Department of Physiology, Medical College of Wisconsin,
Milwaukee, Wisconsin 53226
The unique role of
nitric oxide (NO) in the regulation of renal medullary function is
supported by the evidence summarized in this review. The impact of
reduced production of NO within the renal medulla on the delivery of
blood to the medulla and on the long-term regulation of sodium
excretion and blood pressure is described. It is evident that medullary
NO production serves as an important counterregulatory factor to buffer
vasoconstrictor hormone-induced reduction of medullary blood flow and
tissue oxygen levels. When NO synthase (NOS) activity is reduced within
the renal medulla, either pharmacologically or genetically [Dahl
salt-sensitive (S) rats], a super sensitivity to vasoconstrictors
develops with ensuing hypertension. Reduced NO production may also
result from reduced cellular uptake of L-arginine in the
medullary tissue, resulting in hypertension. It is concluded that NO
production in the renal medulla plays a very important role in sodium
and water homeostasis and the long-term control of arterial pressure.
 |
INTRODUCTION |
THE RENAL MEDULLARY CIRCULATION is now
recognized to be of importance for reasons that extend far beyond
providing the economy of countercurrent exchange to concentrate large
volumes of filtrate to produce small volumes of concentrated urine. It
is now recognized that medullary blood flow (MBF) can play an important
role in determining long-term levels of arterial pressure and that
15-30% reductions of blood flow to the renal medulla in the face
of imperceptibly small changes of total renal blood flow can lead to
the development of hypertension (16, 21). Evidence is
reviewed showing that nitric oxide (NO) plays a unique role in the
acute and chronic regulation of renal MBF, sodium homeostasis, and
arterial pressure. More specifically, the role of NO in the regulation
of MBF, in linking tubular metabolic needs with delivery of nutrients
and as a counterregulatory system to protect the medulla from the consequences of underperfusion, are the focus of this review. The
influence of NO on the renal cortex and such issues that relate to
pressure-natriuresis, autoregulation of total renal blood flow, tubuloglomerular feedback, and glomerular filtration rate (GFR) regulation are important subjects of renal function that either have
been reviewed by others (2, 29, 43, 55,
96) or are beyond the scope of this relatively brief review.
| |
NO PRODUCTION AND THE OVERALL REGULATION OF RENAL FUNCTION |
NO within the kidney is now known to contribute importantly
to low renal vascular resistance (75, 115) and to be able
to stimulate natriuresis and diuresis independent of renal perfusion pressure as reviewed by others (2, 5, 43, 50, 75, 104,
137). The slope of the pressure-natriuresis relationship has
been shown to be reduced with the inhibition of renal NO synthase (NOS)
activity (23, 28, 98). As reviewed recently by Liang and
Knox (50) and by Ortiz and Garvin (80), there
is evidence that NO reduces sodium reabsorption in all tubular
segments, including the proximal tubule (50, 80), thick
ascending limb (78, 79), distal tubules, and cortical
collecting ducts (80), although medullary nephron segments
and collecting ducts remain to be studied in detail. Studies have also
demonstrated that tubuloglomerular feedback control of afferent
arteriolar resistance is influenced by macula densa NO production,
thereby enhancing autoregulation of renal blood flow, GFR, and renin
secretion (30, 37, 53, 75, 122, 128).
| |
INITIAL EVIDENCE THAT NO PARTICIPATES IN THE ACUTE REGULATION OF
MEDULLARY FUNCTION |
Evidence that an endothelial-derived relaxing factor (EDRF) within
the renal medulla could play a role in medullary oxygenation was first
reported by Biondi et al. (6). Functional evidence that
basal levels of NO contribute to MBF resulted from a series of studies
in our laboratory that examined the effects of a variety of vasoactive
compounds on MBF. These compounds were assessed by direct delivery into
the medullary interstitial space of rats by use of a small implanted
catheter (51, 52), whereas MBF responses were determined
using external and implanted optical fibers and laser-Doppler flowmetry
techniques (52). Substances infused into the medullary
interstitium became trapped and concentrated by the countercurrent vasa
recta circulation, enabling responses to be determined independent of
the cortex.
It was found that the most dramatic reduction of MBF was obtained with
medullary infusion of the nonspecific NOS inhibitor NG-nitro-L-arginine methyl ester
(L-NAME) (62), a compound that reduced flow to
this region by nearly 40% (61). No changes of cortical
flow were observed. Similarly, infusions of bradykinin (57) and acetylcholine (62), NO-dependent
vasodilators, also produced substantial increases of MBF. Several
important conclusions were drawn from these in vivo studies in
anesthetized rats. First, MBF could be modulated independently of
cortical blood flow (CBF). Second, the medullary vasculature could
respond to many of the same vasoactive compounds as arterioles in other
regions of the body. Third, NO exerted a tonic influence on the renal
medullary vasculature.
| |
LOCALIZATION OF NO PRODUCTION WITHIN VASCULAR AND TUBULAR
STRUCTURES OF THE KIDNEY |
The extent to which the antinatriuresis associated with medullary
NOS inhibition was a direct consequence of reduction of MBF (16,
102) or was related to changes of intrarenal paracrine/autocrine systems (95, 100) or was a direct consequence of tubular
epithelial transport actions of NO (80, 136) awaited a
better understanding of the intrarenal sites of NO production. Initial
efforts to determine regional NO biosynthesis by determining tissue
nitrite/nitrate and cGMP levels in slices of dissected renal zones,
isolated glomeruli, and various cell cultures systems. These studies
indicated that tissues of both the renal cortex and medulla were indeed
capable of generating NO in abundant amounts (6, 10, 13, 25, 36,
69, 93, 108).
