Comparative analysis of methods used for assessing blood-brain barrier permeability with fluorescent probes in laboratory animals

Full Text

Introduction

The blood-brain barrier (BBB) as part of the neurovascular unit of the brain (at the capillary level) represents a complex multicellular structure comprising cerebral endothelial cells, pericytes, perivascular astroglia, and an extracellular matrix organized as a thin porous membrane. This structure maintains chemical homeostasis in the central nervous system (CNS). Along with other types of histohematic barriers in the brain (e.g., blood-CSF, blood-arachnoid barriers), it regulates controlled entry of chemicals into brain tissue, metabolizes hydrophobic xenobiotics, removes metabolites and bioactive molecules into peripheral blood, and integrates systemic metabolic/immune responses with local processes governing brain plasticity [1].

Barriergenesis (BBB formation) exhibits unique features in pre- and postnatal development: in rodents, BBB develops on embryonic days 10–17 (initially incorporating endothelial cells and pericytes, later astrocytes), with controlled permeability established by embryonic day 21, while endothelial junction maturation continues into early postnatal stages. Notably, humans demonstrate BBB marker emergence at 8 weeks of embryogenesis, with cerebral angiogenesis persisting until 2–3 weeks postnatally. Cerebral barriergenesis commences only after establishing a pool of neural stem cells and neuronal progenitors. Postnatal brain barriergenesis correlates with neoangiogenesis (new capillary formation during experience-driven plasticity or tissue regeneration) [2–4].

Transport of substances and cells across the BBB is known to be provided by two types of permeability:

a) paracellular permeability (between endothelial cells of cerebral capillaries), which is limited by tight intercellular junctions formed by various proteins (claudins, occludins, etc.), where the molecular organization of tight junctions depends on the activity of endothelial mechanoreceptors activated by fluid flow during shear stress, conferring high electrical resistance to the endothelial layer within the BBB (over 1500–2000 Ω/cm²);

b) transcellular permeability (through endothelial cells of cerebral capillaries), which restricts penetration of hydrophilic molecules with molecular mass exceeding 400 Da into brain tissue and is mediated by the activity of a wide range of transporter proteins — for transfer of organic anions and cations, nucleosides, and nucleotides [5].

The limited transport of chemical compounds and water across the BBB also depends on effective interaction of endothelial cells with perivascular astrocyte end-feet, pericytes, and the integrity of the extracellular matrix layer within the basement membrane, which has a thickness of 30–200 nm, pores ranging from 5 nm to 8 μm, and includes various types of collagens, laminin, nidogen, perlecan, agrin, thrombospondins, and other proteins [6]. At the same time, the use of certain fluorescent molecules in BBB in vitro models has demonstrated the possibility of transporting compounds with molecular mass above 600 Da across the BBB (e.g., 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein tetraacetoxymethyl ester — BCECF-AM, molecular mass 809 Da) [7].

The barrier functions within the BBB are primarily attributed to cerebral endothelial cells, which have evolutionarily acquired properties enabling effective regulation of interactions between blood flow and the CNS, thereby establishing a BBB with limited permeability (in early evolutionary stages, this role was assigned to astroglial cells) [6]. Cerebral endothelial cells possess both properties common to endothelial cells in other locations (secretory activity, including production of vasoactive molecules such as nitric oxide; antithrombogenic activity; ability to interact with blood cells; angiogenic potential) and properties that significantly distinguish them from endothelial cells of other organs and tissues. Specifically, the endothelial layer in brain microvessels is virtually non-fenestrated (facilitated by low expression of MECA-32 protein), intercellular interactions are controlled by expression of tight and adherens junction proteins, pinocytotic capacity is limited, and transcellular/paracellular permeability depends on efficient mitochondrial function — mitochondria being highly abundant in cerebral capillary endothelial cells. Furthermore, BBB endothelial cells express biotransformation complex enzymes (including cytochrome P450 isoforms — CYP1A1, CYP1B1, CYP3A4, CYP2C9, CYP2C19, CYP2A6, CYP2E1 — which facilitate phase I xenobiotic metabolism). Cerebral microvascular endothelial cells exhibit high metabolic activity (combining glycolysis and oxidative phosphorylation, with the ability to utilize ketone bodies and fatty acids that undergoes significant changes during angiogenesis and barriergenesis) [8–10].

