other brain areas such as the lyzed amyloid deposition in vivo, using longitudinal mul- hippocampus, cerebellum, basal ganglia, thalamus, brain tiphoton microscopy, in murine models that harbor AD stem or white matter [3, 4]. Both types of Aβ deposit and prediabetes or T2D. can occur either relatively independently of each other or they can overlap [2]. Offering conclusive evidence, a Methods recent meta-analysis has reported that when based on Animals and treatments neuropathological examination, the prevalence of mod- Prediabetes was induced by administering high-fat diet erate to severe CAA in AD is 48%, and 23% in the general (HFD) (60% Kcal from fat, OpenSource, New Brunswick, population [5]. CAA worsens AD pathology by limiting NJ, USA) to wildtype and APP/PS1 immediately after perivascular drainage, a major pathway for Aβ clearance weaning (3 weeks of age) and up to the end of the experi- out of the brain [2]. ments by 32 weeks of age. These become overweight and Previous work has supported involvement of paren- present hyperinsulinemia, indicative of the prediabetic chymal Aβ and CAA [6] in oxidative stress-mediated estate [16]. Untreated animals received regular diet (RD) vascular damage in the brain [7]. This may ultimately (SAFE A04. Augy, France). Imaging acquisition started at contribute to CAA-related cerebrovascular dysfunction, 26 weeks of age and amyloid deposition was followed up cerebral hemorrhage, and further increase of CAA [6, to 32 weeks of age. MMP and oxidative stress were ana- 8]. Other underlying mechanisms include altered matrix lyzed also at 32 weeks of age. metalloproteinases (MMP) activity in association with AD-T2D mice (APP/PS1xdb/db) were produced by CAA. MMP2 and MMP9 are affected in AD mice that crossbreeding APPswe/PS1dE9 mice [17] with db/db reproduce amyloid pathology, and have been related to mice [18] as described [19]. Plaques can be detected tight junction disruption and increase in blood–brain by 6 months of age in APP/PS1 mice, when exponen- barrier (BBB) permeability [9]. Also, Aβ oligomers have tial amyloid deposition commences [20]. Therefore, been directly related to pericyte contraction which led to the direct effect of T2D on amyloid deposition in APP/ capillary stenosis, thus resulting in a reduction of the cer- PS1xdb/db mice was analyzed from a slightly earlier ebral blood flow [7]. Although age remains the main risk stage, at 22 weeks of age, and up to 32 weeks of age. Since factor for AD, advances in dementia research support different genetic backgrounds can significantly determine that metabolic disorders also play an important role in pathological outcomes, studies on prediabetes and T2D the development of AD. Insulin resistance, hyperinsuline- animals (on C57BL/6xC3H and C57BLKS/J backgrounds, mia and type 2 diabetes (T2D) are metabolic alterations respectively) were performed in parallel, unless initial strongly associated with AD [10, 11]. The prevalence of comparison between wildtype mice could be performed metabolic disorders has continued to increase over the (metabolic studies) and no differences were detected. last years. The latest edition of the Diabetes Atlas shows All studies were approved by the Animal Care and Use that 463 million adults are currently living with diabetes Committee of the University of Cadiz in accordance with worldwide [12], and T2D accounts for the vast majority the guidelines for care and use of experimental animals of the cases (90% approximately). Sedentary habits and (European Commission Directive 2010/63/UE and Span- progressive aging of the population contribute to the ish Royal Decree 53/2013). current diabetes pandemic that is usually preceded by a prediabetic state. Prediabetes includes impaired glu- Metabolic assessment cose tolerance and/or impaired fasting glucose that leads Body weight, glucose and insulin levels were determined to maintained hyperinsulinemia and translates to an in at 30–32 weeks of age, before sacrifice [14]. Blood glu- increased risk of the future development of T2D. cose levels were measured using the glucometer Optium Prediabetes and T2D lead to several macro and micro- Xceed (Abbott, UK) and plasma insulin samples were vascular complications resulting in cardiovascular and collected into tubes with potassium-EDTA (Sarstedt, cerebrovascular diseases. Specifically in the brain, these Nümbrecht, Germany). Blood samples were centrifuged conditions can cause neuronal pathology, alterations during 7 min, 6500 rpm at 4 °C, and plasma fraction was in synaptic communication as well as amyloid and tau stored at −80 °C until processed. Plasma insulin lev- pathologies [13, 14]. Moreover, diabetes and even predia- els were measured using ultrasensitive mouse enzyme- betes impair memory processing [15]. Whereas all these linked immunosorbent assay (ALPCO Diagnostics, data support common links between AD and metabolic Salem, NH, USA). Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 3 of 13 Oligomeric Aβ, Aβ40 and Aβ42 surgery and allowed to recover. They were also reimaged Immediately after the completion of the last imaging every 2 weeks up to 32 weeks of age, or until regrowth session, animals received an overdose of pentobarbital of the underlying tissue did not allow the imaging ses- (100 mg/kg). Brains were harvested, right hemispheres sion. All animals were imaged at least twice and isoflu- were fixed in PFA and left hemispheres were dissected. rane anesthesia was used in all imaging sessions. Aβ Cortex was fresh frozen at −80 ºC until used. Human deposition was visualized with methoxy-XO4 (Abcam, Amyloid β Oligomers (82E1-specific) Assay Kit (IBL Cambridge, UK) intraperitoneally injected (5 mg/Kg) the International Corp, Hamburg, Germany) was used for day before each imaging session. Angiograms were per- Aβ oligomers quantification in the cortex. Tissue was formed with 80 µl (12.5 mg/ml) i.v. retro-orbital injection 20 mM Tris–HCl, 140 mM NaCl, pH 7.5) with Halt™ homogenized (1/5 w/v) in Tris-buffered saline (TBS; of Texas Red dextran 70KD (Thermo Fischer Scientific, Spain) preceding the imaging session. In vivo imaging phosphatase and protease inhibitor cocktail (Thermo was conducted using a 20X (1.05 NA, Olympus) water Fisher Scientific Slu., Madrid, Spain). Homogenates were immersion objective in a Fluoview 1000 MPE micro- centrifuged (14,000 rpm) for 60 min at 4 °C. The super- scope (Olympus, Spain). Two-photon fluorescence was natant was diluted 1:2 with EIA buffer from the kit, fol- generated with 800 nm excitation, and collected emitted lowing manufacturer´s indications. Aβ40 and Aβ42 levels light in the range of 380–480, 500–540 and 560–650 nm. were also quantified in the cortex by colorimetric ELISA Along with the last imaging session, at 32 weeks of age, kits (Wako, Japan) as previously described [21]. Tis- red blood cell (RBC) velocity was assessed in 29–94 seg- with ­Halt™ phosphatase and protease inhibitor cocktail sue (5–10 mg) was homogenized in 50 μl of lysis buffer ments per group, using repeated line scans of the same area at 800 nm (5X zoom) through the horizontal part and centrifuged (14,000 rpm) for 12 min at 4 °C. Super- of the selected region of interest as previously described natants were used to measure soluble Aβ40 and Aβ42. [22] and measured [23] using ImageJ software. Ves- The pellet was extracted with 65 μl of 70% formic acid sel diameter changes [24] was also analyzed at 32 weeks and centrifuged (14,000 rpm) for 10 min. Samples were of age, with 3 s line scans, and data were expressed as then neutralized with 1 M Tris (pH 11) and diluted 1:40 % diameter change for each vessel. For Aβ deposition to measure insoluble Aβ levels. Human Aβ40 and Aβ42 615 × 615 μm fields, z/step 5 μm, depth ~ 200 μm were provided in the kit were used for standard curves. All imaged. Maximum intensity projections of z-series were absorbances were measured spectrophotometrically at obtained using ImageJ software. All stacks had 40 sec- 450 nm (MQX200R2, Biotek instruments, Burlington VT, tions and were used to measure plaque size and num- USA) and data were expressed as pmol/g tissue. ber after 2D projections. Plaque size was measured by thresholding, segmenting and measuring using the blue Lipid peroxidation and DNA oxidation fluorescence generated by methoxy-XO4. CAA burden Cortical lipid peroxidation was assessed by malondialde- was