Heart and Circulatory Physiology

Po2 profiles near arterioles and tissue oxygen consumption in rat mesentery

Aleksander S. Golub, Matthew C. Barker, Roland N. Pittman

This article has a correction. Please see:


A scanning phosphorescence quenching microscopy technique, designed to prevent accumulated O2 consumption by the method, was applied to Po2 measurements in mesenteric tissue. In an attempt to further increase the accuracy of the measurements, albumin-bound probe was topically applied to the tissue and an objective-mounted pressurized bag was used to reduce the oxygen transport bypass through the thin layer of fluid over the mesentery. Po2 was measured at multiple sites perpendicular to the blood/wall interface in the vicinity of 84 mesenteric arterioles (7–39 μm in diameter) at distances of 5, 15, 30, 60, 120, and 180 μm in seven anesthetized Sprague-Dawley rats, thereby creating Po2 profiles. Interstitial Po2 above and immediately beside arterioles was found to agree with known intravascular values. No significant difference in Po2 profiles was found between small and large arterioles, indicating a small longitudinal Po2 gradient in the precapillary mesenteric microvasculature. In addition, the Po2 profiles were used to calculate oxygen consumption in the mesenteric tissue (56–65 nl O2·cm−3·s−1). Correction of these values for contamination with ambient oxygen yielded an oxygen consumption rate of 60–68 nl O2·cm−3·s−1, the maximal limit for consumption in the mesentery. The results were compared with measurements made by other workers in regard to the employed techniques.

  • microcirculation
  • oxygen tension gradients
  • phosphorescence quenching microscopy

oxygen exchange between arterioles and tissue can be studied in the rat mesentery, where the density of microvascular networks is low and the spatial separation between arterioles and other vessels is such that gradients of tissue oxygen pressure (Po2) can be measured at sites radiating out from vessels of interest with good spatial resolution. The loose connective tissue surrounding arterioles in the mesenteric “windows” does not contain parenchymal cells, which have a high oxygen consumption rate. Therefore, the transparent, thin tissue anatomy and the low oxygen consumption of the tissue permit the utilization of a simple diffusion-consumption model for interpretation of the spatial distribution of oxygen in the tissue (i.e., Po2 profiles) and the recovery of oxygen transport parameters (e.g., oxygen permeability and consumption). All of these features make the rat mesentery ideal for the study of basic mechanisms of oxygen transport to tissue (18, 51).

Previous studies using distinctly different modifications of luminescence quenching techniques have resulted in reported values of oxygen consumption that differ by 30-fold (41, 49). Since the interpretation of the measured Po2 profiles obtained perpendicular to the vessel axis was calculated similarly in both studies, it is likely that one or both of these studies is plagued by a measurement artifact possibly related to the experimental method itself. Close examination of the techniques used in these studies provides some insight into the source of the measurement error.

Itoh et al. (22) and Yaegashi et al. (49) employed a fluorescent oxygen probe adsorbed into 3-μm-diameter silica gel beads, embedded in a 20-μm-thick silicone elastomer film. The average distance between the surface of the silicone film and the mesenteric tissue was 38.2 μm (range 30–70 μm), while the average thickness of the mesentery tissue was reported as 58.6 μm. The intensity of fluorescence, evoked by continuous epi-illumination, was used to obtain a map of the Po2 distribution over the rat mesentery. The Po2 profile was used for the computation of oxygen diffusivity and consumption by fitting experimental data with the theoretical parabola calculated from a one-dimensional model of diffusion and consumption in a slab of tissue. The Po2 values measured above the arterioles (animals breathing room air) in this study were between 60 and 70 mmHg and gradually decreased to a minimal level of ∼10 mmHg, 300–500 μm away from the vessel. The oxygen consumption (V̇o2) was estimated to be 8.2 nl O2·cm−3·s−1.

In another study, Tsai et al. (41) employed a phosphorescence quenching microscopy (PQM) technique and time domain analysis of the phosphorescence decay curves for the determination of Po2. Tsai et al. infused an albumin-bound phosphorescence probe into the circulation, where the probe became distributed throughout the plasma compartment. As a result of probe extravasation from microvessels into the perivascular interstitial fluid 20 min after injection, it was possible to collect phosphorescence signals from both the blood and the interstitium. The placement of a 5 × 20-μm detection window inside the image of arterioles chosen for study was associated with the signal coming from the blood, while positioning the window outside of the vessel image was associated with tissue Po2 measurements. Tsai et al. (41) reported intravascular Po2 values from mesenteric arterioles similar to those reported by the previous group of researchers (49). However, their averaged Po2 profile had a large Po2 drop across the vessel wall. Perivascular Po2, measured 5 μm from the exterior of the vessel wall, was found to be 18 mmHg lower than intra-arteriolar Po2, measured 5 μm from the blood/wall interface, and fell to near zero at a distance of 50–80 μm. This reported Po2 difference was greater than that found in measurements with microelectrodes in either skeletal muscle (11) or brain tissue (12, 44). The V̇o2 of rat mesenteric connective tissue (240 nl O2·cm−3·s−1) obtained by Tsai et al. (41) is 30 times higher than that reported by Yaegashi et al. (49). The V̇o2 of the arteriolar wall (65,000 nl O2·cm−3·s−1) calculated from the data of Tsai et al. (41) is 278 times higher than their tissue V̇o2.