Recently, techniques have enabled discrete measurements of segmental
tubular and microvascular measurements of NOS enzyme activity,
expression of the mRNA and protein levels, and interstitial fluid NO
concentrations ([NO]). NOS activity assays incubated tissue protein
with L-[3H]arginine and appropriate cofactors
and the conversion of L-arginine (L-Arg) to
L-citrulline (Cit) was quantified by reverse-phase HPLC
and radiochemical detection (63, 68). Evidence was
obtained showing that the renal medulla is the major source of NO
production in the rat kidney. Regional comparisons of NOS activity and
[NO] are summarized in Fig. 1. NOS
activity was found to be 26 times higher in the inner medulla and 4 times higher in the outer medulla than in the renal cortex of
Sprague-Dawley rats (130). In vivo microdialysis
techniques using oxyhemoglobin in the dialysate to trap tissue NO found
that concentrations of interstitial fluid NO averaged nearly 105 ± 18 nmol/l in the renal medulla compared with 62 ± 16 nmol/l in
the cortex (138). These observations conform with data
showing that NOS I and NOS II mRNA levels were more than five times
greater in the outer medulla of Brown Norway (BN) rats compared with
cortex (132). Moreover, the protein expression levels of
all three isoforms determined by Western blot analysis were observed to
be three to five times higher in this region in BN rats (Cowley,
unpublished observations), as also reported for NOS III in kidneys of
Sprague-Dawley rats (77).
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 1.
Top: results are summarized for regional
nitric oxide synthase (NOS) activity determined by HPLC as the
formation of L-citrulline in homogenates of renal
cortex (CX), outer medulla (OM), and inner medulla (IM).
L-Arginine content in the reaction mixture of each tissue
sample was adjusted to ~23 mmol/l by addition of exogenous cold
L-arginine to OM and IM samples. * Significantly
different (P < 0.05) from CX. Data regraphed from the
original manuscript by Wu and colleagues (130).
Bottom: regional concentration of NO, as determined by an
oxyhemoglobin trapping technique, is shown. Data regraphed from the
original manuscript by Zou et al. (136).
|
|
The capacity to produce NO is best represented by NOS activity, and
studies have been carried to determine activity levels in
microdissected glomeruli, tubular, and microvascular segments of the
cortex and medulla of Sprague-Dawley rat kidneys (130). Inner medullary collecting ducts (IMCD) were found to contain the
greatest levels of NOS activity (11.5 fmol
Cit · mm
1 · h1),
greatly exceeding the activity found in the glomeruli (1.9 fmol
Cit · glomerulus1 · h1).
All other tubular elements exhibited significantly less NOS activity,
including proximal convoluted tubule, pars convoluta, pars recta,
cortical thick ascending limb, cortical collecting duct, outer
medullary collecting duct, outer medullary thick ascending limb, and
inner medullary thin limb. The medullary vasa recta bundles of the
outer medulla contained relatively high levels of total NOS enzymatic
activity (3.2 fmol
Cit · mm1 · h1),
exceeded only by that in the IMCDs. Only about one-third as much was
found in the afferent arterioles and considerably less in the arcuate
and interlobular arteries (63). Similar results were found
when relative enzyme activity levels were expressed based on segment
length or as total protein.
NOS I and NOS III mRNA was found to be expressed in both the glomeruli
and in the vasa recta, whereas the IMCD contained mRNA for all three of
the NOS isoforms (130). As initially observed by Terada et
al. (121) and later confirmed by Ugie et al.
(125) and ourselves (63), there appears to be
little or no NOS mRNA for any of the known isoforms expressed in rat
early proximal tubule, although NOS II mRNA was reported to be
expressed in the S3 segment of the proximal tubule, based on in situ
hybridization studies (121). Taken together, the data
indicate that the NOS activity in IMCDs contributes importantly to the
higher levels of interstitial [NO] observed in the medulla compared
with the cortex.
| |
EVIDENCE FOR PARACRINE AND AUTOCRINE NO FUNCTION IN THE RENAL
MEDULLA ("TUBULOVASCULAR CROSS-TALK") |
The use of fluorescent dyes for detection of NO and
Ca2+ ions has begun to provide insights regarding the
ability of medullary tubular epithelial cells to act as a paracrine NO
source for the regulation of vasa recta function. The focus of these
studies has been on the larger diameter vasa recta vessels of the outer medulla for reasons related to the unique structure and function of the
medullary vasa recta circulation. Before the results of these studies
are reviewed, it is important to first appreciate the microvascular
anatomy of the renal medulla, as described in detail by others
(3, 4, 40, 44, 64, 81-83, 127). In brief, it is
recognized that blood flow to the medulla of mammalian kidneys is
supplied by postglomerular microvessels derived from juxtamedullary
glomeruli, the deep glomeruli of the inner cortex that receive their
blood supply from the interlobular branches of the arcuate arteries, as
described by Kaissling and Kriz (39). A photomicrograph of
one of these juxtamedullary glomeruli (Glm) is shown in Fig.
2A. As seen in this figure,
the efferent arterioles of these juxtamedullary glomeruli immediately
radiate into the vasa recta capillary bundles of the outer medulla
(OMDVR) that then coalesce into long descending vasa recta (DVR) loops
supplying the inner medulla and papilla. Although the blood flow to the medulla emanates from the renal cortex (CX) and changes of cortical flow can importantly influence blood flow to the medullary region, there is evidence as presented in this review that flow to the medulla
can be independently modulated. Regional regulation of MBF is important
to supply nutrients and remove tubular reabsorbate. It is important to
preserve the interstitial gradients for NaCl and urea established by
the loops of Henle and collecting ducts for urine concentration.
View larger version (79K):
[in this window]
[in a new window]
|
Fig. 2.
Fluorescent microscopic image of outer medullary region
with thick ascending limbs illuminated by superoxide
(O )-sensitive fluorescent dye dihydroethidium (DHE)
and with the vasa recta filled with latex microspheres (0.20 µm
sized) as described in Ref. 21. The left kidney was
removed, sliced into ~500-µm-thick sections, and imaged using a
cooled charge-coupled device camera equipped with Nikon E600
fluorescence microscope to capture the fluorescent signals.
Fluorescence was stimulated with 490-nm excitation filter and collected
through 605/40-nm band path emission filter with images captured by
MetaMorph software and enhanced by Adobe Photoshop software.
A: a ×40 magnification of a juxtamedullary region of CX and
OM region. Medullary thick ascending limbs (mTAL) are illuminated by
DHE, and the outer medullary descending vasa recta (OMDVR) bundles are
seen as a dark stripes filled with latex microspheres. B:
×100 magnification of vascular bundles (OMDVR) with surrounding mTAL.
