Measuring K_f in Intact Tissues

The original experiments to measure K_f were performed by Pappenheimer, Renkin, and Borrero [338] in perfused hindlimbs of cats. Amputated hindlimbs of dogs or cats were suspended from a balance to gravimetrically determine changes in tissue fluid content [339]. Both arterial and venous vessels were cannulated and perfused with physiological solutions (filtered blood plasma, concentrated by evaporation, or diluted with Ringer's solution) at controlled fluid pressures (in vs. out) to establish a fixed pressure gradient across the microvascular circulation. Hence both ΔP and ΔΠ could be controlled experimentally. Under these conditions, fluid filtered out of the microvasculature into the tissue was measured as the rate of weight gain of the suspended tissue. These measurements were easily obtained because the hind limb microvasculature is a relatively simple model of endothelial barrier function. Other tissues that include endothelial and epithelial barriers, as well as multiple tissue compartments (e.g., lung or intestine), presented a greater challenge.

Lungs are exceptionally susceptible to edema during local or systemic inflammation or infection. Guyton and Lindsay [174] estimated fluid filtration and K_f in the intact lungs of dogs by controlling the left atrial perfusion pressure, and then measuring differences in wet vs. dry weight of the lung tissue. Differences in wet/dry weights reflected differences in water accumulation in the tissue, proportional to fluid filtration out of the microvasculature. By measuring the arterial and venous blood pressure, as well as colloid osmotic pressure of the blood plasma, these investigators could approximate K_f of the pulmonary microvasculature. However, because fluid accumulation in the lungs in these experiments includes filtration across the endothelium, as well as subsequent filtration across the epithelium into the alveolar air space, these measurements are underestimations of K_f. For this reason, Drake, Gaar, and Taylor [114] devised an isolated lung perfusion method similar to that used for hindlimbs. Lungs and heart were removed together (from dogs), and the lungs were mechanically ventilated while the heart was perfused separately through the pulmonary artery and the pulmonary vein. These tissues were perfused by gravity from a reservoir on the arterial side, and outflow was drained to a fluid reservoir on the venous side. Tissues were suspended from a force transducer to measure changes in weight due to fluid accumulation in the tissues. The tissues were initially perfused at a constant pressure until the tissue weight reached steady-state. Next, a small step change in pressure (ΔP) was introduced across the system (by equally elevating both venous and arterial perfusion reservoirs). Following this step change, these investigators observed both a rapid and a slow phase of weight gain (fluid accumulation), which are attributed to microvessel filling and microvascular filtration, respectively. The rate of steady-state weight gain in the slow phase (Δw/Δt) was used to calculate:

14

Measuring L_p With Colorimetric Dyes

Levick and Michel [263] described an optical method for measuring the fluid filtration coefficient (hydraulic conductivity (L_p)) across the wall of an individual perfused microvessel in intact frog mesentery. This method used a colorimetric (densitometric) microscope assay to measure fluxes of labeled macromolecules across the microvessel wall, under various constant pressure conditions. An individual cannulated microvessel within the mesentery was perfused at one end and occluded at the opposite end. Using this configuration, similar to the red blood cell method, fluid flow into the microvessel can only occur if there is fluid leakage (outflow) across the capillary wall. A colored macromolecule tracer was then introduced into the perfusion solution under constant pressure, and the rate of accumulation of the tracer in the microvessel lumen was determined by optical densitometry. This accumulation rate reflects the flow of solution into the microvessel and is equal to the fluid outflow through leak pathways across the microvessel wall. Upon a step change in perfusion pressure, there is a proportional change in the rate of fluid flow into the microvessel, which can be measured as an increased rate of dye accumulation (optical density) in the microvessel lumen. At two separate fixed, constant transluminal osmotic/oncotic pressures, the measured filtration rates (J_v/A) can be used to calculate L_p of the microvessel wall:

(17)