Localized measurements of Po2 and determination of V̇o2 provide fundamental information required for the advancement of our understanding of oxygen transport to tissue. In light of the large discrepancy in V̇o2 values reported by these investigators (41, 49), it is necessary that we ascertain which, if either, provides the more accurate value. To this end, we have carefully analyzed each employed technique and have subsequently refined the PQM technique to minimize the impact of oxygen consumption by the method itself. We then applied this modified technique to a similar evaluation of oxygen consumption in the rat mesentery.


Several features in our design of the phosphorescence quenching method make it suitable for microscopic measurements of Po2 in a stationary tissue. To improve the localization of the signal and avoid unnecessary irradiation of the tissue and vasculature, a small aperture in the epi-illumination light path of the microscope was used. The adjustable aperture was parfocal with the objective lens and formed the size and shape of the excitation spot. The detection arm of the optical path was maximally opened for collection of emitted light so that detection efficiency was determined mainly by the aperture of the objective lens. Such a design provides high sensitivity that allows a good signal-to-noise ratio to be obtained at relatively low flash energy. The scanning excitation spot, in concert with a small excitation area and a low flash rate, makes it possible to attain Po2 measurements in a stationary tissue with minimal distortion of the measured values. Topical application of the probe directly to the tissue and elimination of residual fluid above the tissue by pressing the barrier film against the mesentery improve conditions for the accurate interpretation of periarteriolar Po2 profiles (in terms of the 1-dimensional diffusion-consumption model). A short, rectangular excitation pulse and a short delay in signal acquisition improve the quality of the signal necessary for exponential analysis of heterogeneous decays.

Phosphorescence microscope.

The phosphorescence quenching instrument used in the present study (see Fig. 1) consists of an AxioImager-A1m microscope (Zeiss, Germany) equipped on its two optical ports with a video camera and a custom-built photodetector. The excitation epi-illumination was provided by a transistor-transistor logic (TTL)-modulated diode laser (model PPMT25, Power Technology, Little Rock, AR; 410 nm) that generated short rectangular light pulses where frequency and duration were determined by the clock unit. The central part of the laser beam was isolated by an adjustable rectangular diaphragm to provide the necessary uniformity and a 5 × 5-μm excitation area. The beam was reflected by a SC-20 resonant optical scanner (Electro-optical Products, Glendale, NY) into the epi-illumination port of the microscope. A rotary Dove prism (Edmund Optics, Barrington, NJ) was installed into the beam path to control the direction of the scan. The resonance frequency of the oscillating mirror was 10.3 Hz, and the excitation flash rate was set at 11 Hz. The beat frequency between 11 and 10.3 Hz was 0.7 Hz, providing a sinusoidal scan for 1.43 s along the 57-μm scanning distance, practically excluding overlap of sequential excitation areas. Thus 1 scan cycle produced 11 discrete phosphorescence decays along the scan line.

Fig. 1.

Diagram of the laser scanning phosphorescence microscope for Po2 measurements in a stationary tissue and fluid. See text for details. PMT, photomultiplier tube.

The laser beam was reflected by a dichroic mirror (505 DCXR, Chroma Technology, Rockingham, VT) through the objective lens (×40, water immersion, Zeiss) onto the tissue. Phosphorescence emission was detected by a photomultiplier tube (PMT) (model R3896 with a high-voltage socket HC123-01, Hamamatsu, Bridgewater, NJ). A barrier filter (reflecting interference cut-on 650 nm; Thermo Oriel, Stratford, CT) was located in front of the PMT in order to provide viewing of the excitation light spot and microvessels in the same image via a video camera. A mirror was used to toggle between two fixed positions, one for imaging and the other for phosphorescence detection by the PMT.

The PMT signal was directed to a gated amplifier, digitized, and monitored with an oscilloscope (model 72-3060, Tenma, Springboro, OH). The 12-bit analog-to-digital converter (ADC) was controlled by a custom program written in LabVIEW-7 (National Instruments, Austin, TX). For each Po2 measurement, a sequence of 100 curves, composed of 400 data points per curve (curve duration 2 ms), was recorded and averaged, so that the original set of curves and the averaged curve were available for review and analysis. The acquired phosphorescence decay curves were analyzed off-line with Origin 7.0 (OriginLab, Northampton, MA) and with an automated fitting program constructed in LabVIEW.

Linear dimensions and distances were measured on a video monitor with a magnification of 0.35 μm/mm and a field of view of 180 μm (diagonal). The excitation scan line in the tissue was oriented parallel to the axis of the arteriole with the rotary Dove prism, and its position was marked on a thin plastic screen placed on the flat panel monitor.

The clock unit (see Fig. 1) produced trains of synchronous TTL pulses of 2-μs and 4-μs duration at a frequency of 11 Hz. Each 2-μs pulse induced a laser light pulse of the same duration. The 4-μs rectangular pulses were used to disable the PMT amplifier (gating) for the period covering the duration of the excitation pulse and most of the fast postexcitation transient. Since the sampling period of the ADC was set at 5 μs, the phosphorescence decay curve was acquired from the first point, as it is necessary for exponential analysis (21).

The energy density of the excitation light pulse was 28 pJ/μm2 (measured with a Lasercheck power meter; Coherent, Auburn, CA). The pulse energy density was adjusted to the lowest value delivered to tissue by a Xe-flash lamp reported by other workers (see Table 1) in order to have the same level of oxygen consumption by the method (∼0.3 mmHg/flash; Ref. 31).

View this table:
Table 1.

Excitation energy density in PQM instruments

Animal preparation.

