Endotoxin-induced changes in intrarenal pO2, measured by in vivo electron paramagnetic resonance oximetry and magnetic resonance imaging

Endotoxin-induced changes in intrarenal pO2, measured by in vivo electron paramagnetic resonance oximetry and magnetic resonance imaging

Free Radical Biology & Medicine, Vol. 21, No. 1, pp. 25-34, 1996 Copyright © 1996 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5...

848KB Sizes 0 Downloads 35 Views

Free Radical Biology & Medicine, Vol. 21, No. 1, pp. 25-34, 1996 Copyright © 1996 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/96 $15.00 + .00 ELSEVIER

SSDI 0891-5849(95)02221-X

Original Contribution ENDOTOXIN-INDUCED CHANGES IN INTRARENAL PO2, MEASURED BY IN VIVO ELECTRON PARAMAGNETIC RESONANCE OXIMETRY AND MAGNETIC RESONANCE IMAGING PHILIP E. JAMES, * GORAN BACIC, * OLEG Y. GRINBERG, * FUMINORI GODA, * JEFFREY F. DUNN, * SIMON K. JACKSON, t and HAROLD M . SWARTZ * • EPR Center and NMR Center, Radiology Department, Dartmouth Medical School, Hanover, NH 03755-3863, USA •Medical Microbiology, University of Wales CoUege of Medicine, Cardiff, Wales, CF4 4XN, UK

(Received 19 June 1995; Revised 25 September 1995; Accepted 6 November 1995)

Abstract--Electron Paramagnetic Resonance (EPR) oximetry was used to measure tissue oxygen tension (pO:-partial pressure of oxygen) simultaneously in the kidney cortex and outer medulla in vivo in mice. pO2 in the cortex region was higher compared to that in the outer medulla. An intravenous injection of endotoxin resulted in a sharp drop in pO2 in the cortex and an increase in the medulla region, resulting in a transient period of equal pO2 in both regions. In control kidneys, functional Magnetic Resonance (MR) images showed the cortex region to have high signal intensity (T2 *-weighted images), indicating that this region was well supplied with oxygenated hemoglobin, whereas the outer medulla showed low signal intensity. After administration of endotoxin, we observed an immediate increase in signal intensity in the outer medulla region, reflecting an increased level of oxygenated blood in this region. Pretreatment of mice with N°-monomethyl-L-arginine prevented both the changes in tissue pO2 and distribution of oxygenated hemoglobin, suggesting that localized production of nitric oxide has a critical role to play in renal medullary hemodynamics. In combining in vivo EPR with MR images of kidneys, we demonstrate the usefulness of these techniques for monitoring renal pO2 and changes in the distribution of oxygen. Keywords---Renal tissue oxygen, Renal medulla, Free radicals

developed, 5 which have high sensitivity to oxygen, resistance to chemical reactions, and are inert within biological systems. For this study we selected crystals of lithium phthalocyanine (LiPc) as the oxygensensitive probe. Their small size enables highly localized, multiple site measurements of pO2 in tissues in vivo to be made) MR imaging has been extensively used in the assessment of a wide variety of renal pathologies, including ischemia. 9-1~ Frequently, it has been emphasized that alterations of cortico:medullary contrast on MR images are a sensitive, but not a very specific indicator of renal damage or disease. Recently, however, it has been demonstrated that intrarenal modulations in blood oxygenation can be noninvasively monitored by measuring cortico:medullary contrast with MR techniques that are sensitive to oxygenation changes (oxy- vs. deoxy-hemoglobin) J 2'~3 The method is based on the fact that changes in blood oxygenation level can serve

INTRODUCTION

The pO2 in tissues reflects the balance between oxygen delivery and consumption of oxygen in viable cells and tissues. 1'2 However, direct measurements of pO2 in renal tissue are sparse and inconsistent. Moreover, the methodology employed has involved invasive techniques (such as tonometry), which could produce artefactual r e s u l t s . 3'4 Recent advances in the development of Electron Paramagnetic Resonance (EPR; or completely equivalently Electron Spin Resonance, ESR) oximetry 5-7 provide a more direct measure of renal pO2 in vivo. EPR oximetry utilizes the effect of oxygen on the spectra of EPR active (paramagnetic) substances. Recently, a new class of stable paramagnetic particles has been Address correspondence to: Harold M. Swartz, EPR Center and NMR Center, Radiology Department HB 7252, Strasenburgh Hall, Rm 308, Dartmouth Medical School, Hanover, NH 03755-3863. 25

26

P.E. JAMESet al.

as an endogenous contrast agent by producing characteristic changes in the MR signal intensity. 14'15 In this study we assessed the usefulness of EPR oximetry and MR imaging for monitoring renal tissue pO2. Changes in the supply of oxygen to the kidney and the distribution within were induced by systemic injection of endotoxin (lipopolysaccharide, LPS) to mimic the septic state. MATERIALS AND METHODS