On the one hand, the limited BBB permeability hinders the penetration of large numbers of endogenous molecules and xenobiotics, including pharmacological agents, into brain tissue; on the other hand, excessively high and uncontrolled BBB permeability is an obligatory component in the pathogenesis of CNS diseases characterized by neuroinflammation, hypoxia/ischemia, autoimmune response, mechanical damage, and stress-induced tissue disorders [1]. Experimental and clinical studies confirm the contribution of endothelial layer integrity disruption, dysmetabolic events in endothelial cells, loss of pericyte and astrocyte contacts, and basement membrane protein degradation to the pathogenesis of increased BBB permeability [11–13]. Such mechanisms are typical, for example, for neurodegeneration progression, particularly Alzheimer’s disease, where endothelial dysfunction, impaired mitochondrial dynamics in brain capillary endothelial cells, pericyte loss, extracellular matrix damage, and aberrant (non-productive) neovascularization are documented [14–16], leading to neuroinflammation due to BBB integrity impairment [17]. Events associated with high BBB permeability warrant separate mention: those resulting from immaturity of endothelial cell protein machinery (during early postnatal ontogeny or intense neovascularization, including in Alzheimer’s-type neurodegeneration) or occurring transiently and reversibly during synaptic transmission activation in neurons [18, 19].

In clinical practice, BBB permeability is assessed using several protocols:

a) infrared spectroscopy with indocyanine green, which exhibits rapid clearance from brain tissue;

b) high-resolution contrast-enhanced MRI – based on accumulation of gadolinium-based contrast agents (gadopentetate dimeglumine, gadodiamide, gadoterate meglumine, gadobutrol, gadoteridol) in the extracellular perivascular space, leading to increased longitudinal relaxation time and signal intensity on T1-weighted images;

c) positron emission tomography with radioligands enabling BBB permeability assessment, e.g., using 2-amino-3C-isobutyrate [20].

In experimental studies utilizing models of brain diseases in laboratory animals, the challenge of evaluating BBB permeability is typically addressed through dyes or fluorescent tracers, which due to their physicochemical properties exhibit varying penetration capacities across the barrier into brain tissue, and whose tissue accumulation can be detected by relatively simple methods (microscopy, spectrophotometry, spectrofluorimetry).

Fluorescence analysis for assessing BBB permeability in vivo

A classic example of using fluorescent tracer molecules to assess BBB permeability in experiments is Evans Blue (EB), an anionic azo dye (961 Da), synthesized by H. Evans in 1914, which has a high affinity for serum albumin. Upon intravenous administration, EB rapidly binds to albumin, forming a complex (67 kDa) that is normally too large to pass through the tight junctions of the cerebral endothelium. The binding of EB to albumin induces a conformational shift that reduces the rotational freedom of the molecule and triggers fluorescence in the red and far-red spectrum (excitation: 546–620 nm, emission: 680 nm) [21]. As a result of binding to blood proteins, approximately 30–50% of the initial concentration of albumin labeled with EB remains in the blood plasma even 24 hours after administration of the fluorescent dye [22].

If the BBB is compromised, the EB-albumin complex can escape from blood vessels into the brain parenchyma, staining affected areas blue, while albumin undergoes endocytosis by endothelial cells, resulting in staining of the endothelial cell layer by the dye. After a specified circulation period of the complex in the blood, a protocol is performed to remove non-penetrated complexes from cerebral vessels (perfusion), and the amount of EB in the tissue (i.e., the complex captured by damaged brain cells or distributed in the extracellular matrix) is determined microscopically or via spectrophotometry [23]. The staining intensity reflects the degree of vascular permeability. Compared to other dyes, EB has gained the greatest popularity due to its stronger binding to blood albumin and prolonged retention in the bloodstream [24]. EB’s predominant binding to blood albumin makes it a perfect marker for BBB permeability to proteins, whereas, for example, gadolinium-DTPA (552 Da) is typically used as a marker for barrier permeability to low-molecular-weight compounds [22]. Key aspects of the EB-based protocol application are summarized in Table 1.

 

Table 1. Brief characteristics of the main steps in the protocol for assessing BBB permeability using EB in in vivo models [30, 31]