also analyzed in maximum intensity projections. hyde (MDA) content following manufacturer´s indica- Vessels were outlined and the CAA deposits were manu- tions (Lipid Peroxidation Assay Kit, Abcam, Cambridge, ally threshold using Image J free software, as previously UK). Absorbances were measured spectrophotometri- described [6, 25]. The number of pixels in white clusters cally at 532 nm. Data were expressed in nmol/g tissue. composed of at least 6 contiguous pixels was counted. DNA was purified from 5–10 mg cortical tissue by The CAA burden was calculated as the percentage of QIAamp DNA Mini Kit (Qiagen, MD, USA) according to the vessel area affected by CAA. New growth of CAA the manufacturer’s instructions. DNA oxidation was ana- deposits was measured every 2 weeks, using the same lyzed by 8-hydroxy-2’deoxyguanosine (8-OHdG) content approach. Texas-red signal was used to measure the ves- sel area from the initial imaging session as described [25, Kit, Molecular ­Signature®, BC, Canada) and measuring following manufacturer´s indications (8-OHdG ELISA 26]. Vessel diameter was measured at 3 separate points the absorbance at 450 nm. 500 ng DNA per sample were along the vessel segment, perpendicular to the axis. used and data were expressed in ng 8-OHdG/mg DNA. We cannot obviate the effect that metabolic disorders may have in the surgeries, nevertheless the percentage of animals that underwent the complete procedure for Cranial window implantation and multiphoton imaging the long-term in vivo imaging studies was similar in all Animals had permanent cranial windows placed for groups under study (APP/PS1-RD 75%, APP/PS1-HFD chronic in vivo imaging at 26 weeks of age in AD-HFD 75%, APP/PS1 71% and APP/PS1xdb/db 75%). and 22 weeks of age in case of AD-T2D studies, as pre- A second set of APP/PS1-HFD and APP/PS1xdb/db viously described [20]. To assess amyloid plaques and mice were imaged to analyze other functional alterations CAA deposition mice were imaged immediately after the Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 4 of 13 associated with Aβ deposition and vascular damage. weight (p = 0.231), glucose levels (p = 0.187) or insulin Oxidative stress and MMP activity were measured using levels (p = 0.746) were compared. Hence, these mice were Amplex Red (10-acetyl-3,7-dihydroxyphenoxazine) pooled and considered as the Control group of the study strate ­DQ™ Gelatin (Thermofisher, Spain) respectively, (Thermofisher, Spain), and the green fluorescent sub- in the following analyses. Body weight was significantly increased in APP/PS1 mice and their wildtype littermates while thioflavin S (Sigma, St. Louis, MO) was also used on HFD, db/db, and APP/PS1xdb/db mice compared for histochemical confirmation. The fluorophores were to Control mice [14, 19] (Table 1). Glucose levels were imaged in three immediately consecutive sessions so significantly higher in diabetic mice (db/db and APP/ the signal detected was exclusively derived from the PS1xdb/db) when compared with Control and APP/PS1 marker under study at each time, as previously described animals, while insulin levels were significantly increased [6, 27]. Briefly, 32 weeks old animals were anesthetized in diabetic (db/db and APP/PS1xdb/db mice) and predia- as described above, the dura was removed and 10 μl of betic (Control-HFD and APP/PS1-HFD mice (Table 1). Amplex Red in 85 μl of filtered PBS (1 mM) with 0.5 mg/ ml peroxidase was locally applied for 20 min. A cover- Aβ40, Aβ42 and oligomeric Aβ slip (8 mm in diameter) was placed, and animals were A slight increase in Aβ40 and Aβ42 levels was observed green fluorescent substrate ­DQ™ Gelatin (1 mg/ml) was imaged. The coverslip was removed and 50 μl of the in the cortex from APP/PS1 mice on HFD when com- pared with animals treated with RD, although differences applied for 20 min. After the second imaging session ani- were not statistically significant (Fig. 1). As previously mals were incubated with thioflavin S (0.1 mg/ml) for [14, 19], we observed a shift in Aβ levels; soluble Aβ was 10 min before the final imaging session. Amyloid plaques increased in APP/PS1xdb/db mice, reaching statistic and CAA imaging was performed as described above and significance in case of Aβ40 levels, while insoluble lev- outlined and Amplex Red, ­DQ™ Gelatin and thioflavin images were analyzed using Image J [6, 27]. Vessels were els were reduced, especially when insoluble Aβ40 was assessed. No differences were detected when oligomeric S signals (oxidative stress, MMP activity and Aβ depos- Aβ levels were measured (Fig. 1). Amplex Red, the green fluorescent substrate ­DQ™ Gela- its respectively) were quantified. In each vessel segment, CAA deposition is increased in larger caliber vessels tin and thioflavin S signal were threshold and signal-pos- from AD‑ T2D mice itive regions were quantified. The number of continuous Assessment of CAA deposition revealed that HFD- Amplex Red and ­DQ™ Gelatin signals were expressed pixels larger than 6 were counted. In each vessel segment, induced prediabetes as well as T2D significantly increased CAA deposition as disease progressed and we detected as a percentage of thioflavin S signal [6]. Image volumes a significant treatmentXweek effect [­ F(11,3393) = 8.79, focused on leptomeningeal vessels and not parenchymal **p < 0.001] (statistical power 1.0) (Fig. 2A and B). Indi- plaques, in order to obtain sufficient high quality images, vidual weekly assessment revealed increased CAA depo- however, a small set of amyloid plaques were analyzed sition in APP/PS1xdb/db mice by 24 weeks of age and and Amplex Red and MMP intensities were measured this increase was observed up to 32 weeks of age. Also, and normalized to thioflavin S intensity [27]. long-term HFD significantly increased CAA deposition Statistical analysis One-way ANOVA for independent samples followed by Tukey B test or Tamhane tests was used when four Table 1 Metabolic assessment groups were under study (metabolic assessment, RBC Body weight (g) Glucose (mg/dl) Insulin (ng/dl) velocity and vessel diameters), and Student t test for independent samples was used when only two groups Control 28.50 ± 1.84 122.71 ± 9.58 0.75 ± 0.29 were compared (Aβ levels, plaques, CAA, oxidative stress Control-HFD 44.50 ± 4.80†† 114.00 ± 7.15 5.52 ± 2.28†† and MMP activation). Two-way ANOVA (groupXweek) APP/PS1 28.61 ± 0.87 129.83 ± 4.20 0.69 ± 0.12 was used to analyze amyloid plaques and CAA progres- APP/PS1-HFD 42.95 ± 2.54†† 118.17 ± 9.94 7.13 ± 0.99†† sion rates. SPSS v.24 software was used for all statistical db/db 48.88 ± 2.83†† 484.50 ± 33.73‡‡ 6.90 ± 1.76†† analysis and results are expressed as mean ± SEM. APP/PS1xdb/db 42.87 ± 2.09†† 448.40 ± 68.20‡‡ 6.68 ± 2.08†† Body weight ­[F(5,35) = 13.04, ††p < 0.01 vs. Control and APP/PS1] and insulin levels ­[F(5,44) = 5.32, ††p < 0.01 vs. Control and APP/PS1] were significantly Results increased in animals on HFD, db/db and APP/PS1xdb/db mice, while glucose Metabolism levels were significantly higher in diabetic mice ­[F(5,39) = 52.19, ‡‡p < 0.01 vs. Control, Control-HFD, APP/PS1 and APP/PS1-HFD]. Data are representative of Initial comparison between wildtype mice from AD-T2D 5–12 mice (Control n = 7, Control-HFD n = 7, APP/PS1 n = 12, APP/PS1-HFD and AD-HFD studies revealed no differences when body n = 6, db/db n = 8, APP/PS1xdb/db n = 5) Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 5 of 13 Fig. 1 Aβ levels are altered in prediabetic and T2D mice. Soluble Aβ40 is increased in APP/PS1xdb/db mice when compared with APP/PS1 and prediabetic-APP/PS1 animals, ­[F(3,44) = 5.03, **p = 0.004 vs. rest of the groups] while differences did not reach statistical significance for soluble Aβ42 ­[F(3,36) = 2.20, p = 0.105]. On the other hand insoluble Aβ40 is reduced in APP/PS1xdb/db mice ­[F(3,40) = 5.064 **p = 0.003 vs. rest of the groups] and a similar trend is observed for insoluble Aβ42 although differences are not statistically significant [­ F(3,27) = 0.798, p = 0.506] (APP/PS1-RD n = 9–12, APP/PS1-HFD n = 7–12, APP/PS1 n = 8–12 and APP/PS1xdb/db n = 7–12) (mean ± SEM). While oligomeric Aβ is slightly higher in APP/PS1xdb/db mice, differences are not statistically significant ­[F(3,16) = 1.25, p = 0.323] (APP/PS1-RD n = 5, APP/PS1-HFD n = 5, APP/PS1 n = 4 and APP/PS1xdb/db n = 6) (mean ± SEM) Fig. 2 Prediabetes and T2D accelerate CAA progression rates in APP/PS1 