Materials

Lithium phthalocyanine (LiPc) was a generous gift of Dr. M. Moussavi (Centre D'Etudes Nucleaires De Grenoble, France) and was synthesized by a reported method. 16The EPR line width of LiPc is a linear function of pO2 and is independent of local metabolic processes, the presence of other paramagnetic species, and pH. 6 It has also been showm that LiPc crystals equilibrate with local tissue pO2 within 30 sec, and the linewidth response to changes in pO2 are stable for at least 72 h in tissue. The small size (approx. dimensions 0.1 x 0.05 x 0.05 mm) and high unpaired spin density enable accurate measurements of local pO2 to be made using a single implanted crystal. All other chemicals were obtained from the Sigma Chemical Company (St. Louis, MO) unless otherwise stated; lipopolysaccharide (isolated from E. coli O11 l:B4; Sigma) was used throughout as the source of endotoxin. Animal preparation

Anesthesia was induced in male mice (Balb/C; approximate weight 25 g) by intraperitoneal injection of a mixture of Xylazine/Ketamine (20 mg and 100 mg/ ml, respectively, in saline; 0.5 ml/30 g b.wt.) and supplemented by small doses thereafter when needed. The left kidney was exposed retroperitoneally through a flank incision and the LiPc crystals were implanted. This was achieved by placing the crystal inside the end of a 26 gauge needle, which was fitted with a length of wire (0.2 mm dia.). The needle was inserted to a depth of 1 mm (cortex) and 2.5 mm (outer medulla) as indicated in Fig. 1. The wire was used to displace the crystal from within the needle to implant the crystal at the desired location. At the end of each experiment the kidney was removed from the animal and dissected to confirm the position of the crystals. The incision was covered with clear plastic film and the animal placed between the poles of the 1.1 GHz EPR spectrometer magnet. Body temperature was measured using a rectal probe and maintained at 37°C using a heating lamp. The animal was allowed to reach equilibrium for 30 min (so as to reach thermal

equilibrium after placement in the EPR cavity, and to minimize the effects of postsurgical trauma), at which point an additional dose of anesthesia (halfdose) was given intraperitoneally [ no further anesthesia was required for the course of the experiment (60 min)]. LPS (200/~g/150 #1) in sterile 0.9% saline was injected in a bolus dose via the tail vein. Respiratory rate (approximate 104/min) was monitored throughout using a small pressure-sensitive transducer (Abbott Critical Care Systems, North Chicago) linked to a Hewlett-Packard monitor. This did not change significantly (following the initial equilibrium period of anesthesia) for up to 40 min post-LPS. We observed two deaths per five animals treated with endotoxin (they did not recover from anesthesia and died within 3 h post-LPS injection), which is in close agreement with previous studies using similar LPS doses and route of administration in model systems of septic shock. 5'17'18 All animals given LPS showed typical symptoms of endotoxic shock including ruffled fur, hunched back, and diarrhea. EPR oximetry

The extended loop detector ( 10 mm dia.) was placed over the exposed kidney and EPR spectra of LiPc were recorded using a homemade EPR spectrometer equipped with microwave bridge operating at 1.1 GHz. 8,w The number of spin centers within each crystal was 2 x 1015 (based on a crystal size of 0.1 × 0.05 x 0.05 mm, and the number of spins per gram quoted for LiPc in Ref. 6). The sensitivity of our spectrometer operating with the loop detector was 5 × 1013 spins/ Gauss (signal:noise = 1 ). This enabled recording of EPR signals from small crystals of LiPc with a signal:noise ratio of 100 or better. After measuring baseline pOz for 10 min, LPS or saline alone as a control, was injected and spectra were recorded continuously (20 or 40 s per scan) for up to 1 h. The spectrometer settings were as follows: incident microwave power = 10 mW; modulation amplitude = 32 mGauss; scan range = 1 Gauss; time constant = 0.1 s. A linear magnetic field gradient along the main magnetic field (Fig. 1 ) was used to separate the signals originating from crystals in the cortex and outer medulla, thus enabling simultaneous measurement of pOz in these two sites. 8 The gradient was produced by a pair of coils in an anti-Helmholz configuration and its magnitude was calibrated by measuring the peak-topeak splitting of two capillaries containing LiPc. The magnitude of the magnetic filed gradient was kept at a minimum to avoid distortion of the line width while maintaining adequate separation of the signals arising from two sites. Line widths were measured from first derivative spectra, and converted to pO2 values from

Renal oxygenation after endotoxin Ureter

27

Renal vein and arter ::i~i~i Renal Corte~

iiii

°----

(outer and Inner s~pe} UPc in cortex

UPc I

cortex

II

%

medulla

(a)

[

I

[

424.50

l

424.75

i

425.00

425.25

425.50

Gradient coils

Magnetic

field ( G a u s s )

i000

~-~

750 mHg) 600

800

/I ~ / I •' ~

4.5 ~

500

°e',~ 250

(b)

.

.

.

.

.

.

,

.

50

C.'/~-i

i00

pO 2

l 150

l

,

l

oo

~

200

(mmHg)

Fig. I. A schematic diagram illustrating placement of the animal between the L-band spectrometer magnet. The position of the two crystals of lithium phthalocyanine (LiPc) within the kidney and their position relative to the gradient coils (which separate the two signals) is shown in the magnified view (box). A crystal was inserted to a depth of 1 mm from the kidney surface (cortex) and another at a depth of 2.5 mm (outer medulla). Insert: (a) A typical EPR spectrum obtained from two crystals of LiPC (one placed in the cortex, the other in the outer medulla). Line widths were measured for cortex (high-field peak) and outer medulla (low-field peak) for each time point, and a pO2 value calculated from a standard curve of LiPc line width against varying pO2 (shown in (b) (Reproduced with kind permission from ref. 6.)

a s t a n d a r d c u r v e o f E P R spectra o f L i P c at v a r i o u s pO2 ( s e e Fig. 1 i n s e r t ) .