Seven female Sprague-Dawley rats (246 ± 5 g; Harlan, Indianapolis, IN) were used in this study. Animals were initially anesthetized with a combination of ketamine and acepromazine (75 and 2.5 mg/kg ip). After catheterization of the right jugular vein for administration of supplemental anesthetic, surgical preparation and measurements were conducted under a continuous intravenous infusion of alfaxalone-alfadolone (Saffan, Schering-Plough Animal Health, Welwyn Garden City, UK; 0.1 mg·kg−1·min−1). On conclusion of experimentation, the animals were euthanized with intravenous administration of 0.4 ml/kg Euthasol (pentobarbital 390 mg/ml and phenytoin 50 mg/ml; Delmarva Laboratories, Midlothian, VA). All procedures and protocols were approved by the Virginia Commonwealth University Institutional Animal Care and Use Committee.

Initially, the trachea was cannulated (PE-240, Clay Adams, Parsippany, NJ) to ensure a patent airway for spontaneous breathing of room air. Catheterization of the left carotid artery with PE-50 tubing allowed for monitoring of mean arterial pressure (CyQ 103/301, Cyber Sense, Nicholasville, KY). The carotid artery was also used for arterial blood sampling, where the measured parameters included Po2, Pco2, and pH (ABL 705, Radiometer America, Westlake, OH).

Immediately before the mesentery was prepared for intravital microscopy, a thin, 2-mm-diameter oxygen electrode (InO2; Innovative Instruments, Tampa, FL) was inserted through a small hole in the abdomen allowing access to the intraperitoneal space, so that Po2 measurements of peritoneal fluid could be made. The animal was then placed on its side on a thermostatic animal platform close to the organ pedestal (16). The heated sapphire window was painted with a drop of yellow fluorescent ink, covered with a polyvinylidene chloride (PVDC) film (Saran Wrap; Dow Chemical, Indianapolis, IN), and then the film was fixed with a neoprene ring around the pedestal, as shown in Fig. 2 [the PVDC film used in this study provided a superior oxygen barrier compared with the chlorine-free Saran film distributed by S. C. Johnson & Son (Racine, WI)]. The transparent layer of fluorescent ink under the film provided visualization of the excitation scanning spot during experiments. Fluorescence from the light spot was visible to the video camera, but was blocked from signal acquisition by the emission filter and by amplifier gating.

Fig. 2.

Mesentery preparation used for Po2 measurements. The loop of intestine was placed around the thermostatic pedestal, and the mesentery was covered with a circular piece of Saran film. The entire intestinal loop was then covered with a larger sheet of Saran film, stretched, and fixed to a neoprene ring with pushpins. The Saran bag was attached to the objective lens and pressurized to reduce the thickness of the fluid layer between the film and mesentery. A thin water layer was added to the bag for use with the water immersion objective.

A loop of small intestine with mesentery was exteriorized via a cauterized midline abdominal incision and placed around the pedestal window on the neoprene ring (see Fig. 2). The phosphorescence probe was then applied to the mesentery topically by saturating a small circle of filter paper with probe (6.7 mg/ml) and placing the paper in contact with the mesentery under the Saran film covering the tissue. This technique is based on reports of high albumin conductivity in the mesentery (8, 30) and the known diffusion coefficient of albumin in the interstitium of the rat mesentery (Da = 0.38 × 10−7 cm2/s) (15). The estimated time (Ta) for one-dimensional diffusion of albumin-bound probe from the surface into the mesentery tissue to attain 90% of the concentration at the base of the tissue sheet is given by Ta = 1.03h2/Da (9), which is equal to 15.5 min, where the thickness of the mesentery h is 58.6 × 10−4 cm (49). This method of administration provides uniform distribution of the probe throughout the interstitial space and results in the absence of a phosphorescence signal arising from the intravascular compartment.

After 30 min of exposure, the filter paper was removed and the excess probe was washed away with a phosphate-buffered saline solution (PBS). A Saran Wrap film circle of the approximate size of the pedestal window was placed on the mesentery, carefully ensuring that no air bubbles were trapped between the mesentery and the film. To protect the intestine from desiccation and immobilize the segment of intestine, a larger sheet of Saran was placed over the initial piece and its edges were wrapped around the intestine and attached to the neoprene ring with several pushpins (see Fig. 2). The amplitude of the phosphorescence signal from the mesentery, under such circumstances, was high enough for Po2 measurements, while lower than in the applied probe solution.

The design of an objective-mounted pressure bag is shown in Fig. 2. A silicone rubber ring, with Tygon plastic tubing threaded through it, was fixed onto the objective barrel. A bag made of Saran film was fixed to the ring with a second ring, and electrical tape was wrapped around the outer ring to ensure that there was no outflow of air from the bag. The area of contact between the bag and mesentery preparation was ∼20 mm in diameter, while the diameter of the thermostabilized sapphire window of the organ pedestal was 37 mm (16). The bag was filled with a thin layer of water for immersion of the objective lens. A bag pressure of 8–12 mmHg was supplied from a pump connected to a pressure gauge. The time necessary to equilibrate the temperature of the water in the bag and tissue (37°C) was 16 min, so the bag was warmed before the beginning of measurements and remained in permanent contact with the tissue and heated pedestal during the entire period of experimentation. A small amount of transparent lubricant was added between the Saran bag and the Saran film covering the tissue.