mice in vivo and in real time. A Assessment of CAA deposition revealed that HFD-induced prediabetes as well as T2D significantly increased CAA deposition as disease progressed. When individual imaging sessions were compared no differences were detected between APP/PS1 and APP/PS1xdb/db mice at 22 weeks of age (p = 0.099), however significant differences were observed at 24 weeks of age when APP/PS1 and APP/PS1xdb/db mice were compared (**p < 0.01 vs. APP/PS1). By 26 weeks of age CAA burden was higher in APP/PS1-HFD animals and even higher burdens were observed in APP/PS1xdb/db mice [­ F(3,690) = 15.056, **p < 0.001 vs. rest of the groups; ††p < 0.01 vs. APP/PS-RD]. From 28 weeks of age on, both HFD-induced prediabetes and T2D significantly increased CAA deposition in APP/PS1 animals (week 28 ­[F(3,617) = 18.42, ‡‡p < 0.0.1 vs. APP/PS1-RD and APP/PS1]; week 30 ­[F(3,589) = 25.98, ‡‡p < 0.0.1 vs. APP/PS1-RD and APP/ PS1]; week 32 [­ F(3,503) = 26.16, ‡‡p < 0.0.1 vs. APP/PS1-RD and APP/PS1]) (APP/PS1-RD n = 4 and 42–81 vessel segments, APP/PS1-HFD n = 4 and 86–118 vessels segments, APP/PS1-RD n = 7 and 160–262 vessel segments, APP/PS1xdb/db and 233vessel segments) (mean ± SEM). B Illustrative example of individual vessels labeled with methoxy-XO4 to quantify CAA deposition along imaging sessions. CAA is isolated from the background and set white in binary images. Scale bar = 100 µm. Illustrative pseudocolor images of cortical regions (upper panel) were used to analyze CAA deposition along the biweekly imaging sessions in all groups under study (bottom panel). Blood vessels are labeled in red (Texas Red dextran 70KD) and amyloid in blue (methoxy-XO4). Scale bar = 100 µm. C) CAA deposition was increased in smaller vessels from APP/PS1-HFD mice whereas larger vessels from AD-T2D mice were more severely affected (diameter ≤ 25 µm: 26 weeks of age ­[F(3,256) = 2.49, p = 0.061], 28 weeks ­[F(3,224) = 3.89, ‡p = 0.010 vs. APP/PS1-RD and APP/PS1], 30 weeks ­[F(3,256) = 9.67, ‡‡p < 0.01 vs. APP/PS1-RD and APP/PS1], 32 weeks ­[F(3,250) = 10.71, **p < 0.01 vs. rest of the groups]; diameter > 25 ≤ 50 µm: 26 weeks of age ­[F(3,267) = 7.86, **p < 0.01 vs. rest of the groups], 28 weeks ­[F(3,236) = 8.13, ‡‡p < 0.01 APP/ PS1-RD and APP/PS1, ††p < 0.01 vs. APP/PS1-RD], 30 weeks ­[F(3,223) = 13.94, ‡‡p < 0.01 APP/PS1-RD and APP/PS1], 32 weeks ­[F(3,174) = 18.44, ‡‡p < 0.01 APP/PS1-RD and APP/PS1]; diameter > 50 µm: 26 weeks of age ­[F(3,159) = 14.39, **p < 0.01 vs. rest of the groups], 28 weeks ­[F(3,149) = 14.82, ‡‡p < 0.01 vs. APP/PS1-RD and APP/PS1], 30 weeks ­[F(3,102) = 8.78, **p < 0.01 vs. rest of the groups], 32 weeks ­[F(3,71) = 12.79, **p < 0.01 vs. rest of the groups]) (mean ± SEM) Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 6 of 13 as disease progressed (Fig. 2A and B). Interestingly, we disease progressed (Fig. 3A and B). We did not detect observed that while the prediabetic process seems to significant differences in amyloid plaque burden between affect smaller caliber vessels (diameter ≤ 25 µm) the dia- APP/PS1 and APP/PS1-RD animals when all four groups betic process significantly increased CAA deposition in under study were analyzed; nevertheless individual com- larger vessels. In vessels > 25 and ≤ 50 µm, the amount of parisons between these 2 groups revealed increased bur- CAA was increased in APP/PS1-HFD and more robustly dens in APP/PS1-RD mice at the beginning of the study in APP/PS1xdb/db mice and this effect was more severe (week 26 p = 0.021 and week 28 p = 0.016) when differ- in larger vessels from APP/PS1xdb/db mice (diame- ences in amyloid plaque size are larger, however differ- ter > 50 µm) (Fig. 2C). ences in amyloid plaque burden were not detected by 32 weeks of age, when amyloid plaque size is also similar In vivo amyloid plaque deposition in APP/PS1 and APP/PS1-RD. When we analyzed plaque burden (% of cortical volume) In vivo multiphoton microscopy studies also allowed we observed that progression rates were lower in AD- us to follow amyloid plaque size evolution. Plaque size T2D mice when compared with AD animals, in line with followed a similar trend in all groups under study and, previous postmortem observations in this mouse model since plaque size remained constant along the experi- [14, 19], whereas plaque deposition rates were higher in ments no significant groupXweek effect was observed prediabetes animals as previously observed in postmor- ­[F(11,9132) = 1.002, p = 0.441] (statistical power 0.573) in tem studies [16]. We detected a significant weekXgroup line with previous observations showing similar out- effect ­[F(11,1465) = 2.43, p = 0.005] (statistical power 0.962) comes in APP/PS1 animals [28, 29]. However, as previ- and individual biweekly assessment confirmed an accel- ously described amyloid plaques size was smaller in erated increase in the rate of deposition after long-term the mixed colony and individual biweekly assessment HFD and lower deposition in APP/PS1xdb/db mice as revealed that plaque size was even further reduced in Fig. 3 Long term HFD increases amyloid plaques deposition and T2D reduces amyloid plaques deposition in APP/PS1 mice. A Long-term HFD administration increased plaque burden (% of cortical volume) while it was significantly reduced in APP/PS1xdb/db mice. No differences were observed at 22 (p = 0.277) and 24 (p = 0.430) weeks of age. However, by 26 weeks of age amyloid plaque burden was significantly increased on animals on HFD and reduced burdens were observed in APP/PS1xdb/db animals ­[F(3,311) = 13.447, ₸₸p < 0.001 vs. APP/PS1-RD and APP/PS1-HFD, ##p < 0.01 vs. APP/PS1-HFD] and similar profiles were observed afterwards (week 28 [­ F(3,280) = 6.73, **p < 0.001 vs. rest of the groups], week 30 ­[F(3,289) = 13.25, **p < 0.001 vs. rest of the groups, ##p < 0.01 vs. APP/PS1-HFD], week 32 ­[F(3,299) = 17.99, **p < 0.001 vs. rest of the groups, ##p < 0.01 vs. APP/PS1-HFD]. B Illustrative images of cortical regions used to analyze plaque deposition in 32 week-old mice. Blood vessels are labeled in red (Texas Red dextran 70KD) and amyloid in blue (methoxy-XO4). Scale bar = 50 µm. C Amyloid plaques size was smaller in the mixed colony and individual biweekly assessment revealed that plaque size was even further reduced in APP/PS1xdb/db mice when compared with the rest of the groups (week 22 p = 0.561 vs. APP/PS1, week 24 ╫p = 0.021 vs. APP/PS1, week 26 ­[F(3,1638) = 16.079, **p < 0.001 vs. rest of the groups, ‡‡p < 0.001 vs. APP/PS1-RD and APP/PS1-HFD], week 28 ­[F(3,1638) = 15.371, **p < 0.001 vs. rest of the groups, ‡‡p < 0.001 vs. APP/PS1-RD and APP/PS1-HFD], week 30 ­[F(3,1995) = 11.003, **p < 0.001 vs. rest of the groups, ‡‡p < 0.001 vs. APP/PS1-RD and APP/PS1-HFD], week 32 ­[F(3,2980) = 5.076, ‡‡p < 0.001 vs. APP/PS1-RD and APP/PS1-HFD]). D Biweekly assessment revealed an increase in the number of new amyloid plaques in APP/PS1 animals on HFD when compared with the rest of the groups in the long term (week 22 p = 0.501, week 24 p = 0.976, week 26 [­ F(3,315) = 2.35, p = 0.079], week 28 ­[F(3,284) = 4.20, ##p = 0.006 vs. APP/PS1-HFD], week 30 ­[F(3,302) = 8.76, **p < 0.001 vs. rest of the groups], week 32 ­[F(3,316) = 27.70, **p < 0.001 vs. rest of the groups]) (APP/PS1-RD n = 4 and 47–65 fields, APP/PS1-HFD n = 4 and 49–79 fields, APP/PS1 n = 7 and 77–112 fields, APP/PS1xdb/db n = 4 and 47–91 fields) (mean ± SEM) Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 7 of 13 APP/PS1xdb/db mice when compared with APP/PS1 anesthesia. A general decrease in vessel diameter change mice (Fig. 3C). Whereas we cannot point towards a spe- was observed in vessels affected by CAA in prediabetic cific reason, it is feasible that using two mice colonies AD animals. We observed larger reductions in vessel might account for the differences observed in amyloid diameter change in APP/PS1xdb/db mice compared to plaque size and therefore control APP/PS1 mice are used AD mice (Fig. 4A). RBC velocity was reduced in small for both colonies. Moreover, differences are observed caliber vessels (25–50 µm) in APP/PS1xdb/db (i.e. AD- when APP/PS1xdb/db and APP/PS1-HFD mice are com- T2D mice) compared to APP/PS1 mice, but not in pre- pared with their respective controls. diabetic mice (Figs. 4B and C). The number of vessels In prediabetic mice, we observed that increased plaque available for analysis precluded in depth assessment burden