Inhibition of nitric oxide production A group of five a n i m a l s were pretreated with N % m o n o m e t h y l - L - a r g i n i n e ( N M M A ; S i g m a ) , 2.5 m g / 2 5 g b o d y w e i g h t g i v e n intraperitoneally 30 m i n prior to

surgery (approx. 1 h prior to LPS i n j e c t i o n ) to inhibit nitric oxide synthesis. T h e surgical procedure a n d meas u r e m e n t s of pO2 in vivo were carried out as described above.

Measurement of nitric oxide A t 5, 10, a n d 30 m i n after i n j e c t i o n o f L P S or saline only, b l o o d was r e m o v e d b y retro-orbital b l e e d i n g into

28

P.E. JAMESet al.

heparinized syringes and placed on ice. The proteins of the blood samples were precipitated with 30% zinc sulfate (10 #1 per 100 #1 blood), centrifuged, and filtered with a 0.22 # m filter prior to analysis. The concentration of nitrite ( N O z - ) , a stable metabolite of NO, was determined by a modification of the Griess reaction. 2° Briefly, the filtered samples were mixed with an equal volume of Griess reagent [0.1% N-( 1Naphthyl) ethylenediamine dihydrochloride, 1% sulfanilamide, and 2.5% phosphate in distilled water] and incubated at room temperature for 10 min. The resuiting pink color was read at 543 nm. In a second set of experiments, the total amount of nitrite and nitrate (NO3-) was measured in these same blood samples, because NO is converted predominantly to NO3-in the blood. This was achieved by pretreatment with nitrate reductase prior to the Griess reaction used to assay for NO2-.2° The concentration of NO2- was calibrated from a standard curve of sodium nitrite in serum. The results are from six mice in each group.

Magnetic resonance imaging

MR imaging was performed with a 7.0 T horizontal bore Magnex magnet and a SMIS console. The mice were anesthetized (as for EPR experiments) and placed supine on a foam bed within a 6 cm "birdcage" RF coil. A reference standard (a water filled capillary) was placed retroperitoneally to the rear of the left kidney. After recording control images, the animal was removed from the magnet, LPS was injected, and the animal was immediately returned into the magnet for post LPS imaging. Two imaging sequences were used: a) a sequence that was sensitive to the degree of oxygenation of hemoglobin [so-called T2*-weighted gradient echo sequence ( T R / T E / f l i p angle = 2 0 0 / 1 0 / 4 5 ) ] , and b) a sequence sensitive to changes in water content (T1weighted spin-echo sequence ( T R / T E = 400/16)). Selection of sequences was based on the properties of deoxyhemoglobin. Deoxygenation of oxyhemoglobin changes the paramagnetic properties of the blood resuiting in changes of the local magnetic field. It has been demonstrated that gradient-echo sequences with long TE are extremely sensitive to these magnetic susceptibility-induced changes (T2* effect) in the MR signal intensity in tissues surrounding blood vessels. 12-15 The signal intensity (SI) of a particular tissue on any MRI sequence, however, is a complex function of many other parameters such as intrinsic relaxation times and particularly tissue water content. Therefore, we also used spin-echo sequences, which are substantially less sensitive to susceptibility changes, in an attempt to decouple effects arising from changes of

blood oxygenation from those arising from changes in total water content. Axial images were obtained using the following parameters: slice thickness = 1.5 ram; field-of-view = 4 cm; image matrix = 128 × 128, providing in-plane spatial resolution of 0.31 × 31 mm 2, sufficient to resolve cortex, outer medulla, and inner medulla. The SI of these tissues were measured by placing eliptical regions of interest over the desired region; typically four in the cortex and the outer medulla, and three in the inner medulla. Sis for each region were averaged and normalized using the SI of the reference. RESULTS EPR oximetry

A typical EPR spectrum showing simultaneous acquisition of the signal from both cortex and outer medulla is shown in Fig. 1 (insert a). Separation of the two peaks in the spectrum depends on the distance between the two crystals in the direction of the magnetic field gradient and its magnitude. Knowing the magnitude of the gradient and by measuring the spectral splitting of the two lines, we estimated that the separation of the two LiPc crystals was approximately 2.8 mm. This was in close agreement with microscopic examination and dissection after the experiment (measured at 3 mm). Measurements of line widths in the presence of a magnetic field gradient are always subject to potential errors due to the additional line broadening caused by the gradient. However, it has been shown both theoretically and experimentally that a gradient of 0.07 m T / c m (such as used here) does not cause significant distortion of the EPR line of LiPc. 8 The changes in line width with time observed from the kidney cortex and outer medulla, are shown in Fig. 2. Data from two mice are shown; one treated with LPS, the other given the same volume of pyrogenfree sterile saline. We observed an immediate, sharp decrease in the pO2 of the cortex after injecting LPS into the tail vein, followed by a slow recovery toward the initial values. The pO2 of the outer medulla initially was low (compared to the pO2 measured in the cortex), but after administration of LPS it increased to a level equal to the lowest cortex pO2 value and then returned towards baseline values. The mean differences between the pOe in the cortex and outer medulla for two groups of mice (control and LPS-treated animals (five in each group)) are given in Fig. 3. Baseline values for the cortex were 22.5 ± 1.3 mmHg (mean ___SD), whereas those for the outer medulla were 15.2 ± 1.3 mmHg. The average difference prior to injection of LPS was 7.3 ± 1.1 mmHg, whereas 5 min postinjection this difference was 1.6 ___ 1.4 mmHg. The effect of an IV injection of pyrogen-