C: a light microscopic image of the pericytes of OMDVR
(filled with microspheres) as seen in a micro-tissue strip
isolated from the inner stripe of OM. [From Dickhout et al.
(21).]
|
|
Medullary DVR vessels are unique in that they have characteristics of
true capillaries with continuous endothelium but are also surrounded by
smooth muscle remnants known as pericytes, as shown in Fig.
2C. The DVR pericytes contain 
-smooth muscle actin
(88), and the ability of the pericytes to constrict and reduce the diameter of DVR has been demonstrated by Pallone and colleagues (86) using isolated perfused vasa recta vessels
from the outer medulla. For example, these vessels have been shown to
respond to vasoconstrictors such as ANG II, as shown in Fig. 3A (85), AVP
(124), and endothelin (112), as well as
vasodilators such as bradykinin (83), prostaglandin
E2 (112), and adenosine (111).
The ability of the DVR to behave like small arterioles accounts for the
ability of the medullary flow to be controlled independently from the
cortical circulation when vasoactive substances are endogenously
released or infused. This also accounts for the in vivo medullary flow
responses observed when vasoactive agents are delivered into the renal
medullary interstitium, such as AVP (72), norepinephrine
(NE) (135), and ANG II (138). ANG II-induced reduction of
MBF was also shown to be significantly enhanced by pretreatment with
the cyclooxygenase inhibitor meclofenamate (61).
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 3.
A: light micrograph of an OMDVR under basal conditions
(top) and after addition of 10 nmol/l ANG II
(bottom). Arrows indicate foci of vasoconstriction.
(Courtesy of Dr. T. L. Pallone.) B: top
shows a pseudo-color image of OMDVR pericytes in a control state after
loading with the NO-sensitive dye 4,5-diaminofluorescein deacetate
(DAF-2DA) in microtissue strips. Bottom: same cells after
ANG II (1 µmol/l) administration in the suffusate demonstrating the
increased NO concentration ([NO]). C: time course of
intracellular [NO] ([NO]i) response of pericytes of the
vasa recta with endothelium disrupted by microbeads perfusion in OMVDR
to ANG II (1 µmol/l) measured by changes in DAF-2DA fluorescence
intensity. Included in these micro-tissue strips were vasa recta
surrounded by mTAL. * Bonferroni-corrected statistically significant
increase of the means from 50 to 150 s after stimulation. [From
Dickhout et al. (21).]
|
|
On the basis of these anatomic structures and blood flow responses of
the outer medulla, studies were carried out to determine whether NO
produced by the tubular epithelium of the medullary thick ascending
limbs of Henle (mTAL) might act both as an autocrine factor within
tubular epithelia and as a paracrine factor on surrounding structures.
Although nephron segments are largely excluded from the central core of
the vasa recta bundles, at the periphery of the bundle these vessels
give rise to the dense capillary plexuses that perfuse the outer
medulla. As shown in Fig. 2, A and B, these peripheral vasa recta are surrounded by medullary mTAL, and it was
logical to examine whether these tubules could communicate information
to the DVR via the release of NO.
The movement of NO in vivo has received far less attention than the
biochemical reactions that control its rate of synthesis and
degradation. The importance of the volatility, solubility, and
diffusability of NO in determining its biological effects, however, has
been studied in some detail and reviewed by Lancaster (46). The results have shown that under all physiological
conditions (except when there is a gas-liquid interface as in the
lung), NO exists completely as a dissolved nonelectrolyte. NO is
slightly more soluble in aqueous solution than O2, and in
virtually all its biological actions NO is not a gas unless
concentrations approach 1-2 mmol/l, a highly unlikely situation in
vivo. The solubility of NO, although similar to O2, is
nearly nine times greater in organic solvents, so that NO concentrates
in hydrophobic phases in tissues such as membranes, lipid inclusions,
and lipoprotein complexes. The lipid bilayer of membranes therefore
poses no barrier to the movement of NO in tissue. On the basis of
mathematical modeling of the random diffusion of NO within tissues, it
has been estimated that NO would be able to act at large distances from
a single cellular source and thus affect an enormous number of other
cells surrounding this single source (47). It has been estimated that NO could diffuse in tissues from 0.2 to 0.25 mm in
5-10 s (46), well within the range generally
considered to be the tissue half-life of NO.
Microdialysis of renal interstitial fluid using oxyhemoglobin trapping
has demonstrated relatively high levels of NO as well as nitrite
concentrations in medullary interstitial fluid as shown in Fig. 1
(135, 136, 138). These NO levels were increased by intravenous infusion of L-Arg and decreased by
L-NAME (136). Tubular and vascular segments of
both the cortex and medulla therefore appear to be constantly exposed
to NO that could serve a paracrine function.
Availability of novel fluorescent probes has enabled direct
observation, at high temporal and spatial resolution, of the responses of intracellular NO and Ca2+ within renal tubular and
epithelium vasa recta imaged from thin renal medullary tissue strips.
Intracellular Ca2+ concentration
([Ca2+]i) has been quantified using fura 2-AM
and changes of NO imaged using the NO-sensitive dye
4,5-diaminofluorescein deacetate (DAF-2DA). These techniques were
initially used to study mechanisms of AVP-stimulated NO production in
the IMCD epithelial cells. It was shown that AVP produced a rapid
increase of [Ca2+]i followed by a significant
increase of intracellular NO within these cells (67). The
NO response was abolished both by an NO scavenger, carboxy
2-(4-carboxyphenyl)-4,4,5,5,tetramethylimidazoline-1-oxyl-3-oxide, and by preincubation with the Ca2+-ATPase inhibitor
thapsigargin. These results clearly demonstrated that NO acts as a
signaling molecule in the IMCD.