was due to an increased deposition of new of RBC velocity in larger caliber vessels, although we plaques and we detected a significant groupXweek effect observed no significant differences when APP/PS1 and ­[F(11,1497) = 4.738, p < 0.001] (statistical power 1.000) when APP/PS1-RD mice were compared (p = 0.121). we analyzed amyloid plaque density. Biweekly assessment showed a maintained increase in the number of amyloid Oxidative stress and MMP activation associated plaques in APP/PS1 animals on HFD, while no differ- with amyloid deposition ences were observed when APP/PS1xdb/db mice were Oxidative stress analyzed by Amplex Red intensity was analyzed (Fig. 3D). These observations support that in measured in association with amyloid plaques and CAA APP/PS1xdb/db mice the reduction of plaque burden is in a separate cohort of mice at 32 weeks of age. Amplex due to the presence of smaller plaques, while the number Red intensity ratio (Amplex Red intensity/thioflavin of individual deposits is not affected, in line with previ- S intensity) associated with amyloid plaques was sig- ous studies showing lower plaque burden in this animal nificantly higher in APP/PS1xdb/db and APP/PS1-HFD DQ™ Gelatin signal, associated model [14]. (Fig. 5A and B). We also detected an increase in MMP activity, measured by ­ Vessel diameter and RBC velocity with plaques in HFD-treated mice (Fig. 5A and B). When Besides structural alterations to the vasculature, CAA we analyzed oxidative stress associated with CAA a sig- and diabetes also affect vessel function. To differenti- nificant increase was observed in APP/PS1xdb/db mice ate the role of diabetes and CAA on vessel function, we (Fig. 5C and D) and APP/PS1xdb/db mice also showed an therefore measured the % vessel diameter change (gener- increase in MMP intensity associated with CAA (Fig. 5C ated by the heartbeat) in vessels with CAA in mice with and D). and without prediabetes or diabetes under isoflurane Fig. 4 Vessel diameter change and RBC velocity are affected in diabetic APP/PS1 mice. A When we compared vessels with CAA, we observed a significant reduction in vessel diameter change in prediabetic APP/PS1 animals and larger reductions were observed in APP/PS1xdb/db mice ­[F(3,1172) = 29.87, **p < 0.01 vs. rest of the groups, ‡‡p < 0.0.1 vs. APP/PS1-RD and APP/PS1]. Data are representative of 4–6 mice (APP/PS1 n = 6, APP/ PS1xdb/db n = 4, APPP/PS1-RD n = 5, APP/PS1-HFD n = 5) (mean ± SEM). B A reduction of RBC velocity was observed in APP/PS1xdb/db mice that reached statistical significance when 25–50 µm in diameter vessels were analyzed. Whereas a similar profile was observed in prediabetic mice, differences did not reach statistical significance (< 25 µm ­[F(3,160) = 0.589, p = 0.623]; 25–50 µm ­[F(3,123) = 4.227, ‡‡p = 0.007 vs. APP/PS1-RD and APP/ PS1]). Individual RBC are represented from 5–15 line scans (APP/PS1 < 25 µm n = 15, 25–50 µm n = 11, APP/PS1xdb/db < 25 µm n = 15, 25–50 µm n = 8, APP/PS1-RD < 25 µm n = 9, 25–50 µm n = 6, APP/PS1-HFD < 25 µm n = 5, 25–50 µm n = 9). C) Representative examples of RBC line scans in CAA vessels from APP/PS1-RD, APP/PS1-HFD, APP/PS1 and APP/PS1xdb/db mice Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 8 of 13 Fig. 5 Oxidative stress and MMP activation are altered in prediabetic and diabetic APP/PS1 mice. A Oxidative stress associated with amyloid plaques (Amplex Red intensity/thioflavin intensity) was significantly increased in APP/PS1-HFD and APP/PS1xdb/db mice [­ F(3,117) = 10.009, ‡‡p < 0.001 vs. APP/PS1-RD and APP/PS1, ††p < 0.01 vs. APP/PS1]. MMP activity was also significantly increased in APP/PS1-HFD mice in association with amyloid plaques [­ F(3,126) = 7.71, **p < 0.001 vs. rest of the groups] (APP/PS1-RD n = 3 and 15–16 plaques, APP/PS1-HFD n = 4 and 20–22 plaques, APP/PS1 n = 7 and 50–56 plaques, APP/PS1xdb/db n = 6 and 40–41 plaques) (mean ± SEM). B Illustrative example of oxidative stress and MMP activation associated with amyloid plaques in all groups under study. Scale bar = 50 µm. C No differences were observed when oxidative stress or MMP activity were analyzed in association with CAA in prediabetic mice, although oxidative stress was significantly higher in APP/ PS1xdb/db mice when compared with APP/PS1 animals ­[F(3,387) = 3.32, †p = 0.019 vs. APP/PS1] and a significant increase in MMP activity was also observed in association with CAA in APP/PS1xdb/db mice ­[F(3,283) = 3.88, ‡p = 0.01 vs. APP/PS1 and APP/PS1-RD] (APP/PS1-RD n = 3 and 69–72 vessel segments, APP/PS1-HFD n = 4 and 56–62 vessel segments, APP/PS1 n = 7 and 71–73 vessel segments, APP/PS1xdb/db n = 6 and 88–93 vessel segments) (mean ± SEM). D Illustrative example of oxidative stress and MMP activation associated with CAA in all groups under study. Scale bar = 50 µm. E 8-OHdG was increased in AD colony used to study the effect of prediabetes [­ F(3,16) = 7.93, ‡‡p < 0.01 vs. APP/PS1-RD and APP/ PS1-HFD] F Lipid peroxidation, estimated by MDA content, was significantly increased in APP/PS1xdb/db mice animals [­ F(3,18) = 8.92, **p < 0.01 vs. rest of the groups] Lipid peroxidation, DNA and protein oxidation deposition in AD and animal models that harbor AD and Next, we analyzed postmortem lipid and DNA oxida- metabolic disorders, both prediabetes and overt T2D, tion by MDA and 8-OHdG content. 8-OHdG levels were have been developed [13, 14, 16, 34]. higher in the AD colony used to study the effect of pre- In line with previous studies in APP/PS1xdb/db mice diabetes. We cannot exclude that different colonies might [14, 19], we have observed a change in Aβ species, and account for detected differences, although we cannot soluble Aβ is increased while insoluble Aβ is reduced in point towards a specific cause for this dissimilarity. The AD-T2D mice. Other studies in AD models with meta- fact that no differences were observed when APP/PS1- bolic disorders have also shown changes in amyloid RD and APP/PS1-HFD mice were compared, or when deposition that do not affect total Aβ levels [13, 34] or APP/PS1xdb/db animals were compared with APP/PS1 amyloid burden. By contrast, studies with AD mice on animals, suggests a limited effect of metabolic alterations HFD have shown increased levels of Aβ [16, 35], sug- at this level (Fig. 5E). MDA content was significantly gesting that prediabetes and overt T2D might not nec- increased in the cortex from APP/PS1xdb/db mice when essarily induce the same phenotypic changes. Since the compared with APP/PS1 animals and APP/PS1-HFD ani- actual changes in the natural history of Aβ deposition mals (Fig. 5F). associated with metabolic alterations observed in APP/ PS1xdb/db mice remain elusive, multiphoton microscopy Discussion could permit us to monitoring in vivo the deposition of Previous studies have reported the close relationship amyloid plaques and CAA over time. Most AD patients between AD and T2D, showing that the prevalence of present some grade of CAA, and direct vascular dam- mild cognitive impairment is high in T2D patients [30] age associated with metabolic disorders may contribute and T2D might be a risk factor for AD. Alterations at the to alterations in Aβ deposition. Aβ40 is the main amyloid vascular level are a major link between AD and T2D [31] fragment present in vessel walls [36] and present under- and it has been shown that the increased risk of clinically standing of CAA, although still controversial, has pointed diagnosed AD might be mediated through cerebrovascu- out that Aβ from amyloid plaques in the brains of AD lar pathology [32]. Other potential links include insulin patients [37] may act as a "seed" and initiate the forma- resistance due to insulin signaling impairment [10, 11, tion of vascular deposits of Aβ. Aβ40 is commonly found 33], which is also observed in prediabetes, inflamma- in the walls of leptomeningeal arteries and penetrating tion or oxidative stress alterations [33]. These links may arterioles [2] and in the current study soluble Aβ40 levels ultimately affect neurofibrillary tangles and amyloid were significantly increased in the cortex from AD-T2D Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 9 of 13 mice in agreement with previous studies [34]. In vivo have lower levels of Aβ42 than APP/PS1 animals, may assessment of CAA deposition is significantly increased account for this difference. Nevertheless, the actual com- in APP/PS1 animals