The Isolated, Perfused Microvessel Technique

While the method of Huxley et al. [208] has been very useful for examining microvessel permeability, this procedure is limited in that it can only be applied to semitransparent mesenteric membranes that can be spread across a microscope field of view. For many tissues, this type of preparation is not feasible. Because of the need to study microvascular permeability in parenchymal tissues, Yuan and Granger [513] developed a model measuring albumin permeability in isolated and perfused microvessels (postcapillary venules). Briefly, microvessels of 20 to 70 μm diameter (0.5 to 1.2 mm length) are manually dissected and stripped of extraneous tissue. While submerged in a physiological salt solution, the isolated venule is cannulated with a micropipette-in-pipette at each end of the vessel with each pipette connected to an individual reservoir (Figure 9). One reservoir is filled with physiological wash solution and the other with either the fluorescent tracer (e.g., FITC-albumin) or a chemical treatment. Either reservoir may be selected for perfusion, through a system of appropriate valves. In addition, using the two-cannula system, flow can be arbitrarily directed in either direction as needed. The reservoirs feeding the inflow and outflow pipettes feature adjustable heights such that inflow and outflow pressures can be precisely controlled. Fluorescence video microscopy is used to monitor leaks, and to perform experimental measurements within a selected field of view containing a venule, of known size. Flow rate is determined by an optical Doppler velocimeter by perfusing the system with a dilute suspension of red blood cells. Fixed pressure gradients may be applied across the length of the venule by adjusting the height of the perfusion reservoirs. For example, a pressure gradient (ΔP) of 10 cm H₂O may be achieved by setting the inflow reservoir to 20 cm H₂O, and the outflow reservoir to 10 cm H₂O. In addition, the oncotic pressure gradient (ΔΠ) can be controlled with fixed albumin concentration in the perfusion solution. Hence, permeability measurements can be performed under conditions where all relevant Starling forces are known.

FIGURE 9

The isolated, perfused microvessel permeability assay developed by Yuan and Granger. Upper panel: a schematic diagram of the experimental configuration for isolated venule perfusion. Individual venules are manually isolated from intact living tissues (more...)

The apparent permeability coefficient of albumin is analyzed based on the changes in fluorescence intensity collected from an optical window. Following a step change in fluorescence intensity corresponding to the flow of fluorophore solution through the vessel lumen (ΔI_f), there is a gradual increase in intensity across the entire field of view, corresponding to leakage and the appearance of fluorophore into the bathing solution surrounding the vessel (dI_f/dt). The apparent solute permeability coefficient of albumin (P_a) is calculated according to Equation (19) (Figure 9) [513]. This method allows for precise measurement of permeability in a real microvessel, yet without interference from ancillary parenchymal cells or systemic factors that are present in vivo. In addition, the chemical environment (drugs, etc.) and shear conditions (flow velocity, etc.) can be altered as needed, enabling assessments of their direct effects on endothelial permeability.

Yuan and co-workers have subsequently modified this technique for studying leukocyte–endothelium interactions [518]. During inflammation, leukocytes in the blood circulation respond first by binding to the microvascular endothelium, and then crossing the microvessel wall and invading the extravascular tissue (discussed in a later chapter). Using the isolated microvessel perfusion technique, Yuan et al. introduced fluorescent-labeled neutrophils through the inner inflow pipette, at various constant flow rates, and monitored interactions with the microvessel wall using frame-by-frame analysis of fluorescence microscope high-resolution video images. Under normal physiological treatment conditions, neutrophils in contact with the microvessel wall could be seen moving at a slower velocity (rolling), compared to the flow rate of the perfusate (measured with an optical Doppler velocimeter by perfusing red blood cells at 1% hematocrit). Rolling was quantified as the percentage of neutrophils displaying this behavior during the recorded period. Next, these investigators increased the perfusion pressure gradient by equivalently raising and lowering the inflow and outflow reservoirs, respectively, and observed a decrease in rolling as a function of increasing pressure (2.5 to 20 cm H₂O). Under inflammatory conditions, activated neutrophils can adhere to and cross the microvessel wall. Following preactivation by exposure to complement factor C5a, neutrophils in the microvessel perfusion assay were seen immobilized next to the microvessel luminal surface (adhesion) in isolated coronary venules. By increasing the perfusion pressure (2.5 to 20 cm H₂O), rolling and adhesion events were similarly attenuated in proportion to perfusion pressure. Therefore, adhesion was dependent upon neutrophil activation and normal physiological neutrophil rolling behavior. These investigators also noted that activated neutrophil adhesion occurred in isolated venules, and not in coronary arterioles, supporting the long-held notion that postcapillary venules are the principle site of neutrophil extravasation in the microvasculature. Therefore, the isolated microvessel perfusion method is useful for studying and quantifying leukocyte rolling and adhesion behavior and for observing the differences in these interactions between venules and arterioles.

Intravital Microscopic Measurement of Transvascular Flux

Macromolecules (albumin or dextran) conjugated to a fluorescent probe (e.g., FITC or TRITC) are also frequently used to monitor changes in microvascular leakage in intact tissues conducive to intravital microscopy [183, 328, 487]. Fluorescent-labeled molecules administered intravenously can be visualized in the microvasculature of semitransparent tissues, such as the mesentery, cremaster muscle, and hamster cheek pouch. Ley and Arfors [269] developed a protocol wherein intravenous FITC-dextran was visualized in the microvessels of the hamster cheek pouch by fluorescence microscopy, photographically recorded at various times for densitometric analysis. The greyscale intensities of images of microvessel and extravascular FITC-dextran, as well as a series of cuvettes containing known concentrations of FITC-dextran (standards) were determined by densitometry. These values were then used to calibrate the concentrations of FITC-dextran leakage into the tissues. This method of quantitation was greatly improved by Duran and coworkers [25] who used video recording and digital image analysis to record and calibrate concentrations of FITC-dextran. Using this method, the image optical intensity of a microvessel could be quantified immediately after intravenous injection of FITC-dextran, before any leakage had occurred. Then, knowing the serum FITC-dextran concentration, the concentration of dye leakage outside the microvessel could be determined (provided that fluorescence intensity remained within a predetermined range of linear correspondence to FITC-dextran concentration (0.4 to 3.0 mg/mL in this study). Integrated optical intensities (IOI) are determined by measuring pixel density (and greyscale intensities) of defined regions within an optical window of fixed dimensions and pixel size. IOI values can also be used to express microvascular leakage as the ratio of the transmural intensity difference to the original intraluminal intensity at time = 0 (Figure 10) [205]:

24

FIGURE 10

The use of fluorescence microscopy and digital integrated optical intensity (IOI) to measure macromolecular leakage in intact microvessels in vivo. Upper: fluorescence microscopy images of intact microvessels from mouse gut mesentery with an intravascular (more...)

Using the same data set, microvascular leakiness can also be expressed as the number of leakage sites per defined area (e.g., 100 μm²) in the microscope field of view [205]. Mayhan and co-workers [297] originally described this method using a hamster cheek pouch preparation. FITC-dextran was administered as described in the above preparation, and leaky sites were quantified by acquiring fluorescence microscope images from 10 randomly sampled fields of view within a single preparation at each experimental sampling time. Leaky sites were defined as fluorescent spots greater than 50 μm in diameter. Leaky sites counted from 10 fields were expressed as the number of sites per cm². In this study, the plasma (perfusate) fluorescence and external (superfusate) solution fluorescence were quantified in parallel with imaging, and used to demonstrate disappearance of tracer from the perfusate and appearance of tracer in the superfusate, coinciding with the number of leaky sites. The time dependence and quantitative results from preparations treated with histamine showed excellent correlation between the incidence of leaky sites and the fluorescence leakage measured directly in perfusate/superfusate solutions, indicating the validity of this method.

Transwell Solute Flux Assays

In vivo perfusion models of microvascular permeability have shown that transvascular solute flux is determined by solvent drag in convective fluid flow (J_v), and by diffusive forces. In the absence of convective fluid flow (J_v = 0), solute flux is determined entirely by diffusive forces:

25

26

27

28

FIGURE 11

Transwell solute flux assay for cultured endothelial cell monolayers. Endothelial cells are cultured in growth medium on a porous filter membrane until confluency is attained. This filter membrane forms the floor of a fluid-filled chamber that is then (more...)