The probe application and the phosphorescence registration were carried out in a dark room, and other weak background light sources were shielded to reduce noise caused by stray light. The imaging and positioning of the line of the laser scan in the microvasculature were made under transillumination through an OG-570 orange glass filter (Edmund Optics) to exclude unnecessary excitation of the phosphorescent probe.

Measurements of Po2 profiles were conducted in the vicinity of 84 mesenteric arterioles, 7–39 μm in diameter, located in the transparent connective tissue windows away from fat deposits. Measurement sites were randomly selected and equidistant from branching points in straight segments of arterioles. The direction of the excitation scan was oriented parallel to the axis of the vessel. All distances for the Po2 profiles were measured from the inner wall of the vessel image perpendicular to the axis of the arteriole. The minimal distance to the nearest vessel in the direction of the Po2 profile measurements was 120 μm, which allowed measurements at distances of 5, 15, 30, and 60 μm from the inner vessel wall, leaving at least 60 μm of avascular tissue after the last site. Additional measurements were made at points 120 and 180 μm from the inner wall, if at least a 60-μm length of avascular tissue after the last site was available. The interstitial fluid Po2 measurements inside the image of the arteriole were performed at the −5 μm site (inward from a blood/wall line) or at the center line of vessels below 18 μm in diameter. Data collection for one Po2 measurement at a given site took ∼9 s, and switching from one site to another took ∼10 s; consequently, the total time required to cycle through a set of Po2 profile measurements (5–7 locations) was 2–3 min.

Phosphorescent probe preparation.

The oxygen-sensitive phosphorescent probe palladium meso-tetra-(4-carboxyphenyl)-porphyrin (Pd-MTCPP), obtained from Oxygen Enterprises (Philadelphia, PA), was conjugated with bovine serum albumin (BSA) as described by Vanderkooi et al. (43). Twenty milliliters of 1 mM BSA (fraction V) was dissolved in deionized water, and 200 mg of Pd-MTCPP, dissolved in 4 ml of DMSO, was added drop by drop. The pH was maintained at 7.4 with Tris base (BSA, DMSO, and Tris obtained from Sigma, St. Louis, MO). The probe solution was then dialyzed (Spectra/Por-4 tubing, Spectrum Laboratories, Campton, CA) in PBS with the isooncotic additive polyvinylpyrrolidone (PVP, mol mass 55 kDa; Aldrich, Milwaukee, WI) at 4°C. After dialysis the solution was sterilized by filtration using a 0.2-μm filter unit (Nalgene, Rochester, NY). Aliquots (0.5 ml) of the probe solution with a final concentration of 6.7 mg/ml were placed in plastic microtubes, quickly frozen, and stored at −20°C. For use, the solution was thawed at room temperature and applied to the tissue as stated above.


Oxygen consumption by the method was tested in the mesentery in avascular areas (3–5 mm away from radial vascular bundles) 20 min after the animal was euthanized. To exclude oxygen consumption by the tissue, it was poisoned with 0.1 M NaCN topically applied under the Saran film. After a 15- to 20-min exposure period, the mesentery was briefly opened to the air and then resealed. The Saran air bag was then inflated to a pressure of 25 mmHg. Phosphorescence signals from 1,100 consecutive laser pulses, scanning a 57 × 5-μm strip of the mesentery, were recorded for each of five different sites. The Po2 values were calculated for 10 groups of 110 curves to determine the total trend in Po2. The trend was used to determine the existence of any accumulated oxygen consumption by the method during the period of tissue measurement (9.1 s).

Phosphorescence decay analysis.

The averaged phosphorescence decay curves with a high signal-to-noise ratio were used for Po2 calculations. A nonlinear fitting of the curves was based on the rectangular Po2 distribution model (17): Math where t (μs) is the time from the beginning of phosphorescence decay, I(t) (V) is the magnitude of the phosphorescence signal, I(0) (V) is the amplitude of the phosphorescence signal at t = 0, M (mmHg) is the mean Po2, δ (mmHg) is the half-width of the Po2 distribution, T (μs) is the lifetime of the fast postexcitation transient, A (V) is the amplitude of the fast postexcitation transient, and B (V) is the baseline offset. Phosphorescence decay in the absence of oxygen (K0) = 18.3 × 10−4 μs−1 and quenching constant (Kq) = 3.06 × 10−4 μs−1mmHg−1 were determined by Zheng et al. (50) and were close to values obtained by other workers (28, 36, 37, 48). The offset B consisted of a 10-mV amplifier offset and a small addition of the PMT dark current. In a set of 100 randomly selected decay curves B = 16 ± 1 mV, the fast initial transient lifetime T = 10.6 ± 0.5 μs, and its amplitude was A = 9.6 ± 0.5% of the maximum signal amplitude, I(0).

Calculation of oxygen consumption rate.

The rate of oxygen consumption by the mesentery was calculated from a one-dimensional steady-state diffusion model with zero-order oxygen consumption kinetics, as described in detail by Tsai et al. (41) and Yaegashi et al. (49). The Po2 profile values measured perpendicular to the arteriolar axis were fitted with the parabola (49): Math where Po2(0) and Po2(x) are the oxygen tensions at the arteriolar blood/wall interface (x = 0) and at a distance x from it. The coefficient C1 = V̇o2/2α·Do2, obtained from the parabolic fitting, was used to calculate V̇o2 for known coefficients of oxygen solubility (α) and diffusion (Do2) in connective tissue. Values of α and Do2 were taken from the work of Yaegashi et al. (49) and Tsai et al. (41) for comparison of our data with their reports.