with T2D, when compared with position of the plaques might not be static and amyloid APP/PS1 mice. Interestingly, the prediabetic process plaque pathology in AD patients is generally heterogene- itself is enough to induce similar changes in CAA deposi- ous when the morphologic characteristics of Aβ depos- tion. Previous postmortem studies have shown that CAA its are analyzed [43]. Interestingly, previous studies have is increased in mixed models of AD and T2D [13] and shown that whereas T2D is a risk factor for AD and vas- vascular damage has also been observed in prediabetes cular dementia, the actual cerebral burden of the proto- mice [16]. While studies directly relating insulin or insu- typical AD pathologies is not [44]. In line with this, our lin resistance to CAA are scarce, our data suggest that previous results also show similar outcomes when post- altered insulin homeostasis could be a mediator of the mortem plaque burden is assessed [14, 19]. observed vascular alterations. In this regard, memantine Vascular functionality was also examined in our ani- reduces CAA by enhancing insulin-degrading enzyme mal models given the relevance of vascular risk factors in expression [38], supporting the deleterious cross-talk AD and their role in prediabetes and T2D. In this sense, between insulin homeostasis and CAA. The fact that vascular alterations have been proposed as feasible links both prediabetic and AD-T2D mice show an increase in contributing to an AD phenotype in diabetes, including spontaneous cortical hemorrhages might also underlie vascular endothelial dysfunction, oxidative stress, inflam- both BBB dysfunction and impairment in Aβ drainage, as mation, BBB dysfunction, neurovascular damage and major mechanisms accounting for CAA [39]. MMP alterations [31, 44, 45]. In postmortem AD brains, When CAA-affected vessels were plotted by diam- tight junction proteins, such as occludin, claudin-5, and eter size, in prediabetic mice we observed a preferential zonula occludens-1 are reduced in Aβ-loaded capillaries, Aβ deposition in smaller vessels (< 25 µm) whereas Aβ and affected capillaries are associated with NOX-2-acti- deposition was significantly increased in larger vessels vated microglia and NADPH oxidase-2 [46]. Thus, neu- from AD-T2D mice. In line with these observations, Aβ roinflammation and reactive oxygen species contribute is deposited in the walls of arteries and arterioles, and to Aβ endothelial toxicity and BBB disruption [46]. In less frequently in capillaries and veins [40]. Also, stud- capillaries isolated from AD mouse brains, degradation ies focusing on other brain regions have reported that of P-glycoprotein and low-density lipoprotein receptor- CAA preferentially affects larger vessels in the fore- related protein 1 is facilitated by Aβ40 exposure, which brain from transgenic mice, whereas smaller vessels are reduces the transport activity of P-glycoprotein and in affected much later [41]. Differences in Aβ deposition at turn the elimination of Aβ [47]. Also, in diabetic condi- the vascular level have been previously described, and tions, there is decreased expression of zonula occludens-1 while Aβ40 tends to accumulate in leptomeningeal arter- in brain capillaries as well as increased perivascular ies and penetrating arterioles, capillaries tend to accu- immunoglobulin G leakage, indicating increased BBB mulate Aβ42, resembling the composition in plaques [2, permeability in diabetes [48]. 39]. The fact that Aβ42 is increased in our AD-HFD mice Vessel diameter changes were reduced in CAA-affected might support the detected differences with AD-T2D vessels from AD-HFD mice. This effect was more robust animals. Likewise, other studies have reported differen- in AD-T2D animals, supporting a more severe crosstalk tial effects in vessels depending on APOE alleles. APOE between AD and overt diabetes and a major impact of ε4 inhibits Aβ transport across the BBB, it is associated the diabetic milieu on vascular impairment. While pre- with lower antioxidant activity and mediates BBB deg- vious studies have supported anatomical alterations and radation through a proinflammatory pathway involving changes in the vasculature of diabetic mice [31], to our cyclophilin A in pericytes, impairing Aβ clearance at dif- knowledge, this is the first in vivo assessment of the actual ferent levels [42]. Also, APOE ε4 has been related to CAA functional changes that take place in the brain from pre- in capillaries and larger vessels, while APOE ε2 seems to diabetic, T2D and AD models in real time. Vascular pul- affect larger arteriolar vessels whereas cortical capillaries sation has been described as a mechanism of perivascular are spared [39]. Nevertheless, further studies are needed clearance of Aβ [49, 50] and as disease progresses; arter- to understand the link between APOE and T2D. ies show reduced vascular reactivity affecting perivascu- We also observed that while amyloid plaque numbers lar drainage of fluids and solutes. CAA is associated with were increased in prediabetes animals, plaque size was an increased relative pulsatility index in AD mice and not significantly affected, as previously observed in APP/ patients [49]. Likewise, studies in old mice have shown PS1 animals [28] and this prediabetes model [16]. Aβ42 is reduced vessel wall pulsatility and capacity to clear Aβ more commonly observed in the parenchymal plaques of [50] and similar observations have been reported after AD patients [39], and the fact that APP/PS1xdb/db mice other insults [24, 51]. Vessel pulsation is determined by Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 10 of 13 heartbeat and the role of the different components of a disbalance in MMP expression seems to be associated the neurovascular unit, distinctively affecting arteries, with CAA-related hemorrhages [45] and similar out- arterioles and downstream microvessels. This activity is comes are observed in diabetic models [63]. hampered by CAA, causing vascular tone deregulation, We observed an overall increase in oxidative stress affecting vascular cells or impairing neurovascular cou- associated with amyloid plaques in AD-HFD and AD- pling, ultimately interfering with perivascular clearance T2D mice. Nevertheless, when CAA was analyzed, [52]. As CAA progresses, the neurovascular unit degen- increased oxidative stress only reached statistical sig- erates in amyloid-laden vessels [40, 53], affecting the nificance if overt diabetes was established. Whereas the integrity of the vessel wall and resulting in BBB leakage, amount of CAA is still limited in APP/PS1 by 7 months vessel occlusion or rupture, leading to hemorrhage and of age, the fragility of APP/PS1xdb/db mice precludes decreased cerebral blood flow [2]. Also, under diabetic chronic in vivo studies at later stages. The fact that the conditions, both the capillaries and larger vessels of the presence of CAA is still scarce, limits the capacity of our brain show a thickened basal lamina from collagen depo- approach to robustly discern between different type of sition, accumulated byproducts of lipid peroxidation, and vessels, althoughs the function and compromise of the deterioration of endothelial cells and pericytes [54]. vasculature and amyloid deposition is largely depending Other studies have reported reduced evoked vascular on the type of vessel under study; capillaries, penetrating reactivity in AD patients and transgenic mouse models of arteries or leptomeningeal vessels. AD [55] that might also contribute to impaired perivascu- On the other hand, we have previously observed the lar clearance and subsequent Aβ deposition. In addition, presence of microhemorrhages in the postmortem analy- lower blood flow velocity has been associated with demen- sis of APP/PS1xdb/db mice [14, 19] and, indeed, micro- tia and cognitive decline in the Rotterdam study [56], and hemorrhages have been reported in close association Aβ deposition is associated with impaired vascular reac- with CAA [64]. Previous studies have shown an increased tivity responses in AD mice [57]. Likewise, our data are CAA severity in blood vessels involved in microinfarcts, also in accordance with previous studies reporting that whereas those involved in microhemorrhages had reduced vasodilatation is compromised in aged AD mice [8]. Vas- CAA at the site of the bleeding [65]. It is then feasible that