Protocol Step	Main Procedures
ВDye introduction, sample preparation	EB is administered via the tail vein (80 mg/kg or 2% solution in saline, 2 ml/kg) 30 min, 1 h, 6 h, or 24 h before euthanasia. Intracardiac administration at the same concentration is also possible. Dye penetration into tissues is assessed by blue staining of the paws and ears. Transcardial perfusion with phosphate buffer (to remove residual EB-albumin complexes from the vascular lumen) is performed until the right atrium loses color; followed by perfusion with 4% paraformaldehyde in phosphate buffer (0.1 M, pH 7.4), immersion in 30% sucrose, and embedding in a cryotomy matrix medium (O.C.T. Compound) with freezing to –80°C
Tissue spectral analysis	The brain is extracted, photographed, and homogenized in cold phosphate buffer. Proteins are precipitated with 30% or 50% trichloroacetic acid, and the sample is centrifuged at 14,000 g for 10–30 min. The spectral characteristics of the homogenate are assessed. A calibration curve is constructed using the original EB solution, and the result is expressed as μg of EB per mg of brain tissue
Visualization	30-μm-thick sections are prepared using a cryostat, and nuclei are stained with DAPI dye. The study is conducted using a fluorescence microscope. To assess which cells have taken up EB, immunohistochemistry is performed with detection of markers specific to different cell types, and fluorophore-labeled secondary antibodies can be visualized, for example, in the green fluorescence channel. Visualization can be performed using wide-field microscopy (whole organ) or a fluorescence (confocal) microscope (cryo- or paraffin-embedded sections). To improve visualization of small vessels (capillaries), it is recommended to add glycerol to EB (0.5% EB solution in glycerol)

 

The efficacy of this method for visualizing foci with pathologically high paracellular BBB permeability has been demonstrated in numerous experimental studies. In rodent stroke models, EB extravasation delineates the boundaries of the ischemic core and the area of BBB disruption, which often correlates with regions at risk of hemorrhagic transformation. In traumatic brain injury or concussion models, EB extravasation confirms diffuse microvascular damage and is used to track BBB recovery over time [26]. In neuroinflammatory conditions, such as experimental autoimmune encephalomyelitis, focal EB extravasation corresponding to inflammatory lesions is observed and can be utilized to assess pharmacotherapy efficacy. In tumor progression, visualization of tumor boundaries is achievable by evaluating the pattern of EB spread in tissue. Researchers also combine EB with other tracers (e.g., fluorophore-labeled dextrans) to compare BBB permeability for large versus small molecules.

Another fluorescent probe successfully used for assessing BBB permeability in vivo is sodium fluorescein (NaF) — a small (376 Da) fluorescent dye that was the first low-molecular-weight marker in BBB research [30–32]. NaF exhibits low toxicity even at relatively high doses [33], enabling its use in several clinical protocols [34]. Unlike EB, NaF binds weakly to plasma proteins, remaining predominantly in free form in the bloodstream (up to 70% of administered tracer), and it is this protein-unbound form of the dye that is detected in brain tissue upon BBB damage [35]. When excited at approximately 440 nm, emission at about 525 nm is observed, measurable via spectrofluorimetry or visualizable in tissue sections [36]. In vitro BBB models have demonstrated that cerebral endothelial cells can efflux NaF into the extracellular space through organic anion transporter activity and, to a lesser extent, P-glycoprotein; this mechanism is disrupted by suppression of Na⁺,K⁺-ATPase activity and hypoxic-ischemic tissue injury [37]. Thus, it is reasonable to conclude that NaF accumulation in brain tissue reflects not only BBB permeability but also tissue clearance mechanisms, whose assessment is equally important in the context of neurodegenerative disease pathogenesis. Table 2 summarizes key protocols using NaF for in vivo BBB permeability assessment.

 

Table 2. Brief characteristics of the main protocol steps for assessing BBB permeability using NaF in in vivo models [36, 37, 39]

Protocol Step	Main Procedures
Dye introduction, sample preparation	NaF is dissolved in physiological saline (20 mg/kg animal body weight) and administered via the tail vein, circulating for 1–2 hours, followed by intracardiac perfusion with phosphate buffer in an anesthetized animal. Prior to perfusion, a blood sample (serum) is collected for subsequent analysis. Post-perfusion, the brain is harvested, tissue lysed, and centrifuged (10 min, 13,000 rpm). The supernatant is used for further analysis. In some cases, intracarotid administration of mannitol (20% solution, 0.5 g/kg animal body weight) is recommended to enhance BBB permeability and verify protocol results
Tissue spectral analysis	Using a microplate spectrofluorimeter, dye content in blood serum and brain tissue homogenate is analyzed (excitation: 492 nm, emission: 525 nm). BBB permeability to NaF is calculated as the ratio of fluorescence signal in tissue homogenate to serum. The fiber-optic photometry method may be applied to record fluorescent signals in brain tissue
Visualization	Fluorescence imaging in brain sections is performed using wide-field microscopy (whole organ) or luminescent (confocal) microscopy (cryo- or paraffin sections) in the green channel

 

Interestingly, unlike EB, NaF is predominantly evenly distributed throughout the brain tissue when BBB permeability is impaired, whereas EB accumulates maximally in the prefrontal cortex and cerebellar tissue [38].