Renal oxygenation after endotoxin

:

150

140

j

,

iLK

,

,

TAIL INJECTIO,V



T

v

v

v



29 • 24.0 • v o

Cortex (saline) C o r t e x (LPS) Medulla(saline) Medulla(LPS)



iP

130

20.13

120

19.2

110

17.6

100

16.0

90

14.4

A pa

z

80

5

12.8

10

0

20

30

40

TIME (rain.) Fig. 2. Line width and pO2 values measured from spectra such as in Fig. 1 using simultaneous measurement of two crystals of LiPC in the cortex and outer medulla of mouse kidney. Data from two mice are shown; one given LPS, the other pyrogen-free saline.

taneous measurement of T~, Te, and water content in kidneys 22 show that the parameters increase with cortex{outer medulla{inner medulla. Consequently, the SI of these tissues on T=-weighted images should follow the same order, whereas on T,-weighted images the order should be reversed. In our images, the outer medulla always had the lowest SI, which can be explained if the SI in T,-weighted images at 7 Tesla was also somewhat sensitive to oxygenation of hemoglobin (but to a lesser extent than for T=). These results are

free saline was monitored in a separate group of animals and found to be minimal. This enabled us to rule out the possibility that these changes were due to either a) hemodilution induced hypotension, or b) effects of anesthesia on pO2. MRI

Images obtained using T2-weighted and T,weighted sequences are shown in Fig. 4a and b. Simul-

io[



8

Control mice "(saline)

o Lipopolysaccharide

I

o v

4

o2 ~r

E 2

-2 -4

!

o

t

t

10

20

30

40

TIME (min.)

Fig. 3. Combined data for all experiments expressed as the difference in pO2 between cortex and outer medulla for both control (saline) and LPS treated mice (n = 5 in each group; mean +_ SD for each time point).

30

P.E. JAMESet

al.

a

b

c

cl

Fig. 4. Typical axial MR images of mice kidneys. (a) Trweighted, and (b) T2-weighted images of kidneys prior to LPS. (c, and d) Shows T2*-weighted images taken pre- and postendotoxin, respectively; areas where hemoglobin is in its oxygenated state are seen as regions of high intensity, whereas regions containing deoxy hemoglobin are dark. Note that the reference standard is seen to the left of the animal.

in agreement with our EPR data, indicating lower pO2 (i.e., more deoxygenated blood) in the outer medulla. Typical Tz*-weighted MR images of mouse kidney are shown in Fig. 4c and d. Images obtained prior to injection of LPS showed high cortico:medullary contrast, while after injection, this became less apparent. The changes of relative intensities from both cortex and outer medulla from all T2 *-weighted experiments are shown in Fig. 5 (upper graphs). Immediately after IV injection of LPS, we observed a sharp drop in signal from the cortex region, accompanied by an increase in intensity from the outer medulla. The SI of the cortex returned to control levels after 5 - 1 0 min (n = 5),

while the SI of outer medulla remained higher than the control. The Sis from each region of the kidney obtained using a T I sequence also are shown in Fig. 5 (lower graphs). It can be seen that the SI in all regions decreased immediately after injection of LPS and recovered within 5 min. The SI in the outer medulla continued to increase, indicating that changes observed in this region on T2*-weighted images might have been in part due to increased hydration or that the SI of the outer medulla on T l-weighted images reflects changes in blood oxygenation. The relative increase of SI on T2 *-weighted images (80%) was much higher than on Tl-weighted images (25%), however,

Renal

oxygenation

after endotoxin

31

T2*-weighted sequence i~,,i~

120

2O0

160

140

16O

8O

140

120

120

loo i

i

i

i

i

t

l

~~/i~I ~

100 60

i

0

5

i

i

i

I

i

CORTEX 120

i

80

i

i

i

I

0 5

10 15 2 0 2 5 3 0 3 5

, I I ; 10 15 2 Time/rain.

20

80

;

3

OUTER MEDULLA i

I

INNER 120

t

110

110

100

100

9O

90

80

80

i

I

i

MEDULLA i

i

I

i

i

I

I

I

I

t

70

70 6O

i

5 10 15 20 25 30 3 5

35

I

5

I

I

l

I

I

10 15 2 0 2 5 3 0 3 5

60

i

i

i

~

I

0 5 10 15 20 25 30 35 Time/rain.