cular functional alterations may result in CAA deposition, the combination of CAA with T2D, that affects vessel contributing to the loss of brain homeostasis and neuronal wall integrity and blood flow concomitantly, contributes damage, but vascular damage may also contribute to CAA to increased hemorrhage burden in this model [19]. In deposition by limiting Aβ clearance in a vicious cycle [58]. line with this idea, previous studies have shown that reac- We further analyzed oxidative stress and MMP activa- tive oxygen species might be implicated in the formation tion in vivo and in real time and we observed an over- of CAA and underlie CAA-related microhemorrhages in all increase of oxidative stress associated with amyloid aged AD mice [8]. Thus, increased levels of reactive oxy- deposition, along with metabolic alterations. Previous gen species lead to alterations in blood flow resistance and studies have largely documented increased oxidative endothelial regulatory response is impaired by the disrup- stress in the brain and peripheral tissues in prediabetes tion of endothelium-dependent nitric oxide signaling [66– and T2D. Chronic hyperglycemia and hyperinsulinemia, 68], resulting in increased microvessel permeability and inherent to prediabetes and T2D, favor the formation of impaired microvascular endothelial function [67, 69–71]. advanced glycation end products and the overproduc- Hydrogen peroxide, superoxide derivates, hydroxyl radical tion of reactive oxygen species, leading to extracellular or peroxynitrites increase endothelial calcium concentra- matrix damage [59]. MMPs are required for the integrity tion in time- and dose-dependent manner causing dys- of the extracellular matrix and have been associated with regulation in calcium homeostasis, acting as proapoptotic alterations in BBB permeability and neuropathological factor in cerebral vascular cells [66, 70, 72]. In addition, disorders [60]. We observed that MMP activation is sig- oxidative stress has a key role in insulin resistance and pro- with metabolic disorders. Since ­DQ™ Gelatin, used as a nificantly increased in association with CAA in AD mice gression of diabetes [73, 74] and diabetic mice show vas- cular alterations associated with increased oxidative stress, probe in our study, is preferentially degraded by MMP9, pericyte dysfunction and cerebrovascular integrity com- but it can also be degraded by MMP2 [61], we cannot promise [75]. Likewise, advanced glycation end products, unequivocally point towards one of them. Nevertheless, resulting from hyperglycemia in T2D, alter the structure previous studies have shown that the increase of MMP2 and function of BBB proteins [62] and diabetes-induced and MMP9 activities is associated with the inflamma- inflammation leads to astrocytic feet swelling in small ves- tory process and oxidative stress in APP/PS1 mice [6, 9] sels, separation of the plasma membrane and accumula- and hyperglycemia alters the PKC-β pathway, causing an tion of hypertrophic microglia [54]. increase in MMP2 [62]. In addition, in patients with CAA Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 11 of 13 Our in vivo observations were also confirmed by post- Ministerio de Ciencia e Innovacion. Programa Estatal de Generacion de Conocimiento y Fortalecimiento Cientifico y Tecnologico del Sistema de mortem assessment of oxidative stress and an overall I + D + i y del Programa Estatal de I + D + i Orientada a los Retos de la Socie‑ increase of lipid peroxidation was observed in AD-T2D dad, del Plan Estatal de Investigacion Cientifica y Tecnica y de Innovacion animals, when compared with AD alone. It should be (PID2020-115499RB-I00/AEI/10.130 39/501100011033). Programa Estatal de I + D + I orientada a los Retos de la Sociedad (BFU 2016-75038-R), financed noted that oxidative stress alters lipid-lipid interaction by the Agencia Estatal de Investigacion (AEI) and the Fondo Europeo de leading to impaired membrane permeability [76] and Desarrollo Regional (FEDER), Ministerio de Economia y Competitividad. high levels of markers of oxidative stress-induced lipid Proyectos de I + D + i, en regimen de concurrencia competitiva, destinadas a las universidades y entidades publicas de investigacion calificadas como peroxidation and cytotoxicity such as 4-hydroxynon- agentes del Sistema Andaluz del Conocimiento, en el ambito del Plan Andaluz enal, MDA or thiobarbituric acid reactive substances are de Investigacion, Desarrollo e Innovación (PAIDI 2020). Andalucia se mueve increased in AD [77, 78] and diabetes [79, 80] supporting con Europa (P20-00928). a synergistic effect when both diseases coexist. In addi- Availability of data and materials tion, lipid peroxidation is accompanied by structural and Data are available on reasonable request. functional alterations of the microvasculature [69], that may ultimately contribute to the observed pathology Declarations when AD and T2D coexist. Ethics approval and consent to participate All studies were approved by the Animal Care and Use Committee of the Conclusions University of Cadiz in accordance with the guidelines for care and use of experimental animals (European Commission Directive 2010/63/UE and Span‑ Limitations of this study include the use of isoflurane ish Royal Decree 53/2013). anesthesia, which affects vessel function. Neverthe- less, our data show that prediabetes accelerates amyloid Consent for publication Not applies. pathology, suggesting that reversing the early stages of metabolic disease could slow down brain pathology. Competing interests Functional alterations in the brain vasculature are sig- Authors declare that they have no competing interests. nificantly worsened in AD mice with T2D. Moreover, the Author details presence of T2D has also a relevant impact on oxidative 1 Division of Physiology. School of Medicine, University of Cadiz, Cadiz, Spain. 2 stress associated with amyloid deposition, both as amy- Instituto de Investigacion e Innovacion en Ciencias Biomedicas de la Provin‑ cia de Cadiz (INIBICA), Cadiz, Spain. 3 Currently at Department of Physiology, loid plaques and CAA, providing further support to vas- School of Health Sciences, University of Granada, Granada, Spain. 4 Salus‑Infir‑ cular pathology as underlying mechanisms linking T2D morum, University of Cadiz, Cadiz, Spain. 5 Alzheimer Research Unit, Depart‑ and AD. ment of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston, USA. 6 Wellcome-Wolfson Institute for Experimental Medicine, Queen’s University Belfast, Belfast, Northern Ireland, UK. 7 Diabetes and Metab‑ olism Research Unit, Vall d’Hebron Research Institute, Universitat Autonoma Abbreviations de Barcelona, Barcelona, Spain. 8 Centro de Investigacion Biomedica en Red 8-OHdG: 8-Hydroxy-2’deoxyguanosine; Aβ: Amyloid-β; AD: Alzheimer´s de Diabetes y Enfermedades Metabolicas Asociadas (CIBERDEM), Instituto de disease; BBB: Blood–brain barrier; HFD: High-fat diet; MDA: Malondialdehyde; Salud Carlos III (ISCIII), Madrid, Spain. MMP: Matrix metalloproteinases; RBC: Red blood cell; RD: Regular diet; T2D: Type 2 diabetes. Received: 1 June 2022 Accepted: 26 September 2022 Acknowledgements We thank the Animal Facility of the University of Cadiz and its personnel for their help and support. We thank Servicios Centrales de Investigacion en Biomedicina (SC-IBM) from Universidad de Cadiz for the resources and techni‑ References cal support. 1. Spires-Jones TL, Hyman BT. The intersection of amyloid beta and tau at synapses in Alzheimer’s disease. Neuron. 2014;82(4):756–71. Author contributions 2. Greenberg SM, Bacskai BJ, Hernandez-Guillamon M, Pruzin J, Sperling R, MVS and JJRR experiments design, data acquisition and analysis. ADM van Veluw SJ. Cerebral amyloid angiopathy and Alzheimer disease—one experimental design, data acquisition, analysis and interpretation. CHB, MJCN peptide, two pathways. Nat Rev Neurol. 2020;16(1):30–42. data acquisition and analysis. MCR, SVV, AS, RS and BJB design, analysis sup‑ 3. Rensink AA, de Waal RM, Kremer B, Verbeek MM. Pathogenesis of cerebral port, drafting and critical revision of manuscript for intellectual content. CIG amyloid angiopathy. Brain Res Brain Res Rev. 2003;43(2):207–23. and MGA study concept and design, analysis, drafting and critical revision 4. Yamada M. Cerebral amyloid angiopathy: emerging concepts. J Stroke. of manuscript for intellectual content. All authors provided critical feedback 2015;17(1):17–30. and helped shape the research, analysis and manuscript. All authors read and 5. Jakel L, De Kort AM, Klijn CJM, Schreuder F, Verbeek MM. Prevalence of approved the final manuscript. cerebral amyloid angiopathy: a systematic review and meta-analysis. Alzheimers Dement. 