Fluorescent tracers widely used for in vivo BBB permeability analysis include sulforhodamine B (559 Da), rhodamine-123 (344 Da), and fluorescein isothiocyanate (FITC)-labeled dextrans of varying molecular weights [36]. However, each tracer has specific application nuances: for instance, while NaF accumulation in brain tissue enables differentiation between fluorescence signals from vascular walls and perivascular spaces, sulforhodamine B uniformly labels both microvessels and brain parenchyma. Rhodamine-123 localizes in clusters, likely due to mitochondrial binding, but fails to visualize vessel walls, as is also the case with FITC-labeled dextrans [36].

Among fluorescent exogenous labels for assessing BBB status, fluorophore-labeled dextrans (4–150 kDa) warrant special attention; their penetration into brain tissue through the BBB depends on the degree of barrier damage (higher permeability enables greater transport of lower molecular weight dextrans into brain parenchyma). These tracers do not remain in the vascular wall but accumulate in perivascular spaces when BBB integrity is compromised [22]. When using labeled dextrans, it should be noted that their circulation time in blood after animal administration is significantly shorter than, for example, EB (≤ 2 min), as this probe clears faster from the bloodstream. Furthermore, while EB redistributes from intravascular space to the reticuloendothelial system, FITC-labeled dextrans undergo renal elimination, with excretion rates inversely proportional to molecular weight [22]. It should also be considered that low-molecular-weight dextrans may be washed out from brain tissue during sample processing (due to higher solubility in paraformaldehyde and sucrose), potentially obscuring their fluorescence signal unless specialized protocols are employed (e.g., tissue freezing with 2-methylbutane on dry ice) [22].

As described above, fluorescence microscopy methods, including confocal microscopy, are most commonly used to visualize the results of fluorescent tracer application in brain tissue. At the same time, in this context, fluorescent multiphoton microscopy also finds application, which is based on the effect of two- and three-photon fluorescence, in which a fluorophore molecule simultaneously absorbs 2 or 3 photons, transitioning to an excited state and emitting fluorescence. In such cases, fluorescent marker molecules are used, such as FITC- and Texas Red-conjugated dextrans [39–42]; FITC-albumin [43]; antibodies conjugated with fluorescent labels [41]; and endothelial cell-targeting fluorescent and fluorescently labeled agents [44–46]. Among the advantages of this group of methods, it is worth mentioning first of all the possibility of in vivo observation of the changes in the BBB state in the animal brain with high resolution [42, 47] and in real time [41, 44–46, 48], including in freely moving animals [48]. The disadvantages include a general technical limitation of almost all optical methods — a penetration depth limitation of 0.5–1.0 mm [40, 42], except in cases where excitation in the near-infrared spectral region is used [44].

Wide-field optical imaging in the near-infrared range (NIR-I and NIR-II), unlike high-magnification methods, can capture wider fields of view, such as the entire dorsal surface of a rat brain, without raster scanning [49]. Wide-field infrared imaging employs a range of fluorescent tracers, which can be categorized into low-molecular-weight organic dyes [49–51] and nanoparticle-based probes [52]. Wide-field optical imaging in the near-infrared range is limited to depths exceeding 2 mm. Light scattering is significantly reduced during NIR-II range imaging [53], enabling NIR-II wide-field imaging to detect contrast signals from depths of approximately 3 mm. The primary advantage of wide-field optical imaging in the red-light spectrum is its ability to study the BBB through the animal’s skull [48, 50].

The expansion of real-time in vivo visualization capabilities for blood vessels and brain parenchyma is associated with the use of label-free multiphoton microscopy. This type of microscopy utilizes second harmonic generation effects (two-photon effects, which in brain tissue are characteristic of vascular collagen and the extracellular matrix [54–56]), third harmonic generation (three-photon effects at phase interfaces, which is relevant, for example, for visualizing blood flow in cerebral vessels [57, 58]), and Raman spectroscopy methods (e.g., stimulated Raman scattering allows indirect assessment of BBB integrity through the state of microvessels [59]). All these methods are based on nonlinear optical phenomena in endogenous structures such as collagen, lipids, or cells in general [60], enabling real-time imaging of blood vessels and parenchyma with high resolution without the need for exogenous tracers [57]. However, despite all its advantages, label-free multiphoton BBB imaging has its limitations. Light scattering in the brain limits the imaging depth to several hundred micrometers under a cranial window or on the sample surface, while achieving depths greater than 1 mm requires either increasing the wavelength or using three-photon excitation with specialized optics [61]. Moreover, the field of view in label-free multiphoton microscopy is typically limited to hundreds of micrometers, which complicates the examination of large lesions. Finally, the interpretation of label-free signals can be challenging due to ambiguity in signal drop interpretation — thorough control and verification by histology or standard fluorescence microscopy are often necessary [62].

Table 3 presents a comparative analysis of the advantages and limitations of the main methods for assessing the BBB state and permeability in in vivo models.

 

Table 3. Advantages and disadvantages of the main methods for assessing the BBB state in experimental animals

Method	Advantages	Limitations
Post-mortem visualization with fluorescent probes	Simplicity of implementation, high sensitivity, visualization of extravasation regions	Inability to observe the BBB changes over time, questionable effectiveness of some popular tracers, need for expensive fluorescent optics
Post-mortem fluorimetry with fluorescent tracers	Simplicity of implementation, high sensitivity	Inability to observe the BBB changes over time, questionable effectiveness of some popular tracers, inability to study the micromorphology of damage
Label-free multiphoton microscopy	Real-time visualization, no need for a tracer	Need for craniotomy in in vivo application, compromise between wavelength and resolution, high cost of the study, risk of tissue photodamage
Wide-field imaging in near-infrared light	Intravital visualization of large fields of view in real time, relatively large depth of visualization, possibility of non-invasive use	Exclusively planar representation of the obtained data, small number of specialized tracers, need for highly specialized detectors, compromise between resolution, depth, and signal-to-noise ratio
Fluorescent multiphoton microscopy	Real-time visualization with high resolution	Need for craniotomy, complexity of intravital visualization and equipment, risk of sample overheating

 

Conclusion

Methods utilizing fluorescent probes constitutes a powerful toolkit for dynamic assessment of the BBB status in vivo, which holds fundamental significance for experimental neurobiology. These techniques enable monitoring of the structural and functional integrity of the barrier across a broad spectrum of pathological conditions, including neuroinflammation, chronic neurodegeneration, and acute cerebral tissue injuries. A strategically critical aspect is the appropriate selection of fluorescent probes, as their molecular weight, chemical properties, and specificity determine the feasibility of differential evaluation of key BBB homeostasis parameters. Specifically, variously sized probe allows selective analysis of paracellular (intercellular) and transcellular (through cellular cytoplasm) permeability, reflecting the functional activity of tight junctions and vesicular transport, respectively. Modern experimental approaches entail comprehensive assessment not only of permeability but also of the morphofunctional state of the brain microvascular network. Furthermore, fluorescent tracers have become indispensable for investigating complex physiological processes related to metabolite clearance from brain tissue. A pivotal recent advancement is the protease-activatable probes, whose levels significantly increase during BBB destabilization. Such activatable sensors enable visualization not merely of tracer presence in brain parenchyma but also real-time activity of proteolytic enzymes associated with neuroinflammation and angiogenesis.

Thus, the methodological arsenal based on fluorescent probes is evolving from simple detection of barrier dysfunction to multidimensional analysis of pathogenic mechanisms underlying these impairments, opening new avenues for preclinical evaluation of therapeutic interventions targeting BBB preservation and restoration.

About the authors

Arseniy K. Berdnikov

Russian Center of Neurology and Neurosciences

Email: annaly-nevrologii@neurology.ru
ORCID iD: 0009-0007-4195-2533

postgraduate student, Laboratory of neurobiology and tissue engineering, Brain Institute

Russian Federation, Moscow

Anton S. Averchuk

Russian Center of Neurology and Neurosciences

Author for correspondence.
Email: annaly-nevrologii@neurology.ru
ORCID iD: 0000-0002-1284-6711

Cand. Sci. (Biol.), Assoc. Prof., senior researcher, Laboratory of neurobiology and tissue engineering, Brain Institute

Russian Federation, Moscow

Yulia K. Komleva

Russian Center of Neurology and Neurosciences

Email: annaly-nevrologii@neurology.ru
ORCID iD: 0000-0001-5742-8356

Dr. Sci. (Med.), leading researcher, Laboratory of neurobiology and tissue engineering, Brain Institute

Russian Federation, Moscow

Ilia V. Potapenko

Russian Center of Neurology and Neurosciences

Email: annaly-nevrologii@neurology.ru
ORCID iD: 0000-0001-9743-8700

PhD, researcher, Laboratory of neurobiology and tissue engineering, Brain Institute

Russian Federation, Moscow

Alla B. Salmina

Russian Center of Neurology and Neurosciences

Email: annaly-nevrologii@neurology.ru
ORCID iD: 0000-0003-4012-6348

Dr. Sci. (Med.), Professor, Corr. Member of the RAS, principal researcher, Head, Laboratory of neurobiology and tissue engineering, Deputy director, Brain Institute

Russian Federation, Moscow

References

Supplementary files