Statistical calculations and linear and nonlinear fitting procedures were made with the Origin 7.0 program (OriginLab). All data presented in Table 2 are means ± SE (N = number of measurements), and differences of two means were analyzed with an unpaired t-test. Statistical significance was assigned at the P = 0.05 level.

View this table:
Table 2.

Systemic parameters of experimental animals


The systemic parameters of the experimental animals are presented in Table 2. The mean arterial blood pressure was 97 mmHg, and arterial Po2 was 88 mmHg. The Po2 in peritoneal fluid was 56 mmHg, indicating the absence of hypoxia in the mesenteric environment.

The PQM measurements of Po2 in the centers of avascular mesenteric windows showed that, even at a distance of 4–5 mm from the vascular bundles, oxygen was found in substantial amounts: Po2 = 9.1 ± 2.6 mmHg (N = 10). The Po2 at these sites measured 20 min after euthanasia was reduced because of respiration by the surviving tissue to 1.5 ± 0.3 mmHg (N = 6), showing that the 9-mmHg level was supplied mainly by mesenteric sources of oxygen and not by contamination from the air. The continuous Po2 recording for mesenteric tissue, in which cellular respiration was eliminated with topical cyanide treatment (total of 1,100 light pulses at 11 Hz in 5 series), did not reveal any negative trends in the Po2 level, indicating the absence of accumulated oxygen consumption by the scanning PQM technique. A small positive trend of 0.13 ± 0.02 mmHg/s (N = 5) was found in these tests (see Fig. 3), presumably caused by oxygen inflow to the mesentery due to its imperfect insulation from the atmosphere. The contribution of that inflow to the calculated V̇o2 is estimated below for different values of oxygen solubility in the mesentery.

Fig. 3.

Example of continuous Po2 monitoring in an avascular region of the mesentery treated with cyanide for 1,100 light pulses at a flash rate of 11 Hz. The possibility of accumulated oxygen consumption by the scanning phosphorescence quenching microscopy (PQM) technique was tested, and there was no detectable decrease of Po2 found for a period 10 times longer than the experimental measuring time. A small upward trend of 0.008 mmHg/pulse is observed in these values.

The average data for 84 Po2 profiles are presented in Fig. 4. No difference was found between Po2 values measured inside and just beside the image of arterioles in the total set of profiles. Only a segment of the profile at distances from 5 to 60 μm matches the theoretical profile, allowing application of a one-dimensional diffusion and consumption model. An increasing negative slope appeared only at 120–180 μm from the blood/wall interface. At a distance of >100 μm from arterioles, the Po2 in the mesentery was still much higher than the zero value found by Tsai et al. (41).

Fig. 4.

Average (N = 84) Po2 profile in mesenteric connective tissue around arterioles. The distances were measured from the blood/wall interface line on the image of an arteriole. Negative distance denotes the position of the scanning line inside the image of the vessel, and positive distance denotes the position of measurements in the surrounding tissue.

The set of profiles was sorted according to arteriolar diameter and divided into two groups, large and small arterioles, of 42 profiles each. Averaged profiles for these groups with different mean diameters (27.2 ± 1.0 and 13.2 ± 0.5 μm) are not significantly different, as shown in Fig. 5. Po2 is the same inside and just outside the image of arterioles, and no significant longitudinal Po2 gradient in arterioles was revealed from the comparison between these values for small and large arterioles. The Po2 profile gradient out to a distance of 60 μm was not significantly different between the large- and small-arteriole groups.

Fig. 5.

Average Po2 profiles in connective tissue around two groups of arterioles [diameter 27.2 ± 1.0 (▪) and 13.2 ± 0.5 (•) μm], of 42 profiles each. Po2 values are the same inside and just outside the image of the arterioles. The Po2 profile gradient out to a distance of 60 μm was not significantly different between the groups of large and small arterioles.

The rate of oxygen consumption by the mesentery was calculated based on a segment of the averaged Po2 profile from a distance of 5–60 μm away from the blood/wall interface (Fig. 6). Since the mesentery thickness does not change significantly over that distance, the one-dimensional diffusion model discussed above was used for this determination. The coefficient C1 = 89,382 mmHg/cm2 was ascertained from nonlinear fitting by the parabolic function as shown in Fig. 6. The values of V̇o2 from the current study are presented in Table 3. They were calculated with values of α and Do2 from Yaegashi et al. (49) and Tsai et al. (41). The oxygen contamination determined in tissue treated with cyanide can be converted into the same units as V̇o2 in order to estimate its contribution (negative offset) to the calculated V̇o2 (see Table 3). That offset of V̇o2 in the mesentery is estimated to be within the range of 4–7%, depending on the values of α and Do2 used for calculations. The corrected V̇o2 does not change the large difference between our results and those reported by some previous investigators (Table 3).

Fig. 6.

The least-squares fit to a parabola for the segment of the average (N = 84) Po2 profile between a distance 5–60 μm away from the blood/wall interface.

View this table:
Table 3.

Oxygen consumption in mesentery


Values of Po2 measured at sites inside the image of arterioles (Fig. 4) were in good agreement with results reported by Yaegashi et al. (49) and Tsai et al. (41) obtained in the same locations. In the experiments of Tsai et al. the probe was predominantly confined to the vessel lumen, and results were interpreted as intravascular Po2. Yaegashi et al. measured Po2 based on oxygen-sensing silica gel beads separated from the mesentery by a fluid layer of 38 μm and a <5-μm silicone rubber membrane, so their results reflected extramesenteric Po2 at a remote plane. In our experiments, the probe was distributed only in the interstitial fluid and, therefore, the signal was localized to the extravascular tissue compartment.