2022;18(1):10–28. https://​doi.​org/​10.​1002/​alz.​12366 Funding 6. Garcia-Alloza M, Prada C, Lattarulo C, Fine S, Borrelli LA, Betensky R, et al. University of Cadiz Predoctoral Fellowship (CHB). This study is part of the cur‑ Matrix metalloproteinase inhibition reduces oxidative stress associated rent project (RECOGNISED; Clinical Trials gov registration no. NCT04281186) with cerebral amyloid angiopathy in vivo in transgenic mice. J Neuro‑ funded by the European Commission (H2020 programme-GA 847749) focus‑ chem. 2009;109(6):1636–47. ing on common mechanisms in the pathogenesis of diabetic retinopathy, 7. Nortley R, Korte N, Izquierdo P, Hirunpattarasilp C, Mishra A, Jaunmuktane brain pathology and cognitive impairment, with special interest in the Z, et al. Amyloid beta oligomers constrict human capillaries in Alzheimer’s neurovascular unit, in the T2D population. Agencia Estatal de Investigacion. Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 12 of 13 disease via signaling to pericytes. Science. 2019. https://​doi.​org/​10.​1126/​ 28. Meyer-Luehmann M, Spires-Jones TL, Prada C, Garcia-Alloza M, de scien​ce.​aav95​18. Calignon A, Rozkalne A, et al. Rapid appearance and local toxicity of 8. Han BH, Zhou ML, Johnson AW, Singh I, Liao F, Vellimana AK, et al. amyloid-beta plaques in a mouse model of Alzheimer’s disease. Nature. Contribution of reactive oxygen species to cerebral amyloid angiopathy, 2008;451(7179):720–4. vasomotor dysfunction, and microhemorrhage in aged Tg2576 mice. 29. Garcia-Alloza M, Dodwell SA, Meyer-Luehmann M, Hyman BT, Bacskai Proc Natl Acad Sci U S A. 2015;112(8):E881–90. BJ. Plaque-derived oxidative stress mediates distorted neurite tra‑ 9. Liu Y, Chen Z, Li B, Yao H, Zarka M, Welch J, et al. Supplementation with jectories in the Alzheimer mouse model. J Neuropathol Exp Neurol. gamma-glutamylcysteine (gamma-GC) lessens oxidative stress, brain 2006;65(11):1082–9. inflammation and amyloid pathology and improves spatial memory in 30. You Y, Liu Z, Chen Y, Xu Y, Qin J, Guo S, et al. The prevalence of mild cogni‑ a murine model of AD. Neurochem Int. 2020. https://​doi.​org/​10.​1016/j.​ tive impairment in type 2 diabetes mellitus patients: a systematic review neuint.​2020.​104931. and meta-analysis. Acta Diabetol. 2021;58:671–85. https://​doi.​org/​10.​ 10. Baglietto-Vargas D, Shi J, Yaeger DM, Ager R, LaFerla FM. Diabetes and 1007/​s00592-​020-​01648-9. Alzheimer’s disease crosstalk. Neurosci Biobehav Rev. 2016;64:272–87. 31. Hayden MR. Type 2 diabetes mellitus increases the risk of late-onset 11. De Felice FG, Ferreira ST. Inflammation, defective insulin signaling, Alzheimer’s disease: ultrastructural remodeling of the neurovascular unit and mitochondrial dysfunction as common molecular denomina‑ and diabetic gliopathy. Brain Sci. 2019;9(10):262. tors connecting type 2 diabetes to Alzheimer disease. Diabetes. 32. Vagelatos NT, Eslick GD. Type 2 diabetes as a risk factor for Alzheimer’s 2014;63(7):2262–72. disease: the confounders, interactions, and neuropathology associated 12. International Diabetes Federation. IDF Diabetes Atlas. 9th ed. Brussels; with this relationship. Epidemiol Rev. 2013;35:152–60. 2019. https://​www.​diabe​tesat​las.​org. 33. Holscher C. Insulin signaling impairment in the brain as a risk factor in 13. Takeda S, Sato N, Uchio-Yamada K, Sawada K, Kunieda T, Takeuchi D, et al. Alzheimer’s Disease. Front Aging Neurosci. 2019;11:88. Diabetes-accelerated memory dysfunction via cerebrovascular inflamma‑ 34. Niedowicz DM, Reeves VL, Platt TL, Kohler K, Beckett TL, Powell DK, et al. tion and Abeta deposition in an Alzheimer mouse model with diabetes. Obesity and diabetes cause cognitive dysfunction in the absence of Proc Natl Acad Sci U S A. 2010;107(15):7036–41. accelerated beta-amyloid deposition in a novel murine model of mixed 14. Infante-Garcia C, Ramos-Rodriguez JJ, Galindo-Gonzalez L, Garcia-Alloza or vascular dementia. Acta Neuropathol Commun. 2014;2:64. M. Long-term central pathology and cognitive impairment are exac‑ 35. Guo Y, Ma X, Li P, Dong S, Huang X, Ren X, et al. High-fat diet induced erbated in a mixed model of Alzheimer’s disease and type 2 diabetes. discrepant peripheral and central nervous systems insulin resistance Psychoneuroendocrinology. 2016;65:15–25. in APPswe/PS1dE9 and wild-type C57BL/6J mice. Aging (Albany NY). 15. Ennis GE, Saelzler U, Umpierrez GE, Moffat SD. Prediabetes and working 2020;13(1):1236–50. memory in older adults. Brain Neurosci Adv. 2020;4:2398212820961725. 36. McGowan E, Pickford F, Kim J, Onstead L, Eriksen J, Yu C, et al. Abeta42 16. Ramos-Rodriguez JJ, Ortiz-Barajas O, Gamero-Carrasco C, de la Rosa PR, is essential for parenchymal and vascular amyloid deposition in mice. Infante-Garcia C, Zopeque-Garcia N, et al. Prediabetes-induced vascular Neuron. 2005;47(2):191–9. alterations exacerbate central pathology in APPswe/PS1dE9 mice. Psy‑ 37. Yamada M, Ono K, Hamaguchi T, Noguchi-Shinohara M. Natural phenolic choneuroendocrinology. 2014;48:123–35. compounds as therapeutic and preventive agents for cerebral amyloido‑ 17. Jankowsky JL, Slunt HH, Gonzales V, Jenkins NA, Copeland NG, Borchelt sis. Adv Exp Med Biol. 2015;863:79–94. DR. APP processing and amyloid deposition in mice haplo-insufficient for 38. Inoue Y, Ueda M, Masuda T, Misumi Y, Yamashita T, Ando Y. Memantine, a presenilin 1. Neurobiol Aging. 2004;25(7):885–92. noncompetitive N-Methyl-D-Aspartate receptor antagonist, attenuates 18. Hummel KP, Dickie MM, Coleman DL. Diabetes, a new mutation in the cerebral amyloid angiopathy by increasing insulin-degrading enzyme mouse. Science. 1966;153(3740):1127–8. expression. Mol Neurobiol. 2019;56(12):8573–88. 19. Ramos-Rodriguez JJ, Jimenez-Palomares M, Murillo-Carretero MI, Infante- 39. Gireud-Goss M, Mack AF, McCullough LD, Urayama A. Cerebral amyloid Garcia C, Berrocoso E, Hernandez-Pacho F, et al. Central vascular disease angiopathy and blood-brain barrier dysfunction. Neuroscientist. and exacerbated pathology in a mixed model of type 2 diabetes and 2020;27:1073858420954811. Alzheimer’s disease. Psychoneuroendocrinology. 2015;62:69–79. 40. Magaki S, Tang Z, Tung S, Williams CK, Lo D, Yong WH, et al. The effects 20. Garcia-Alloza M, Robbins EM, Zhang-Nunes SX, Purcell SM, Beten‑ of cerebral amyloid angiopathy on integrity of the blood-brain barrier. sky RA, Raju S, et al. Characterization of amyloid deposition in the Neurobiol Aging. 2018;70:70–7. APPswe/PS1dE9 mouse model of Alzheimer disease. Neurobiol Dis. 41. Domnitz SB, Robbins EM, Hoang AW, Garcia-Alloza M, Hyman BT, Rebeck 2006;24(3):516–24. GW, et al. Progression of cerebral amyloid angiopathy in transgenic 21. Infante-Garcia C, Jose Ramos-Rodriguez J, Delgado-Olmos I, Gamero-Car‑ mouse models of Alzheimer disease. J Neuropathol Exp Neurol. rasco C, Teresa Fernandez-Ponce M, Casas L, et al. Long-term mangiferin 2005;64(7):588–94. extract treatment improves central pathology and cognitive deficits in 42. Tarasoff-Conway JM, Carare RO, Osorio RS, Glodzik L, Butler T, Fieremans APP/PS1 mice. Mol Neurobiol. 2017;54(6):4696–704. E, et al. Clearance systems in the brain-implications for Alzheimer disease. 22. Schiessl IM, Castrop H. Deep insights: intravital imaging with two-photon Nat Rev Neurol. 2015;11(8):457–70. microscopy. Pflugers Arch. 2016;468(9):1505–16. 43. Di Fede G, Catania M, Maderna E, Ghidoni R, Benussi L, Tonoli E, et al. 23. Shih AY, Driscoll JD, Drew PJ, Nishimura N, Schaffer CB, Kleinfeld D. Molecular subtypes of Alzheimer’s disease. Sci Rep. 2018;8(1):3269. Two-photon microscopy as a tool to study blood flow and neuro‑ 44. Biessels GJ, Despa F. Cognitive decline and dementia in diabetes vascular coupling in the rodent brain. J Cereb Blood Flow Metab. mellitus: mechanisms and clinical implications. Nat Rev Endocrinol. 2012;32(7):1277–309. 2018;14(10):591–604. 24. Iliff JJ, Wang M, Zeppenfeld DM, Venkataraman A, Plog BA, Liao Y, et al. 45. Jakel L, Kuiperij HB, Gerding LP, Custers EEM, van den Berg E, Jolink WMT, Cerebral arterial pulsation drives paravascular CSF-interstitial fluid et al. Disturbed balance in the expression of MMP9 and TIMP3 in cerebral exchange in the murine brain. J Neurosci. 