As illustrated in Fig. 3, studies have also shown that NO produced
within the mTAL can serve both a paracrine and autocrine function to
control vascular tone of the surrounding vasa recta. The paracrine
function was demonstrated when techniques were developed to distinguish
the fluorescent signals of the contractile pericytes from the
stronger signals of the underlying vasa recta endothelial cells. To do
this, the fluorescent signals of the underlying endothelial cells were
abolished by perfusing the kidneys with 0.2-µm latex microspheres
before the removal of the microtissue strip from the outer medulla for
study under the fluorescent microscope. This enabled a 10-fold increase
in fluorescent signal gain of the detection unit for the acquisition of
the less intense signals of the vasa recta pericytes without
overexposure of fluorescent signals from the endothelial cells. The NO
response to ANG II of an outer medullary vasa recta pericyte surrounded
by mTAL in the microtissue strip loaded with NO-sensitive dye DAF-2DA
was then measured (Fig. 3C).
These techniques were used to determine if the constrictor actions of
ANG II on the vasa recta pericytes were buffered by NO produced by
either tubular epithelial cells surrounding the vasa recta and/or by NO
produced by endothelial cells within the vasa recta themselves
(21). Using thin microtissue strips, ANG II (1 µmol/l)
produced an increase of intracellular [NO] ([NO]i) in
the vasa recta pericytes only when the vessels were adjacent to the
mTAL. Pericytes of completely isolated vasa recta without surrounding
mTAL showed a rapid peak increase in [Ca2+]i
(21), consistent with observations of Reinhardt et al.
(97). Importantly, however, DVR pericytes did not show any
increase of NO production either in response to ANG II or with
administration of a Ca2+ ionophore, indicating a lack of
Ca2+-sensitive NO production in vasa recta pericytes. Using
isolated vasa recta with intact endothelium, ANG II was found to reduce Ca2+ within the endothelial cells as was also observed by
Pallone and colleagues (85, 86) and failed to increase NO,
although the Ca2+ ionophore was able to increase NO in
these endothelial cells. Tubular epithelial cells responded quite
differently in that increases of NO were observed in single isolated
mTAL in response to both ANG II and Ca2+ ionophore.
Although it is evident that the concentrations of ANG II used in these
studies far exceeded levels in the systemic circulation, it appears
that intrarenal levels may be as high as 109 to
1010 mol/l (73, 107) so the medullary
structures may be exposed to nanomolar levels of ANG II
(75). Pallone et al. (85) observed isolated
DVR diameter reductions at concentrations of 10 nmol/l (see Fig.
3A).
The autocrine function of NO produced in rat thick ascending limb (TAL)
segments was established by Oritz and Garvin (78), who
demonstrated in perfused isolated tubules that endogenously produced NO
inhibited HCO3 transport. Although it is not
stated in this study whether TAL segments were obtained from the cortex
or medulla, it is likely that these events take place in both segments.
On the basis of these studies, autocrine/paracrine concepts emerged as
summarized in Fig. 4. ANG II increases
[NO]i in pericytes only when the vasa recta vessels are
surrounded by renal tubular segments, mostly mTAL in this case. Neither
the endothelium nor pericytes can directly increase NO production in
response to ANG II. ANG II increases intracellular calcium
concentration in the pericytes of the DVR, apparently as part of its
constrictor action on these vessels. This response is buffered by NO
diffusing from tubular elements surrounding vasa recta within the outer
medulla, a paracrine function that we have called tubulovascular
cross-talk. The close proximity of mTAL to the vasa recta bundles in
the outer medulla (~50-150 µm) suggests that these tubular
elements would be of greater importance in this paracrine cross-talk
than the outer medullary collecting ducts, which are of a greater
distance to the vessels (~150-250 µm) (39, 44,
49). The extent to which hemoglobin within the vasa recta
capillary circulation would modify the in vivo NO diffusion profile is
unclear. However, NO derived from mTAL is not exposed to hemoglobin
before reaching the pericytes of the vasa recta. Because there is
considerable interstitial space and fluid surrounding the individual
vasa recta to provide NO to the pericytes (~10- to 30-µm distances
separating the vasa recta; personal observation), intravascular
hemoglobin would not be expected to play a major role in buffering the
medullary interstitial NO gradients. The diameter of the entire vasa
recta bundle does not exceed the estimated diffusion distance of NO within biological fluids (0.2-0.25 mm) as indicated above
(46, 47). Finally, although the deep vasa recta of the
inner medulla remain to be studied, it is likely that these vessels may
be controlled in a similar fashion by the high levels of NO produced by
the epithelial cells of the IMCD, because pericytes have been shown to
exist surrounding these vasa recta, although less densely distributed (88).
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 4.
Schematic summary of the observed actions of ANG II on NO
levels and intracellular Ca2+ concentration
([Ca2+]i) in outer medullary vascular
bundles. ANG II increases [Ca2+]i in
pericytes of the descending vasa recta and reduces
[Ca2+]i in endothelium of the descending vasa
recta. ANG II also increases [NO]i in the pericytes of
the descending vasa recta but only when these cells are in proximity to
the mTAL surrounding the outer medullary vascular bundles. ANG II
increases [Ca2+]i and [NO]i in
mTALs even when these tubules were studied in isolation. These
observations indicate that ANG II exerts a constrictor effect on the
descending vasa recta by direct action on pericytes and that this
constrictor action is buffered by NO diffusing from mTALs to the
pericytes of the descending vasa recta. [Schematic modified from
original in Dickhout et al. (21).]
|
|
| |
IMPORTANCE OF MEDULLARY NO PRODUCTION IN THE LONG-TERM CONTROL
OF MBF AND ARTERIAL PRESSURE |
The first evidence that basal NO production played an important
role in determining arterial blood pressure arose from observations that chronic oral or intravenous administration of NOS inhibitors such
as L-NAME or
NG-monomethyl-L-arginine
(L-NMMA) produced hypertension (28, 32, 55,
123). It was unclear, however, whether the observed hypertension was primarily a result of generalized systemic vasoconstriction, an
effect of NO reduction in brain, or a primary consequence of reduced
renal excretory function. Studies were carried out to determine whether
medullary NO production per se played an important role in the
long-term regulation of renal MBF, sodium excretion, and arterial
pressure. L-NAME was infused chronically into the renal
medullary interstitium of Sprague-Dawley rats for 5 days. As shown in
Fig. 5, this renal interstitial (ri)
infusion of L-NAME resulted in a sustained reduction of
renal MBF (~30%) with no measurable changes of cortical flow as
determined daily in unanesthetized rats with chronically implanted
optical fibers. This reduction of MBF was associated with an immediate
reduction of sodium excretion, retention of sodium, and the development
of hypertension (59). This demonstrated that reduction of
NO production in the medulla alone could reduce MBF and produce
hypertension.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 5.