Although our Po2 measurements were done in the mesentery interstitium, we can infer (based on the knowledge of low V̇o2 of the connective tissue) that measurements above the center line of arterioles are close to values of intravascular Po2. In our experiments, Po2 values above arterioles were found to be in agreement with intravascular Po2 values reported by other workers. However, the Po2 obtained 5 μm away from the arteriolar wall (Fig. 4) in the present study differs from the reports of Tsai et al. (41), which found a Po2 of ∼30 mmHg on the external side of arteriolar walls.

Our Po2 measurements in peritoneal fluid (56 ± 7 mmHg) corroborate the data obtained in cats and support the conclusion that the mesentery is normally exposed to a high-oxygen environment (25). Periarteriolar Po2 values measured by microelectrodes in tissues with high metabolic rates also have been reported to be much higher than 30 mmHg: first-order arterioles of rat spinotrapezius muscle at rest: 50 mmHg (26); above retinal cat arterioles: 55.2 mmHg (5); rat pial arterioles: 68–76 mmHg (44). The low periarteriolar Po2 found by Tsai et al. (41) led them to conclude that a large Po2 gradient exists across arteriolar walls because of a very high oxygen consumption of the vessel wall. These data contradict previous direct measurements of wall Po2 gradients with microelectrodes that showed relatively small differences: 1.4 mmHg intra-/extravascular difference in hamster cheek pouch arterioles (11) and an ∼1 mmHg/μm transmural Po2 gradient in cat pial vessels (12). Furthermore, recent simultaneous PQM measurements of blood and tissue Po2 with two different oxygen probes revealed only a slight Po2 difference (<1.5 mmHg) between intravascular and interstitial compartments in skeletal muscle (47). The results of the present study, which found a high Po2 measured directly adjacent to arterioles, do not support the results reported by Tsai et al. (41), nor do the present study results verify the existence of a significant Po2 gradient across the arteriolar wall.

The comparison of Po2 around large (27.2 ± 1.0 μm) and small (13.2 ± 0.5 μm) arterioles showed no significant difference (Fig. 5), indicating that longitudinal gradients in the arterioles of connective tissue are small. Yaegashi et al. (49) also reported no significant longitudinal Po2 gradients in the mesenteric arterioles under normal conditions. Tsai et al. (41) expressed the longitudinal oxygen losses in the same type of vessels in terms of an oxygen saturation drop (24%/mm at an intravascular Po2 of 43.3 mmHg) that approximately corresponds to the same number expressed in millimeters of mercury. This is much higher than the results reported by the same group (3.0–3.8 mmHg/mm) for the hamster dorsal skin chamber (6, 40).

Profiles of periarteriolar Po2 have been obtained in previous years in tissues with various metabolic rates and oxygen microelectrodes. Buerk et al. (5) found in the eye vitreous that Po2 remains above 45 mmHg at a distance of 500 μm from arterioles. Measurements of transmural Po2 profiles in the rabbit aortic wall were higher than 35 mmHg at a distance of 200 μm from the blood/wall interface (4). Profiles obtained by Yaegashi et al. (49) in the mesentery also showed Po2 levels above 20–30 mmHg at a distance of 200 μm. Conversely, the Po2 measured in the mesentery by Tsai et al. (41) exhibited a decrease in Po2 to near zero at a distance of 50–60 μm from the blood/wall boundary. The steepness of the Po2 profile in mesenteric connective tissue in their work significantly exceeded the decline of Po2 at the same distance from arterioles in intensively consuming retinal tissue (5) and brain cortex (44).

Our results demonstrate a relatively flat Po2 profile from 5- to 60-μm distance, with an increasing slope only at 120–180 μm from the blood/wall interface. A possible explanation for this finding is that Po2 profiles in our experiments were principally measured from arteriole to avascular tissue regions where the thickness of connective tissue was reduced below 20 μm (2, 15), while the two layers of mesothelial cells, the main consumers of oxygen in this tissue, remained the same. Thus, as the connective tissue becomes thinner, the volume-specific oxygen consumption increases. This may create higher specific oxygen consumption in thin regions of the mesentery. A second factor distorting the profile at 120–180 μm is a deviation from the model assumption that the direction of oxygen diffusion is always transverse to the axis of the vessel. These factors limit the radial distance from arterioles that can be used for Po2 profile analysis with the homogeneous one-dimensional model of oxygen diffusion, combined with zero-order consumption kinetics. According to the anatomic data reported previously (2, 15), application of this model may be limited to a distance of ∼100 μm from a vessel. In the present study, this interval contains four experimental points (5–60 μm) that have a good parabolic fit and can be used to calculate the rate of oxygen consumption by the mesentery for comparison with previously reported values of V̇o2 in different tissues (see Fig. 6).