2013;33(46):18190–9. amyloid angiopathy-related intracerebral haemorrhage. Acta Neuro‑ 25. Robbins EM, Betensky RA, Domnitz SB, Purcell SM, Garcia-Alloza M, pathol Commun. 2020;8(1):99. Greenberg C, et al. Kinetics of cerebral amyloid angiopathy progres‑ 46. Carrano A, Hoozemans JJ, van der Vies SM, van Horssen J, de Vries HE, sion in a transgenic mouse model of Alzheimer disease. J Neurosci. Rozemuller AJ. Neuroinflammation and blood-brain barrier changes in 2006;26(2):365–71. capillary amyloid angiopathy. Neurodegener Dis. 2012;10(1–4):329–31. 26. Garcia-Alloza M, Robbins EM, Zhang-Nunes SX, Purcell SM, Beten‑ 47. Yamazaki Y, Kanekiyo T. Blood-brain barrier dysfunction and the patho‑ sky RA, Raju S, et al. Characterization of amyloid deposition in the genesis of Alzheimer’s disease. Int J Mol Sci. 2017;18(9):1965. APPswe/PS1dE9 mouse model of Alzheimer disease. Neurobiol Dis. 48. Acharya NK, Levin EC, Clifford PM, Han M, Tourtellotte R, Chamberlain D, 2006;24(3):516–24. et al. Diabetes and hypercholesterolemia increase blood-brain barrier 27. McLellan ME, Kajdasz ST, Hyman BT, Bacskai BJ. In vivo imaging of reactive permeability and brain amyloid deposition: beneficial effects of the oxygen species specifically associated with thioflavine S-positive amyloid LpPLA2 inhibitor darapladib. J Alzheimers Dis. 2013;35(1):179–98. plaques by multiphoton microscopy. J Neurosci. 2003;23(6):2212–7. Vargas‑Soria et al. Fluids and Barriers of the CNS (2022) 19:88 Page 13 of 13 49. Kim SH, Ahn JH, Yang H, Lee P, Koh GY, Jeong Y. Cerebral amyloid angi‑ 72. Lloret A, Esteve D, Lloret MA, Monllor P, Lopez B, Leon JL, et al. Is oxidative opathy aggravates perivascular clearance impairment in an Alzheimer’s stress the link between cerebral small vessel disease, sleep disruption, disease mouse model. Acta Neuropathol Commun. 2020;8(1):181. and oligodendrocyte dysfunction in the onset of alzheimer’s disease? 50. Kress BT, Iliff JJ, Xia M, Wang M, Wei HS, Zeppenfeld D, et al. Impairment Front Physiol. 2021;12:708061. of paravascular clearance pathways in the aging brain. Ann Neurol. 73. Jha SK, Jha NK, Kumar D, Ambasta RK, Kumar P. Linking mitochondrial 2014;76(6):845–61. dysfunction, metabolic syndrome and stress signaling in Neurodegenera‑ 51. Mestre H, Tithof J, Du T, Song W, Peng W, Sweeney AM, et al. Flow of tion. Biochim Biophys Acta Mol Basis Dis. 2017;1863(5):1132–46. cerebrospinal fluid is driven by arterial pulsations and is reduced in 74. Etchegoyen M, Nobile MH, Baez F, Posesorski B, Gonzalez J, Lago N, et al. hypertension. Nat Commun. 2018;9(1):4878. Metabolic syndrome and Neuroprotection. Front Neurosci. 2018;12:196. 52. Di Marco LY, Farkas E, Martin C, Venneri A, Frangi AF. Is Vasomotion in 75. Bailey J, Coucha M, Bolduc DR, Burnett FN, Barrett AC, Ghaly M, et al. cerebral arteries impaired in Alzheimer’s disease? J Alzheimers Dis. GLP-1 receptor nitration contributes to loss of brain pericyte function in 2015;46(1):35–53. a mouse model of diabetes. Diabetologia. 2022;65:1541–54. https://​doi.​ 53. Morris AW, Carare RO, Schreiber S, Hawkes CA. The cerebrovascular org/​10.​1007/​s00125-​022-​05730-5. basement membrane: role in the clearance of beta-amyloid and cerebral 76. Gaschler MM, Stockwell BR. Lipid peroxidation in cell death. Biochem amyloid angiopathy. Front Aging Neurosci. 2014;6:251. Biophys Res Commun. 2017;482(3):419–25. 54. Li X, Cai Y, Zhang Z, Zhou J. Glial and vascular cell regulation of the blood- 77. Amin FU, Shah SA, Kim MO. Vanillic acid attenuates Abeta1-42-induced brain barrier in diabetes. Diabetes Metab J. 2022;46(2):222–38. oxidative stress and cognitive impairment in mice. Sci Rep. 2017;7:40753. 55. van Veluw SJ, Hou SS, Calvo-Rodriguez M, Arbel-Ornath M, Snyder AC, 78. Liu Z, Kumar M, Kabra A. Cucurbitacin B exerts neuroprotection in a Frosch MP, et al. Vasomotion as a driving force for paravascular clearance murine Alzheimer’s disease model by modulating oxidative stress, in the awake mouse brain. Neuron. 2020;105(3):549–61. inflammation, and neurotransmitter levels. Front Biosci (Landmark Ed). 56. Poels MM, Ikram MA, Vernooij MW, Krestin GP, Hofman A, Niessen WJ, 2022;27(2):71. et al. Total cerebral blood flow in relation to cognitive function: the Rot‑ 79. Aggarwal H, Gupta S, Sharma P, Sharma BM, Sharma B. Neurobehavioral terdam Scan Study. J Cereb Blood Flow Metab. 2008;28(10):1652–5. and neurobiochemical effect of atomoxetine and N-acetylcysteine in 57. Shin HK, Jones PB, Garcia-Alloza M, Borrelli L, Greenberg SM, Bacskai BJ, streptozotocin diabetes induced endothelial dysfunction and related et al. Age-dependent cerebrovascular dysfunction in a transgenic mouse dementia. Physiol Behav. 2022;249:113767. model of cerebral amyloid angiopathy. Brain. 2007;130(Pt 9):2310–9. 80. Ayala A, Munoz MF, Arguelles S. Lipid peroxidation: production, metabo‑ 58. Keable A, Fenna K, Yuen HM, Johnston DA, Smyth NR, Smith C, et al. lism, and signaling mechanisms of malondialdehyde and 4-hydroxy- Deposition of amyloid beta in the walls of human leptomeningeal 2-nonenal. Oxid Med Cell Longev. 2014;2014:360438. arteries in relation to perivascular drainage pathways in cerebral amyloid angiopathy. Biochim Biophys Acta. 2016;1862(5):1037–46. 59. Squitti R, Mendez A, Ricordi C, Siotto M, Goldberg R. Copper in glucose Publisher’s Note intolerance, cognitive decline, and Alzheimer disease. Alzheimer Dis Springer Nature remains neutral with regard to jurisdictional claims in pub‑ Assoc Disord. 2019;33(1):77–85. lished maps and institutional affiliations. 60. Rempe RG, Hartz AMS, Bauer B. Matrix metalloproteinases in the brain and blood-brain barrier: Versatile breakers and makers. J Cereb Blood Flow Metab. 2016;36(9):1481–507. 61. Ugarte-Berzal E, Vandooren J, Bailon E, Opdenakker G, Garcia-Pardo A. Inhibition of MMP-9-dependent degradation of gelatin, but not other MMP-9 substrates, by the MMP-9 Hemopexin domain blades 1 and 4. J Biol Chem. 2016;291(22):11751–60. 62. Van Dyken P, Lacoste B. Impact of metabolic syndrome on neuroinflam‑ mation and the blood-brain barrier. Front Neurosci. 2018;12:930. 63. Zhang Y, Liu H, Chen Z, Yu M, Li J, Dong H, et al. TLR4-mediated hip‑ pocampal MMP/TIMP imbalance contributes to the aggravation of perioperative neurocognitive disorder in db/db mice. Neurochem Int. 2020;140:104818. 64. van Veluw SJ, Frosch MP, Scherlek AA, Lee D, Greenberg SM, Bacskai BJ. In vivo characterization of spontaneous microhemorrhage formation in mice with cerebral amyloid angiopathy. J Cereb Blood Flow Metab. 2021;41(1):82–91. 65. van Veluw SJ, Scherlek AA, Freeze WM, Ter Telgte A, van der Kouwe AJ, Bacskai BJ, et al. Different microvascular alterations underlie microbleeds and microinfarcts. Ann Neurol. 2019;86(2):279–92. 66. Carvalho C, Moreira PI. Oxidative stress: a major player in cerebrovascu‑ lar alterations associated to neurodegenerative events. Front Physiol. 2018;9:806. 67. Zinkevich NS, Gutterman DD. ROS-induced ROS release in vascu‑ Ready to submit your research ? Choose BMC and benefit from: lar biology: redox-redox signaling. Am J Physiol Heart Circ Physiol. 2011;301(3):H647–53. • fast, convenient online submission 68. Moretti R, Janjusevic M, Fluca AL, Saro R, Gagno G, Pierri A, et al. Common shared pathogenic aspects of small vessels in heart and brain disease. • thorough peer review by experienced researchers in your field Biomedicines. 2022;10(5):1009. • rapid publication on acceptance 69. Obadia N, Lessa MA, Daliry A, Silvares RR, Gomes F, Tibirica E, et al. Cer‑ • support for research data, including large and complex data types ebral microvascular dysfunction in metabolic syndrome is exacerbated by ischemia-reperfusion injury. BMC Neurosci. 2017;18(1):67. • gold Open Access which fosters wider collaboration and increased citations 70. Zhou X, Qian Y, Yuan D, Feng Q, He P. H(2) O(2) -induced microvessel • maximum visibility for your research: over 100M website views per year barrier dysfunction: the interplay between reactive oxygen species, nitric oxide, and peroxynitrite. Physiol Rep. 2019. https://​doi.​org/​10.​1007/​ At BMC, research is always in progress. s00592-​020-​01648-9. 71. He P, Talukder MAH, Gao F. Oxidative stress and microvessel barrier dys‑ Learn more biomedcentral.com/submissions function. Front Physiol. 2020;11:472.