Changes in mean arterial pressure (MAP), renal medullary
blood flow (MBF), and cortical blood flow (CBF) in conscious,
uninephrectomized Sprague-Dawley rats during renal medullary
interstitial (ri) infusion of
NG-nitro-L-arginine methyl ester
(L-NAME; 120 µg/h). Blood flow was measured by
laser-Doppler flowmetry with implanted optical fibers. * Significant
difference from 3rd control day. [Regraphed from original data
(59)].
|
|
Consistent with these observations, it was also found that
L-NAME administered in small doses by chronic intravenous
infusion reduced MBF alone, independently of any changes in CBF, and
this again led to sodium retention and hypertension (71).
It had been assumed that hypertension resulting from systemically
administered L-NAME was probably related to a reduction of
total renal blood flow (75) and an overall increase in
systemic vascular resistance related to vascular and central nervous
system NOS inhibition. Together these studies with both systemic
(intravenous) and renal medullary (ri) delivery of L-NAME
demonstrated that the production of NO specifically within the medulla
was of primary importance in the long-term regulation of MBF and sodium
homeostasis and that reduction of NOS activity in the renal medulla
alone could produce chronic hypertension.
| |
INFLUENCE OF MEDULLARY NO PRODUCTION ON BLOOD PRESSURE SALT
SENSITIVITY |
Consistent with the influence of NO on tubular sodium reabsorption
summarized above, it was found that systemic administration of
nonspecific NOS inhibitors resulted in a sodium-dependent form of
hypertension (28, 55, 70, 123). The extent to which reduction of medullary NO independent of the renal cortex could produce
blood pressure salt sensitivity was determined by a series of studies
In our laboratory. The effects of inhibiting two of the specific NOS
isoforms, NOS I and NOS II, were examined. A salt-dependent form of
hypertension occurred in Sprague-Dawley rats when an antisense
oligonucleotide complementary to the code of the initiating region of
the mRNA for NOS I synthesis was infused chronically into the medullary
interstitial space (56). Similar responses were observed
using a mechanistically different inhibitor of NOS I,
7-nitroindazole (7-NI) (56). The antisense probe
reduced NOS I synthase activity by 53% and the 7-NI by 37%. In both
cases, mean arterial pressure rose over 3 to 5 days to ~15 mmHg above resting control levels in rats receiving a high (4%)-salt diet, with
no changes in pressure being observed in rats receiving the low-salt diet.
In another study, the effects of chronically inhibiting the inducible
NOS II isoforms were examined (60). Chronic intravenous infusions of aminoguanidine in Sprague-Dawley rats that resulted in a
49% reduction of Ca2+-independent NOS (inducible NOS) in
the renal medulla, produced a salt-dependent form of hypertension with
mean arterial pressure rising nearly 15 mmHg when these rats were fed a
high-salt (4%) diet. The results also suggested that these events
occurred independently of renal vascular changes, because intravenous
infusion of this inhibitor at the same dose in anesthetized rats
resulted in no significant changes in mean arterial pressure, CBF, or
MBF, whereas urine flow and sodium excretion were significantly
decreased (60).
Thus it appears that reduction of NO production within the renal
medulla alone can lead to a salt-sensitive form of hypertension. Tan et
al. (120) observed hypertensive responses in the absence of any measurable changes in renal plasma flow or GFR with chronic intravenous administration of aminoguanidine in salt-resistant Dahl R
rats exposed to a high-salt diet, consistent with predominant tubular
effects. Recent work by Kakoki et al. (42) and Mattson et
al. (60) using a variety of NOS isoform-specific
inhibitors indicated that NO derived from NOS I and/or NOS II has
minimal effect on blood flow in the renal medulla (42,
60). The results also suggest that the enhanced salt sensitivity
with NOS inhibition may be largely a consequence of a reduction of NOS
I and NOS II activity within the deep nephrons of the renal medulla.
The role of NOS II, however, remains controversial given the lack of
complete specificity of these inhibitors. The antinatriuretic effects
of aminoguanidine have not been consistently observed by others
(43), and firm conclusions regarding the specific roles of
NOS II in the regulation of medullary function await final resolution.
The lack of availability of NOS III-specific inhibitors has precluded
pharmacological approaches to examine this isoform. One would expect,
however, that specific reductions of endothelial NOS would be
accompanied by a lowering of MBF and chronic hypertension as seen with
L-NAME administration. Huang et al. (33)
observed hypertension in NOS III-knockout mice, demonstrating the
importance of NOS III in long-term blood pressure regulation, but this
study did not address the role of NOS III in the renal medulla.
Inasmuch as NOS III is found in both endothelium of the vasa recta and the epithelial cells of mTAL and IMCD (see below), the NO produced by
this NOS isoform could influence both tubular sodium reabsorption and
MBF. Thus it will not be so easy to separate the relative contributions
of NO from these specific regions upon the long-term control of
arterial pressure.
The role of medullary NO in response to changes of sodium intake and
the extent to which such changes participate in the regulation of MBF
remain unclear. Studies in chronically instrumented Sprague-Dawley rats
in our laboratory found no changes in MBF as daily sodium intake was
increased from 0.4 to 4.0% (27). However, rats fed a
high-sodium diet (4.0%) exhibited an increase in NOS-immunoreactive protein for all isoforms in the medullary tissue (58). Ni
and Vaziri (76) observed a decrease in medullary protein
expression of NOS I and II after 3 wk of an 8% sodium diet. Others
found no change in medullary mRNA expression of any of the NOS isoforms (98, 113).
| |
ACUTE COUNTERREGULATORY ACTION OF RENAL MEDULLARY NO
ON CIRCULATING VASOCONSTRICTORS |
The vascular counterregulatory functions of NO were first
suspected, but the data supporting this role for NO were not initially forthcoming. Reduction of tissue NO production with NOS inhibition was
found by some investigators to enhance ANG II-induced reduction of
renal blood flow (1) and to potentiate the vasoconstrictor effects of ANG II in isolated perfused kidneys (105).