Transendothelial Electrical Resistance Measurements

In pore theory, increased transendothelial permeability is due to an increase in the number of open pores at endothelial cell–cell junctions. Open pores are fluid-filled pathways that allow the passage of water and aqueous solutes across the endothelium. In contrast, the endothelial membranes are composed of lipophilic molecules (phospholipids, cholesterol, etc.) that restrict the passage of aqueous solutions, and that function as electrical insulators to maintain the physiological endothelial membrane electrical potential. Taken together, this implies that fluid-filled pores at cell–cell junctions also function as electrical conductance pathways across the endothelium. Therefore, an increase in the number of open pores will be manifested as a decrease in electrical resistance across the microvascular endothelium. The electrical resistance across microvessel walls has been directly measured in vivo in a variety of animal models. For example, Crone and Olesen [92] measured the electrical resistance of the walls of brain surface pial microvessels in live frogs by impaling the vessels with glass microelectrodes at specified distances apart. Knowing the resistivity of the plasma, and the dimensions of the microvessel (i.e., cable properties), they measured the decay in electrical current across the interior of the microvessel, and calculated the electrical resistivity of the microvessel walls: 1870 Ohm cm². They noted that this measurement corresponded well to prior estimations of transendothelial electrical resistance (TER) performed by calculating membrane conductance based on impedence to the flow of small inorganic ions across the microvessel wall, suggesting that TER represents pores for solute movement across the microvessel wall.

Measurements of TER are challenging to perform in vivo, but are relatively simple to perform with cultured endothelial cell monolayers. TER can be measured using the two-compartment transwell configuration (Figure 11), where electrical resistance is measured directly across the cell monolayer using an electrical resistance meter, by inserting probes into the luminal and abluminal compartments. The resistance of the filter membrane is first measured in the absence of cells, and then is subtracted from the total resistance in the present of a confluent cell monolayer to yield the resistance of the cell monolayer, per unit surface area (specific resistance; Ohm cm²). This method is often used in conjunction with transwell solute flux assays (described above), to determine the monolayer confluency/tightness prior to experimentation, or as an additional measurement to accompany PS determinations.

A more sophisticated, quantitative method for determining the transendothelial electrical resistivity is with an electrical cell–substrate impedance sensor (ECIS) [445]. The ECIS device, invented by Giaever and Keese [156] to quantify cultured cell density and migration, consists of a small cell culture chamber with two gold plate electrodes embedded in the bottom surface (Figure 12). Endothelial cells are grown to confluency in growth medium solution, directly on top of the electrodes. ECIS experiments are performed by applying alternating current (AC) across the electrodes and measuring electrical impedance. Impedance is analogous to resistance, but accounts for all resistances and reactances in an AC circuit. In the ECIS circuit, current flows through the medium and across the cell monolayer, such that the cell membranes function as a capacitor, and the pores or gaps between cells function as resistors. Electrical capacitance is proportional to the amount of cells or confluency of the monolayer. Opening or closing of intercellular gaps or pores across a confluent endothelial monolayer will be manifested as changes in ECIS resistance [445].

FIGURE 12

Electrical cell impedance sensor (ECIS) assay for cultured endothelial cell monolayers. Endothelial cells are grown to confluency on a cell culture plate with two gold electrode contact surfaces. These electrodes are connected to electrical alternating (more...)

The resistance component of ECIS measurements includes resistance attributed to fluid-filled spaces between cells (R_b) (cell–cell adhesive barrier) as well as that of the fluid volume trapped in the space between the basolateral cell surface and the ECIS electrode (α) (indicator of cell–substrate adhesion) [156]. These properties (R_b and α) can be determined by examining the impedance at varied frequencies and fitting the data to the mathematical model described by Giaever and Keese. The cell radius (r) is an empirically measured value (modeled as a sphere in this example), the medium resistivity (ρ) is measured directly in the absence of cells, and the relationship, α = r * (ρ/h)^0.5 is used to calculate the height (h) of the subcellular space. The value of h will increase as cell focal adhesions detach from the electrode surface. Therefore, ECIS can be used to quantitatively model changes in endothelial barrier properties attributed to intercellular gap formation vs. detachment at focal adhesions [223].