With values of the parameters α and Do2 employed by Yaegashi et al. (49) and Tsai et al. (41) V̇o2 for the rat mesentery was calculated to be 55.8 and 65.0 nl O2·cm−3·s−1, respectively. Correction for the oxygen contamination offset from the air does not increase V̇o2 substantially (60–68 nl O2·cm−3·s−1). Compared with other values of V̇o2 presented in Table 4, our value is much lower than that reported by Tsai et al. (41) and quite close to the V̇o2 of the renal capsule and aortic valve cusp, which consist mostly of loose connective tissue. Our calculated mesenteric V̇o2 is also much lower than that found in rat skeletal muscles but significantly higher than results obtained by Yaegashi et al. (49) for the rat mesentery. Discrepancies in V̇o2 between our data and results reported by other workers seem likely to be due to differences in instruments, methods, and experimental conditions that should be scrutinized further.

View this table:
Table 4.

o2 in different tissues

The mapping of the spatial Po2 distribution by Yaegashi et al. (49) employed a 5-μm-thick silicone rubber film containing 3-μm silica gel beads containing a luminescence oxygen probe. The method was designed to be resistant to photoactivated oxygen consumption, so the singlet oxygen (1O2) generated by the photosensitizer (probe) remained trapped inside the beads and deactivated to the ground state because of the short lifetime of 1O2 (4 μs in water, 99.9% quenched within 250-nm distance; Refs. 27, 38, 46). In the work of Yaegashi et al. (49), the most probable source of the discrepancy with our results was the separation of the probe membrane and the mesentery by an estimated 38-μm fluid layer [approximately the equivalent thickness of rat mesentery in avascular, adipocyte-free regions (17.4–58.6 μm) (2, 15, 49)], which could cause a substantial bypass for oxygen diffusion ultimately resulting in a three times higher rate of diffusion (Do2 in water at 37°C is 3.3 × 10−5 cm2/s; Refs. 1, 49). The fluid layer above the mesentery could also be agitated by pulsation of the mesenteric arteries, adding to the convective transfer of oxygen in the horizontal plane. Both mechanisms of extramesenteric oxygen transfer could have acted to “flatten” the Po2 profiles, reducing the value of the calculated V̇o2 by seven- to eightfold compared with our data. In our experiments, the fluid layer bypass was minimized by the application of a low external pressure (i.e., a pressurized Saran bag) that provided a tight contact between the overlying Saran Wrap film and the mesenteric tissue without affecting blood flow. Furthermore, localization of the interstitial signal in our study was ensured by the direct application of probe to the mesentery rather than the use of an oxygen-sensitive membrane remotely located above the mesentery.

The low interstitial Po2 in the rat mesentery measured by Tsai et al. (41) resulted in the conclusion that the transmural Po2 drop was created by the very high V̇o2 of the arteriolar wall. These results can be accounted for by the application of their PQM method to flowing and motionless media.

Energy transfer from the photoexcited probe to dissolved oxygen in water creates highly reactive oxygen species, mainly 1O2 (13, 14, 29, 43, 46), which has a lifetime of ∼4 μs (46). In blood or tissue 1O2 is consumed by the oxidation of organic molecules (13, 14, 29, 38), so the lifetime of 1O2 in blood plasma is only 1 μs (23). The inventors of PQM, Vanderkooi et al. (43), found that oxygen consumption by the macroscopic application of the PQM method was 60 nM/min (∼25 × 10−6 mmHg/flash). The microscopic application of the PQM method deals with a very small volume of blood or tissue, so that the energy of the phosphorescence signal is very low. To improve the signal quality, the probe concentration was increased from 1 μM to 1 mM (39, 43) and the excitation light from a flash lamp source was condensed by an objective lens. Both efforts to amplify the signal resulted in increased oxygen consumption by the PQM method, an inevitable factor affecting microscopic Po2 measurements. The oxygen consumption in plasma caused by the flash of a conventional xenon lamp, delivered by epiillumination optics to the microscopic field, was determined to result in a Po2 decrement of ∼0.3 mmHg per flash (31).

In the study of Tsai et al. (41), the oxygen consumption artifact was discounted on the basis of in vitro testing of consumption in a solution of oxygen probe with a concentration similar to that found in the interstitial fluid. The solution of oxygen probe was sealed in a 75-mm (L)-long glass tube, which was masked, leaving 0.5 mm (l) of tube length exposed to flash illumination. The fluid remained motionless during a 45-min (t) excitation period and then was immediately mixed after 81,000 flashes (N) (at a 30-Hz flash rate). The Po2 values measured in the tube before and after the 45-min period of excitation were 37 and 31 mmHg (ΔPo2 = −6 mmHg), respectively. The Po2 change per flash was calculated as δ = [ΔPo2·(L/l)]/N = −6 mmHg·(75 mm/0.5 mm)/81,000 flashes = −0.01 mmHg/flash (41). The paradox is that the numerator [ΔPo2·(L/l) = −6 mmHg·(75 mm/0.5 mm) = −900 mmHg] demonstrates that for 81,000 light flashes Po2 within the exposed tube segment was reduced to a physically impossible level. Even for the reported δ = −0.01 mmHg/flash, all oxygen within the illuminated segment would be depleted after 3,700 flashes (only 2 min). In the remaining 43 min, oxygen consumption within the exposed segment of the tube would be very low (proportional to the Po2) and limited by the diffusion of oxygen from the masked part of the tube. The mean square displacement for oxygen molecules during a period of illumination (t = 45 min) is Ld = 2Do2·t = 3.7 mm, where Do2 = 2.5 × 10−5 cm2/s, the oxygen diffusion coefficient in water at 20°C (1). The length of the fluid column in the tube involved in oxygen depletion by illumination in this test was ∼2 × 3.7 + 0.5 = 7.45 mm or 10% of the total volume. Therefore, the actual oxygen consumption per flash in the test was at least 10 times higher than was reported. Also, in consideration of the reported factor of 2 for initial consumption at Po2 = 37 mmHg (41), the resultant consumption per flash would be δ ≈ 0.2 mmHg/flash.