I

60 0

5

10 15 2 0 2 5 3 0 3 5

T1-weighted sequence Fig. 5. Signal intensities for cortex, inner, and outer medulla, calculated as percentage change relative to the pre-LPS signal intensities. Data from five mice are shown (results are from both kidneys because there was no difference between right and left kidneys). Data obtained using T2*-weighted (sensitive to hemoglobin oxygenation) and T~-weighted (sensitive to amount of water) sequences are shown in the upper and lower graphs, respectively. For each region of the kidney, a mean MR signal intensity was calculated from the values measured in 4 ROls (each containing 20 to 30 pixels).

indicating that the former images predominantly reflect changes in blood flow and/or oxygenation. INFLUENCE OF NITRIC OXIDE

Mice given NMMA to inhibit nitric oxide synthesis showed no change from baseline pO2 values in both cortex and outer medulla, and no significant difference was observed between mice given the inhibitor and the control group given saline (pre-LPS difference was 7.6 +_ 1.4; post-LPS difference was 7.4 +_ 1.7). Measurement of nitrite concentration in the blood (or measurement of the sum of NO2- and NO3-), as a marker of NO production in mice treated by the same LPS dose, showed no significant changes from baseline control values at incubation times up to 30 min ( 7 - 1 3 nmol/ml). In this model, nitrite increased only 2 - 3 h post-LPS injection ( 5 0 - 1 0 0 nmol/ml, data not shown).

DISCUSSION

In these experiments we have demonstrated the usefulness of in vivo EPR oximetry and MR imaging to measure pO2 and blood oxygenation levels in two regions of the same kidney simultaneously. Measurements in control kidneys (both EPR and MRI) prior to LPS injection show a relatively well-oxygenated

cortex region and less oxygenated outer medulla. This is consistent with a picture of the kidney, in which the outer medulla, although having a high density of capillaries (compared to the inner medulla or cortex regions) and high oxygen demand by urine concentrating cells, 1'38has very slow blood flow, 23 and results in decreased pO2 in the outer medulla, which operates on the verge of hypoxia. Our values compare closely with those quoted by previous workers using oxygen electrodes to measure pO2 in the rat kidney in vivo (approx 10 mmHg). 24'38 We observed significant, time-dependent changes in the pO2 in both cortex and outer medulla after administration of L P S - - a n immediate decrease in the pO2 of the kidney cortex and an initial rise then gradual return to baseline in the pO2 of the outer medulla. The pO2 in both regions recovered almost to control values by 40 min after injection of LPS. It has been suggested that poor tissue perfusion is the cause of multiple organ failure associated with septic shock, 3'25-27 and much evidence suggests that LPS plays a critical role in the development of acute renal failure, a common characteristic of the shock state. 28-3~ The mechanisms underlying this acute renal failure are not completely understood, but renal ischemia resuiting from impaired renal perfusion is often suggested as a cause. 4'32'33

32

P.E. JAMESet

The kidneys receive 25% of the total cardiac output under normal conditions, and possess a countercurrent microvascular arrangement where oxygen diffuses from arteriole to venule. Blood flow in the medulla is derived from efferent arterioles of juxtamedullary glomeruli (in the cortex) and supplies the medulla with approximately 10% of renal blood flow] 8 Therefore, changes in systemic hemodynamics have a profound effect on the supply of blood to the kidney and, consequently, its function. 27 Studies in isolated dog kidneys and in vivo in rats have shown that LPS causes a progressive fall in renal blood flow and glomerular filtration rate, 28 a decrease in sodium reabsorption, 34 and regions of localized tissue hypoxia. ~ Our EPR and MRI results show that changes in renal pO2 occur within 5 min after injection of LPS. We have previously demonstrated in vitro that endotoxin has a delayed toxic effect on cellular oxygen consumption (but required 20 min incubation) 35, and from this conclude that decreased consumption due to direct toxicity of LPS cannot explain the rapid changes in renal pO2 reported here. We think it more likely that these LPS-induced changes in pO2 are the result of changes in renal blood supply (either to the kidney or redistribution therein). It is also possible that decreased medullary 02 requirements from reduced GFR and solute delivery may give rise to improved medullary pO2 after endotoxin. 36'37Although several methods have been developed to estimate regional renal cortex and medullary blood flow (such as laser-Doppler flowmetry and radiolabeled microspheres), it should be noted that these have been unsuccessful in determining an absolute flow value in vivo] 8 Our MRI measurements provide qualitative assessment of renal hemodynamics--in control kidneys, the cortex region had high S1 (indicative of a high oxygenation level of hemoglobin in this region), whereas the outer medulla showed low SI on these images. Changes in MRI signal intensity were observed immediately after injection of LPS, indicating that they reflect functional, for instance, hemodynamic derangements. Previous MRI measurements of renal T2* during apnea-induced ischemia ~3 have shown that decreased oxygenation of blood results in a simultaneous fall in the SI in both the cortex and outer medulla. Systemicinduced changes in the oxygenation of inflowing blood might, therefore, have caused the observed decrease in pO2, but can be ruled out because the SI of the outer medulla showed a continuous increase on T2* weighted images (see Fig. 5). The transient decrease of SI in all three regions of the kidney after injection of LPS (on T r w e i g h t e d images) can be explained in terms of a systemic decrease in blood flow. The continuous increase of SI in the outer medulla on T2*weighted images, however, can be explained best by

al.

a combination of the effects of LPS-induced redistribution of oxygenated blood, and possibly decreased oxygen demand in the outer medulla, which normally functions on the verge of hypoxia. 38 This is consistent with the increase in pO2 measured by EPR in the outer medulla. These results are in agreement with previous publications: studies using radionuclide-labeled microspheres have shown a 15% increase in renal arteriovenous shunting 2 h after onset of endotoxic shock 39 and significant redistribution of blood from the cortex to the outer medulla regionY Carriere et al. 17 reported decreased outer cortical blood flow but relatively stable inner cortical and outer medullary flow during prolonged hemorrhagic hypotension in the dog. Passmore and Baker 4° confirmed these observations and indicated a return of outer cortical flow during the irreversible stages of hemorrhagic shock. LPS induces production of a variety of mediators with hypotensive effects; vasoactive peptides 41'42 and compounds such as NO are mediated through LPS effects on circulating monocytes and tissue macrophages or endothelial cellsY Cytokines (such as TNFa, IL-I/3) and leukotrienes are known to be produced in the kidney early after injection of LPS in mice and have since been regarded as one of the earliest mediators of s h o c k . 43 All of these mediators combine to cause vasodilation and localized increased perfusion. We were able to prevent the LPS-induced changes in pO2 by pretreating mice with NMMA to inhibit the production of NO. This suggests that the latter is active in inducing localized vasodilation in the outer medulla region of the kidney after LPS, thereby increasing medulla pO2, and is consistent with a model for endotoxic shock in which LPS stimulates NO synthesis. 44 We were unable to detect increased levels of circulating NO in the blood of LPS treated mice until 2 - 3 h postLPS; however, this might reflect the lack of specificity of the assay rather than little production of NO. Westenberger et al., 45 measured formaton of an NO derivative of hemoglobin in the blood of rats during the shock syndrome, and found that NO detection in blood was evident only after 2 - 3 h (peaking at 6 - 8 h). What is more, they demonstrated that pretreatment with LNMMA could prevent appearance of this complex in the blood. The drop in cortex pO2 induced by LPS also was prevented by pretreatment with NMMA; this might have been due to maintenance of a high vascular resistance in the medulla as the result of inhibition of NO synthesis (but might also reflect inhibition of NO production systemically). Our results suggest that these effects on renal pO2 were caused not by induction of NO synthesis systemically (which takes up to several hours) but probably

Renal oxygenation after endotoxin reflect the stimulation of constitutive ( p r e - f o r m e d ) nitric oxide synthase within the renal outer medulla. It is possible that endotoxin upregulates constitutive nitric oxide synthase in the kidney, but we can find no evidence in the literature to support this finding. However, these effects on NO synthesis m a y not necessarily reflect direct effects of endotoxin, but may be the result of stimulation by e n d o t o x i n - i n d u c e d cytokines. Brezis et al. 38 have shown that N M M A decreases the medullary pO2 by inhibiting production of N O and thereby decreasing medullary perfusion. Our results are in agreement with others 46 in suggesting that rapid localized production of N O from preformed e n z y m e in the renal m e d u l l a plays an important role in the regulation of renal blood flow and function. The recovery of pO2 in both regions to control values at around 40 m i n post-LPS might have been due to a direct toxic effect of LPS on cellular mitochondrial function, 34 resulting in decreased oxygen utilization. It is also possible that vascular endothelial cells lining the entire circulatory system function as effective functional vascular oxygen sensors 47 reacting to increased oxygen c o n s u m p t i o n or reduced oxygen content of blood by causing local vasodilation and increased regional blood flow. W e have shown that E P R oximetry can be used to measure tissue pO2 n o n i n v a s i v e l y at several sites simultaneously within the kidney, and have shown that endotoxin causes changes in renal pO2 very early ( w i t h i n m i n u t e s ) during sepsis. Future experiments c o m b i n i n g both EPR and M R I techniques will be valuable in obtaining direct in vivo measurements in disease states in which tissue pO2 will be used as an index of damage or improved therapy.

7.

8. 9.

10. 11.

12. 13.

14.

15. 16. 17. 18.

Acknowledgements - - This research was supported by NIH Grant

19.

GM-34250 and used the facilities of the Dartmouth Medical School EPR Research Center, which is supported by NIH Grant RR 0181 l. The authors would like to thank Professor Heinz Valtin (Department of Physiology, Dartmouth Medical School) for his enthusiasm for this work and critical reading of the manuscript.

20.

21. REFERENCES

1. Epstein, F.H.; Balaban, R. S.; Ross, B. D. Redox state of cytochrome aa3 in isolated perfused rat kidney. Am. J. Physiol. 243:F356-F363; 1982. 2. Wilson, D. F.; Erecinska, M.; Drown, D.; Silver, I. A. Effect of oxygen tension on cellular energetics. Am. J. Physiol. 233:C135-C 140; 1977. 3. Antonsson, J. B.; Fiddian-Green, R. G. The role of the gut in shock and multiple system organ failure. Eur. J. Surg. 157:312; 1991. 4. Gullichsen, E. Renal perfusion and metabolism in experimental endotoxic shock. Acta Chir. Scand. (Suppl.) 560:7-31; 1991. 5. Glockner, J. F.; Swartz, H. M. In vivo EPR oximetry using two novel probes: Fusinite and lithium phthalocyanine.In: Erdmann, W.; Bruley, D. F., eds. Oxygen transport to tissue X1V. New York: Plenium Publishing; 1992:221-228. 6. Lui, K. J.; Gast, P.; Moussavi, M.; Norby, S. W.; Vahidi, N.;

22. 23. 24. 25. 26. 27.

33

Walczak, T.; Wu, M.; Swartz, H. M. Lithium phthalocyanine: A probe for electron paramagneticresonance oximetry in viable biological systems. Proc. Natl. Acad. Sci. USA 90:5438-5442; 1993. Swartz, H. M.; Bacic, G.; Friedman, B.; Goda, F.; Grinberg, O. Y.; Hoopes, P. J.; Jiang, J.; Liu, K. J.; Nakashima, T.; O'Hara, J.; Walczak, T. Measurement of pO2 in vivo, including human subjects by electron paramagnetic resonance. In: Hogan, M. C., et al., ed. Oxygen transport to tissue XVI. New York: Plenium Publishing; 1995:119-128. Smirnov, A,; Norby, S. W.; Clarkson, R. B.; Walczak, T.; Swartz, H. M. Simultaneous multi-site EPR spectroscopy in vivo. Magn. Reson. Med. 30:213-220; 1993. Lohr, J.; Mazurchuk, R. J.; Acara, M. A.; Nickerson, P. A.; Fiel, R. J. Magnetic resonance imaging (MRI) and pathophysiology of the rat kidney in streptozotocin-induced diabetes. Magn. Reson. Imaging 9:93-100; 1991. Marotti, M.; Hricak, H.; Terrier, F.; McIaninch, J. W.; Phuroff, J. W. MR in renal disease: Importance of cortical medullary distinction. Magn. Reson. Med. 5:160-172; 1987. Terrier, F.; Lazeyras, F.; Posse, S.; Aue, W. P.; Zimmerrnann, A.; Frey B. M.; Frey, F. J. Study of acute renal ischemia in the rat using magnetic resonance imaging and spectroscopy. Magn. Reson. Med. 12:114-136; 1989. Nagrani, N. K.; Mattison, D.; Jesmanowicz, A.; Greene, A. S.; Cowley, A. W.; Hyde, J. S. In vivo high resolution echo-planar imaging of rat kidney. SMRM 1019;1994. Vexler, V. S.; de Crespigny, A. J. S.; Wendland, M. F.; Kuwatsum, R.; Muhler, A.; Brasch, R. C.; Moseley, M. E. MR imaging of oxygenation-dependentchanges in focal renal ischemia and transplanted liver tumor in rat. J. Magn. Reson. Imaging 3:48349O; 1993. Hoppel, B. E.; Weisskoff, R. M.; Thulborn, K. R.; Moore, J. B.; Kwong, K. K.; Rosen, B. R. Measurement of regional blood oxygenation and cerebral hemodynamics. Magn. Reson. Med. 30:715-723; 1993. Ogawa, S.; Lee, T.-M.; Nayak, A. S.; Glynn, P. Oxygenationsensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn. Reson. Med. 14:68-78; 1990. Turek, P.; Andre, J. J.; Giraudeau, A.; Simon, J. Preparation and study of a lithium phthalocyanineradical: Optical and magnetic properties. Chem. Phys. Lett. 134:471-476; 1987. Carriere, S.; Thorburn, G. D.; O'Morchoe, C. C. C.; Barger, A. C. Intrarenaldistributionof blood flow in dogs during hemorrhagic hypotension. Circ. Res. 19:167-179; 1966. Hansell,P. Evaluationof methods for estimatingrenal medullary blood flow. Renal Physiol. Biochem. 15(5):217-230; 1992. Nilges, M. J.; Walczak, T.; Swartz, H. M. 1 GHz in vivo ESR spectrometer operating with a surface probe. Phys. Med. 5:195201; 1989. Green, L. C.; Wagner, D. A.; Glogowski, J.; Skipper, P. L.; Wishnok, J. S.; Tannenbaum, S. R. Analysis of nitrate, nitrite, and [tSN] nitrate in biological fluids. Anal. Biochem. 126:131138; 1982. Verdon, C. P.; Burton, B. A.; Prior, R. L. Sample pretreatment with nitrate reductase and glucose-6-phosphate dehydrogenase quantitatively reduces nitrate while avaoiding interference by NADP + when the Griess reaction is used to assay for nitrite. Anal. Biochem. 224(2):502-508; 1995. Reidy, M. A.; Schwartz, S. M. Endothelial injury and regeneration. VI. Endotoxin: A nondenuding injury to aortic endothelium. Lab. Invest. 48:25-34; 1983. Pallone, T. L.; Robertson, C. R,; Jamison, R. L. Renal medullary microcirculation. Physiol. Rev. 70:885-920; 1990. Leichtweiss, H. P.; Lubbers, D. W.; Weiss, C. H.; Baumgartle, H.; Reschke, W. Pflugers Arch. 309:328-349; 1969. Cavagh, D.; Rat, P. S.; Sutton, D. M. C.; Bhagat, B. D.; Bachmann, F. Pathophysiology of endotoxin shock in the primate. Am. J. Obstet. Gynecol. 108:705-722; 1970. Ramsay, G.; Runcie, C. Hepatic dysfunction in shock. In: Update in intensive care and emergency medicine. New York: Springer Verlag; 1989:368-375. Vincent, J. L.; De Backer, D. Initial management of circulatory

34

28. 29.

30. 31. 32.

33.

34.

35.

36.

37.

P.E. JAMESet al. shock as prevention of MSOF. Crit. Care Clin. 5:369-378; 1989. Cohen, J. J.; Black, A. J.; Wertheim, S. J. Direct effects of endotoxin on the function of the isolated perfused rat kidney. Kidney Int. 37:1219-1226; 1990. Hinshaw, L. B.; Spink, W. W.; Vick, J. A.; Matlet, E.; Finstad, J. Effects of endotoxin on kidney function and renal hemodynamics in the dog. Am. J. Physiol. 201:144-148; 1961. O'Hair, D. P.; Adams, M. B.; Tunberg, T. C.; Osborne, J. L. Relationships among endotoxemia, arterial pressure and renal function in dogs. Circ. Shock 27:199-210; 1989. Wardle, E. N. Endotoxin and acute renal failure. Nephron t4:321-332; 1975. Frey, L.; Kreimeier, U.; Schwartz, G.; von Hirschhausen, E,; Messmer, K. Deterioration of renal function during hyperdynamic endotoxemia correlates with redistribution of intra-renal blood flow. Eur. Surg. Res. 22(Suppl. 1):66; 1990. de Navasquez, S. Experimental symmertical cortical necrosis of the kidney produced by staphylococcus toxin. A study of the morbid anatomy and associated circulatory and biochemical changes. J. Pathol. Bact. 46:47-65; 1938. Baumgartl, H.; Leichtweiss, H. P.; Lubbers, D. W.; Weiss, C.; Huland, H. The oxygen supply of the dog kidney: Measurements of intrarenal pO2. Microvasc. Res. 4:247-257; 1972. James, P. E.; Jackson, S. K.; Grinberg, O. Y.; Swartz, H. M. The effects of bacterial endotoxin on oxygen consumption of various cell types in vitro: An EPR oximetry study. Free Radic. BioL Med. 18(4):641-647; 1995. Brezis, M.; Heyman, S. N.; Epstein, F. H. Determinants of intrarenal oxygenation.II. Hemodynamic effects. Am. J. Physiol. 267(6pt2) :F1063-8; 1994. Atkins, J. L.; Lankford, S. P. Changes in cytochrome oxidation

38.

39. 40. 41. 42. 43.

44. 45.

46.

47.

in outer and inner stripes of outer medulla. Am. J. Physiol. 261 (5pt2) :F849-57; 1991. Brezis, M.; Heyman, S. N.; Dinour, D.; Epstein, F. H.; Rosen, S. Role of nitric oxide in renal medullary oxygenation. Studies in isolated and intact rat kidneys. J. Clin. Invest. 88:390-395; 1991. Archie, J. P. Anatomic arterial-venous shunting in endotoxic and septic shock in dogs. Ann. Surg. 186:171-176; 1977. Passmore, J. C.; Baker, C. H. Intrarenal blood flow distribution in irreversible hemorrhagic shock in dogs. J. Trauma 13:10661074; 1973. Dinarello, C. A. Interleukin-1 and its biologically related cytokines. Adv. ImmunoL 44:153-205; 1989. Jaattela, M. Biology of disease: Biologic activities and mechanisms of action of tumor necrosis factor-alpha/cacbectin. Lab. Invest. 64:724-742; 1991. Xia, Y.; Feng, L.; Yoshimura, T.; Wilson, C. LPS-induced MCP-I, IL-1B, and TNF-alpha mRNA expression in isolated erythrocyte-perfused rat kidney. Am. J. Physiol. 264 (33) :F744 F780; 1993. Wang, Q.; Jacobs, J.; DeLeo, J.; Kruszyna, H.; Kruszyna, R.; Smith, R.; Wilcox, D. Nitric oxide hemoglobin in mice and rats in endotoxic shock, iLife Sci. 49:55-60; 1991. Westenberger, U.; Thanner, S.; Ruf, H. H.; Gersonde, K.; Sutter, G.; Trentz, O. Formation of free radicals and nitric oxide derivative of hemoglobin in rats during shock syndrome. Free Radic. Res. Commun. 11(1-3):167-178; 1990. Mattson, D. L.; Lu, S.; Nakanishi, K.; Papanek, P. E.; Cowley, A. W., Jr. Effect of chronic renal medullary nitric oxide inhibition on blood pressure. Am. J. Physiol. 266(5 Pt 2):H1918H1926; 1994. Pohl, U. Endothelial cells as part of a vascular oxygen-sensing system: Hypoxia-induced release of autacoids. Experientia 46:1175-1179; 1990.