Similar findings were obtained using the in vitro perfused
juxtamedullary preparation (35) and in isolated perfused
afferent arterioles (37). Others found no evidence of
enhanced vasoconstriction to either ANG II or NE after NOS inhibition
using a perfused split kidney juxtamedullary preparation when vessel
diameter was restored to normal by sodium nitroprusside
(35), or in studies using an in vitro hydronephrotic
kidney preparation (87).
Considerable evidence has now emerged supporting the view that NO does
serve an important counterregulatory role within the renal medullary
circulation. In this regard, ANG II has been studied in some detail.
There is abundant evidence that ANG II can stimulate NO production in
many cell types, as shown by measurements of nitrate and cGMP
production in neural, cardiac, vascular, and renal cells (8, 14,
34, 45, 106, 126, 133). ANG II also increases the renal
excretion of nitrate and nitrite (20) and cGMP
concentrations in renal cortical interstitial fluid as determined by
microdialysis (110).
Results of these earlier studies were unable to provide conclusive
evidence that NO served as an important counterregulatory system to
vasoconstrictor agents in the renal medulla, because neither tissue
[NO] nor MBF were measured directly. Both of these are needed to
ascertain changes in local NO production. Blood flow measurement is
important because [NO] alone need not reflect increased NO
production, because medullary interstitial tissue and fluid [NO] can
be hemodynamically influenced by the rate of blood flow through the
countercurrent vasa recta circulation ("washout") and by shunt
pathways in the outer medulla, as known to be the case with oxygen gradients.
Studies were therefore carried out to determine the in vivo effects of
ANG II on both MBF and interstitial fluid [NO] (via microdialysis
oxyhemoglobin trapping technique; Ref. 136). As shown in
Fig. 6A, intravenous infusion
of a subpressor dose of ANG II (5.0 ng · kg1 · min1)
produced no reduction of CBF, MBF, or medullary
PO2 levels. This level of infused ANG II,
however, did produce nearly a 150% rise of [NO] within the medullary
interstitial fluid (138). This ANG II-induced rise of
tissue fluid [NO] was abolished when medullary NOS activity was
blunted in these same rats by medullary infusion of a low dose of
L-NAME, a dose carefully determined to have no measurable
effect on basal MBF. In the absence of the ANG II-stimulated increase
of [NO], there was a reduction of both MBF and tissue PO2 of nearly 30%. These results demonstrated
the importance of the NO counterregulatory actions on the regulation of
MBF and oxygen delivery. Since Navar et al. (74) found
that, even under normal conditions, medullary tissue ANG II
concentrations are significantly higher than those of the renal cortex,
the regional generation of NO within the medulla could be of
substantial importance even in reducing the basal vasoconstrictor
actions of ANG II in this region of the kidney.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 6.
Left: effect of intravenous (iv) infusion of
L-NAME ( ) or saline ( l) on
ANG II response of MAP, CBF, and MBF and medullary oxygenation
(PmO2) after 2 control periods (C1 and C2). CBF
and MBF were measured by using laser-Doppler flowmetry, and medullary
tissue oxygen was measured by microelectrode polargraphic technique.
* Significant difference from the control periods (P < 0.05). [Reprinted with permission from Zou et al.
(138).] Right: using a similar study design,
the effect of iv infusion of L-NAME () or
saline () on norepinephrine (NE) response of MAP, CBF,
MBF, and PmO2 is summarized. [From Zou and
Cowley (135).]
|
|
The effects of NE on medullary NO release have also been examined.
Several studies found that NE can induce NO release through the
2-receptors (35, 68, 135).
2-Activation appears to increase the influence of NO in
the glomeruli and proximal tubules of rats (130), and it
has been shown that the vasculature becomes more sensitive to
2-receptor stimulation with chronic NOS inhibition (31). As summarized in Fig. 6B, studies
have demonstrated the in vivo effects of NE on the medullary
circulation. In response to 1-h intravenous infusions of subpressor
amounts of NE (0.1 µg · kg1 · min1),
medullary [NO] increased 43% (135). This response was
inhibited with a selective 2-adrenergic inhibitor
(135). As was observed in the ANG II studies
(138), when medullary NOS activity was blunted by
medullary infusion of a low dose of L-NAME, the NE-induced rise of tissue [NO] was abolished and a reduction of both MBF and
tissue PO2 of nearly 30% was observed
(135). These results again demonstrated the importance of
the NO counterregulatory actions on the regulation of MBF and oxygen delivery.
Similar findings were also obtained in studies using subpressor amounts
of AVP (2 ng · kg1 · min1)
in which a 35% increase of medullary NO was observed
(89), a response mediated via the V2-type
receptor. These observations were consistent with the direct
observations that AVP stimulated NO production in the epithelial cells
of inner medullary collecting ducts (IMCD) as determined by fluorescent
microscopy (67) and other reports that AVP stimulated
increases of urinary cGMP concentrations (103) and that
the inhibition of NOS acutely enhanced systemic pressor responses to
AVP (35, 109).
Together, these studies demonstrated two important aspects of medullary
NO production. First, subpressor elevations of circulating ANG II, NE,
and AVP result in a substantial increase of renal medullary NO
production. Second, one of the important functions of medullary NO
production is related to the counterregulatory actions of this NO in
response to stimulation by vasoconstrictor stimuli. The rise of
medullary NO stimulated by these vasoconstrictor compounds
significantly buffers both the reduction of MBF and tissue
PO2.
| |
RENAL MEDULLARY NO PROTECTS AGAINST CHRONIC ANG II-, NE-, AND
AVP-INDUCED HYPERTENSION |
On the basis of the acute observations described above, a series
of studies was carried out in chronically instrumented rats to
determine whether the stimulation of NO by vasoconstrictor compounds
would buffer or prevent chronic reductions of MBF and hypertension.
Studies determined whether selective blunting of NOS activity in the
renal medulla would result in a hypersensitivity to small elevations of
circulating vasoconstrictor compounds and consequently result in
hypertension. A dose of L-NAME (75 µg · kg1 · h1)
was used that did not reduce basal MBF >5% when administered into the
medullary interstitial space chronically for 2 wk (118). The same dose of L-NAME that significantly attenuated acute
increases of medullary [NO] produced by intravenous infusions of
subpressor amounts of ANG II, NE, and AVP (116-118).
Doses of ANG II (3.0 ng · kg1 · min1),
NE (0.1 µg · kg1 · min1),
and AVP (2.0 ng · kg1 · min1)
were determined that when administered intravenously did not result in
elevations of mean arterial pressure >5 mmHg when infused for 5-7
days to normal Sprague-Dawley rats (see Fig.
7). Each compound was then infused
intravenously, whereas the nonhypertensive dose of L-NAME
was infused chronically into the medullary interstitium to blunt
medullary NOS activity.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 7.
Influence of NO blockade by L-NAME delivered
chronically into the renal interstitium on the blood pressure response
to vasoconstrictors given intravenously. MAP was normalized to allow
comparison between the individual studies using nonpressor doses of ANG
II (116), NE (118), and arginine vasopressin
(117). , Saline ri; ,
L-NAME ri.
|
|
In the normal state (e.g., an absence of medullary L-NAME
infusion), AVP infusion displayed a well-known response pattern with an
initial rise of arterial pressure for the first several days of
intravenous infusion and then returning to normal (19, 92). Chronic hypertension had never been observed previously with chronic elevations of plasma AVP in rats (19, 92),
humans (48), or dogs (114). However, as seen
in Fig. 7, in rats that received the subpressor chronic medullary
infusion of L-NAME, AVP led to a sustained elevation of
blood pressure (117). Similar studies were carried out to
determine the responses to chronic intravenous infusions of subpressor
doses of ANG II (116) and NE (118). Also
shown in Fig. 7, hypertension occurred associated with sustained
reductions of MBF (not shown) only when medullary NOS activity was
blunted by the medullary interstitial infusion of L-NAME
(116, 117).
The medullary circulation of rats therefore appears to be well
protected from both the acute and chronic vasoconstrictor actions of
small increases in circulating levels of ANG II, NE, and AVP. Although
many of the details of these mechanisms remain to be elucidated, these
studies show that NO-mediated responses play a key role in protecting
the renal medulla from excessive vasoconstriction and chronic hypertension.
Other medullary systems may also be involved in protecting this region
from underperfusion. On the basis of acute studies in anesthetized
rats, one of the other medullary counterregulatory systems may be
prostaglandin release. We have observed that rats pretreated with a
cyclooxygenase inhibitor (meclofenemate) exhibited a significant
reduction of medullary flow with subpressor intravenous infusion of ANG
II (61). Similar responses have been observed with kinin
receptor antagonists (24, 99). It has now also been found
that the hemoxygenase enzymes are present at significantly greater
concentrations in the renal medulla than in the cortex, and the
production of medullary CO via the hemoxygenase pathway may
participate in the control of flow to this region (134). It appears therefore that the renal medulla contains a number of
redundant systems that could serve to protect blood flow to this
region. However, only the NO system at this time has been sufficiently
studied to conclude that it indeed represents a feedback control system
for the long-term regulation of blood flow to the renal medulla.
The NO counterregulatory responses are of seemingly great importance in
maintaining an adequate blood flow to the outer medulla. The delivery
of oxygen to the mTAL is dependent on this blood flow. There is a high
rate of O2 utilization due to the high metabolic ion
transport activity of the thick ascending limbs of Henle that absorbs
nearly 25% of the filtered sodium load. Blood flow to the renal cortex
per gram of tissue is substantially higher in the cortex (700 ml · min1 · 100 g1) compared with the inner medulla (100 ml · min1 · 100 g1) (9); the hematocrit of vasa recta blood
is less than half (26%) of that of arterial blood (101).
The countercurrent flow in the vasa recta and shunting of
O2 between descending and ascending vasa recta account in
part for the falling gradient of PO2 observed from cortex to the tip of the papilla (6, 10, 12). These features result in the observed PO2 levels of
nearly 45 mmHg in the outer medulla falling progressively to <10 mmHg
in the inner medulla and tip of papilla (10, 11, 138).
This contributes to the cells in the deepest zone of the outer medulla
receiving less oxygen, making them more susceptible to ischemic
injury than cells situated in the cortex (12). The
vulnerability of the medulla to underperfusion is expressed by
tubulointerstitial injuries in the juxtamedullary region seen early in
the development of hypertension in rats (22, 26, 38, 94,
131) and humans (7).
| |
GENETIC RAT MODEL OF REDUCED MEDULLARY NO PRODUCTION: THE DAHL
SS RAT |
It has been found that salt-sensitive Dahl rats (SS) exhibit a
reduced capacity to generate NO within the renal medulla. The outer
medulla NOS enzyme activity was found to be nearly one-third less in SS
rats than in normotensive salt-resistant BN rats, although NOS activity
was nearly the same in the inner medulla (119; Fig. 8B). Consistent with these
findings, the mRNA and protein for each of the three NOS isoforms was
also less in the outer medulla of SS rats compared with BN control rats
(Fig. 8, A and C; 119, 132). Most importantly,
although basal medullary [NO] was not significantly lower in SS
compared with BN rats, the ability of ANG II to stimulate an increase
of medullary NO was greatly blunted in SS rats as shown in Fig.
8D (119). These observations predict that the
renal medullary circulation of the SS inbred rat would be
hypersensitive to small elevations of circulating vasoconstrictor agents due to blunting of the NO counterregulatory system.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 8.
Summary of studies demonstrating the reduced capacity of the Dahl
salt-sensitive (SS) rat to generate NO within the outer medulla
compared with the Brown Norway (BN). A: significantly
reduced mRNA expression (determined by Northern blot) of the 2 detectable NOS isoforms in SS rats compared with the BN rats
(132). NOS activity in the outer medulla of SS rats is
also significantly decreased as depicted in B
(119). C: results of the Western blot analysis
of protein expression of the 3 NOS isoforms in which all 3 isoforms
were significantly lower in the SS rat compared with BN rat
(119). D: significant increase in the
concentration of NO in the outer medulla after a 1-h intravenous
infusion of ANG II in the BN rat, with the SS rat showing no change
from control (119). Control [NO] for SS rat was 89 ± 3 and 87 ± 3 nmol/l for the BN rat. * Significant
difference between the SS and BN in A-C. * Significant
difference from the pre-ANG II control in D.
|
|
Studies were carried out to determine if SS rats resembled
Sprague-Dawley rats, in which medullary NOS activity was blunted by
selective delivery of L-NAME chronically into the medullary interstitium. As summarized in Fig. 9A,
chronic intravenous infusion of ANG
II (3.0 ng · kg1 · min1)
failed to induce a rise of mean arterial pressure in the
salt-insensitive BN rats as had been found earlier in Sprague-Dawley
rats (119). However, this same dose of systemically
delivered ANG II resulted in sustained hypertension in SS rats. It is
important to recognize that both the SS and BN rats were maintained on
a low (0.4%) salt intake throughout the study. These results were
consistent with the acute responses of MBF to low-dose acute
intravenous ANG II infusions (5.0 ng · kg1 · min1).
That is, ANG II significantly reduced MBF in SS rats, but not in BN
rats (119). The ability of the BN rats to buffer ANG
II-induced reductions of MBF but not in SS rats was also consistent
with observations that ANG II induced a significant increase of
medullary interstitial [NO] of BN rats, but failed to do so in SS
rats (119).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 9.
Change in MAP in SS and BN rats is shown after
chronic intravenous infusions of a nonpressor dose of ANG II
(119) or AVP (132). Control pressures were
normalized (presented as  %) to better contrast the results from the
4 groups studied.
|
|
Similar studies were carried out with chronic intravenous infusion of
AVP (132). As shown in Fig. 9B, mean arterial
pressure in SS rats maintained on a low (0.4%)-salt diet increased
from a control of 127 mmHg to an average of 147 mmHg after 14 days of
AVP infusion (2.0 ng · kg1 · min1).
These were remarkable observations, because it was the first time that
AVP produced sustained hypertension without pharmacological manipulations or surgical reductions of renal mass (54).
Northern blot analysis of renal outer medullary tissue in this study
demonstrated that SS rats maintained on a low (0.4%)-salt diet
expressed NOS I and NOS II mRNA levels that were significantly less
than that found in BN rats maintained on the same low salt intake (132; Fig. 8A).
Taken together, the results of studies in both Sprague-Dawley rats
receiving a medullary infusion of a low dose of L-NAME and
in SS rats indicate that an impaired NO counterregulatory system in the
renal medulla results in reduction of blood flow to this region and an
increased susceptibility to develop hypertension in face of small
elevations of circulating vasoconstrictor compounds such as ANG II,
AVP, and NE.
| |
EVIDENCE THAT L-ARG UPTAKE IN THE RENAL MEDULLA
PARTICIPATES IN THE REGULATION OF NO PRODUCTION |
Both oral and intravenous administration of the NOS substrate
L-Arg in large amounts have been shown to be capable of
preventing salt-induced hypertension in SS rats (15, 32, 90,
91). It has also been found in anesthetized rats that
pressure-natriuresis and transmission of perfusion pressure into the
renal interstitium were normalized by long-term L-Arg
treatment (90, 91). These observations led us to
hypothesize that the major antihypertensive actions of
L-Arg in SS rats could be mediated largely through actions
of this NO precursor, specifically within the renal medulla. A
threshold dose of L-Arg was found that when infused into
the medullary interstitium increased renal MBF without altering CBF in
anesthetized SS rats. This dose (300 µg · kg1 · min1)
was then infused chronically into the renal medullary interstitium of
SS rats to determine the effects of this regional L-Arg
administration on salt-induced hypertension in these rats (65,
66). Remarkably, as daily salt diet was increased from 0.4% to
4.0%, mean arterial pressure remained nearly unchanged over 1 wk in SS
rats receiving the medullary infusion of L-Arg (Fig.
10). In contrast, in SS rats receiving
the saline vehicle control infusion, mean arterial pressure increased
from a control level of 120 to 165 mmHg over the same time period. The
same dose of medullary infused L-Arg that prevented salt-induced hypertension in SS rats failed to blunt the development of
hypertension when administered intravenously to SS rats. It was found
that SS rats receiving the high-salt diet exhibited a significant
reduction of MBF (65). This reduction of flow to the renal
medulla of the SS rat was prevented by chronic medullary interstitial
delivery of L-Arg with no rise of arterial pressure observed. These data, as summarized in Fig. 10, indicated that the
prevention of salt-sensitivity in SS rats was due specifically to the
action of L-Arg on the renal medullary function and that SS
rats may have a deficit of medullary substrate availability that could
account in part for the observed reduction in NO production in this
region of the kidney.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 10.
effects of renal medullary interstitial infusion of
L-Arg (300 µg/kg/min) in SS rats on high-salt diet
induced changes in MAP, CBF, and MBF. SS rats infused with
L-Arg ri are indicated by . SS rats not infused with
L-Arg but undergoing the same change in diet are indicated
by . [These data are regraphed from the original manuscript by
Miyata and Cowley (65).]
|
|
Studies carried out to characterize the pathway of cellular uptake of
L-Arg determined the influence of cellular
L-Arg uptake on NOS enzymatic activity in IMCD cells from
Sprague-Dawley rat kidneys (129). It was found that
L-Arg uptake was predominantly mediated by a high-affinity
y+ transporter system, a sodium-independent, trans-stimulated cationic transporter encoded by the CAT 1 gene. The y+ transporter system consists of a family of four genes, designated as CAT1, CAT2A, CAT2B, and CAT3, but the only transporter found in IMCD was CAT1. Biochemical transport studies using
L-[3H]arginine found that the
L-Arg transport system was saturable and inhibited by the
c
TOP
ABSTRACT
INTRODUCTION