The oxygen depletion in the tissue that occurs via this methodology cannot be compensated for by oxygen diffusion from the tissue surrounding the excitation spot because of the very wide excitation area and the high flash rate (140 μm, 30 Hz). Limiting relationships between the size of the excitation area and maximal flash rate F (necessary to replenish 90% of the reduced Po2 in the center of a round spot for the time between 2 flashes) can be obtained from a simplified model of diffusion into a cylinder (9), F < Do2/0.54·R2 (where R is the radius of the excitation area and Do2 is the diffusion coefficient for oxygen). Thus, in connective tissue at 37°C where Do2 = 1.04 × 10−5 cm2/s (49), a wide excitation area (R = 70 μm; Refs. 24, 41) limits the maximal flash rate to F = 0.4 Hz, while the maximal flash rate for a small excitation spot (R = 2.5 μm in our instrument) is F = 300 Hz. In our study, a much lower flash rate of 11 Hz was used in order to ensure sufficient time for the recovery of Po2 disturbed by each light pulse.

Differences in the parameters R and F defined by the PQM instruments used must be the main source of discrepancy between results obtained in the same tissue. The mesentery V̇o2 = 240 nl O2·cm−3·s−1 obtained by Tsai et al. (41) is four times higher than the calculated V̇o2 in the present work. Reported oxygen flux [3.6 × 10−5 ml O2·cm−2·s−1; Ref. 7] from mesenteric arterioles with arrested flow measured with the large-spot/high-frequency method (R = 140 μm, F = 10 Hz) was 45-fold higher than the value [8 × 10−7 ml O2/(cm2·s)] obtained with a small-spot/low-frequency instrument (R = half of the arteriolar diameter, F = 1 Hz) (32). Thus the PQM technique under such conditions (wide excitation area/high flash rate) creates a significant measurement artifact due to excessive oxygen consumption and, therefore, contraindicates the usage of such methods for Po2 measurements in stationary tissue, blood, and lymph.

Furthermore, taking into account the oxygen consumption artifact neglected by Tsai et al. (41), the accumulated oxygen depletion for 90 flashes at 0.2 mmHg/flash in the stationary tissue is as high as 18 mmHg. This value is identical to their results reported for the transmural Po2 drop. Since the summation of oxygen depletion did not occur in flowing blood, the apparent “intra-perivascular differences” must have been due to oxygen consumption by the method itself.

Minimizing the size of the excitation spot and the flash rate can be an effective strategy to reduce the artifact only for a short train of flashes. Equally, when the pulse train is long, the tissue surrounding the excitation spot also becomes depleted of oxygen. In this case, oxygen diffusion cannot effectively compensate for the oxygen losses, either. To create measurement conditions in a stationary tissue similar to those in flowing fluids, we applied a method whereby the excitation spot location was shifted between flashes (i.e., tissue scanned in discrete steps). This prevents excitation of the same tissue volume before oxygen diffusion back into the excitation region. The combination of a small excitation area, a low flash frequency, and scanning the tissue with the excitation spot employed in our new method allows for Po2 measurements in stationary tissue or blood without a significant accumulated depletion of oxygen. The in situ test for oxygen consumption by the method did not show a Po2 decrease during a period 10 times longer than the time interval of Po2 measurements in this experiment. A small inflow of oxygen, presumably from ambient air, was detected instead. The contribution of this oxygen contamination may be insignificant for thick tissues with high metabolic rates, but it was taken into account in the case of the mesentery.

In conclusion, there is a principal difference between measuring the concentration of dissolved oxygen in flowing and stationary media with PQM similar to issues that have been raised in the past regarding the oxygen microelectrode technique. The accumulated consumption of oxygen by the method employed could be the main source of error affecting Po2 measurements in the tissue. We conclude that a PQM instrument having a wide excitation area and a high flash rate is not suitable for Po2 measurements in stationary tissues. As such, we have designed a novel PQM instrument that combines a small excitation area, a low flash rate, and a scanning excitation spot in order to minimize the artifact of oxygen consumption by the method to a level of a single flash consumption artifact (∼0.3 mmHg). The direct application of the probe to the tissue and gently pressurizing the barrier film against the tissue ensured the localization of Po2 measurements and eliminated the transport bypass in the fluid between the tissue and gas barrier film. We have found that periarteriolar Po2 values above and beside the mesenteric arterioles are the same and as high as reported for intra-arteriolar Po2. Oxygen consumption in the mesentery was found to be about eight times higher than measured by Yaegashi et al. (49) and four times lower than that reported by Tsai et al. (41). After the correction for oxygen inflow from the air, the only possible source of error in our experiments could be the single flash consumption artifact that cannot be accumulated. In this case, the hypothetical correction would lessen the values of V̇o2 obtained in the mesentery. Thus we conclude that 60–68 nl O2·cm−3·s−1 is the upper limit for the oxygen consumption by connective tissue in the rat mesentery.


This work was supported by Grants HL-18292 and HL-79087 from the National Heart, Lung, and Blood Institute and Grant AHA-0655449U from the American Heart Association, Mid-Atlantic Affiliate.


We thank Dr. Helena Carvalho for help with the polarographic Po2 measurements in peritoneal fluid in rats.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract