点击显示 收起
【摘要】 Multiphoton fluorescence microscopy offers the advantages of deep optical sectioning of living tissue with minimal phototoxicity and high optical resolution. More importantly, dynamic processes and multiple functions of an intact organ can be visualized in real time using noninvasive methods, and quantified. These studies aimed to extend existing methods of multiphoton fluorescence imaging to directly observe and quantify basic physiological parameters of the kidney including glomerular filtration rate (GFR) and permeability, blood flow, urinary concentration/dilution, renin content and release, as well as more integrated and complex functions like the tubuloglomerular feedback (TGF)-mediated oscillations in glomerular filtration and tubular flow. Streptozotocin-induced diabetes significantly increased single-nephron GFR (SNGFR) from 32.4 ± 0.4 to 59.5 ± 2.5 nl/min and glomerular permeability to a 70-kDa fluorophore approximately eightfold. The loop diuretic furosemide 2-fold diluted and increased 10-fold the volume of distal tubular fluid, while also causing the release of 20% of juxtaglomerular renin content. Significantly higher speeds of individual red blood cells were measured in intraglomerular capillaries (16.7 ± 0.4 mm/s) compared with peritubular vessels (4.7 ± 0.2 mm/s). Regular periods of glomerular contraction-relaxation were observed, resulting in oscillations of filtration and tubular flow rate. Oscillations in proximal and distal tubular flow showed similar cycle times ( 45 s) to glomerular filtration, with a delay of 5-10 and 25-30 s, respectively. These innovative technologies provide the most complex, immediate, and dynamic portrayal of renal function, clearly depicting the components and mechanisms involved in normal physiology and pathophysiology.
【关键词】 intravital microscopy diabetic nephropathy singlenephron glomerular filtration rate red blood cell velocity quinacrine rhodamine lucifer yellow proteinuria
RECENT ADVANCES HAVE ADDED microscopic investigative techniques to the arsenal of tools applicable to gaining a more comprehensive understanding and quantifiable benchtop assessment of renal function. The development of micropuncture by Wearn and Richards ( 29 ) in 1924 was one of the most significant advances in renal physiology, leading to the discovery of the phenomena of filtration, absorption, and secretion in the nephron ( 24 ). The localization of these processes and the determination of the handling of specific substances were deduced by comparing fluid compositions in the bladder, kidney, and blood ( 29 ). Eventually, manipulation of renal physiological functions was made possible with the development of the isolated perfused tubule by Burg and colleagues ( 3 ) in 1966. Micropuncture and microperfusion techniques can be used to study the end results of renal physiological function, but they do not permit direct visualization of ongoing mechanisms or other details of the pathways. Renewed interest in employing conventional methods to study nephron function in genetically altered animal models has prompted innovations like the use of FITC-labeled inulin, rather than radioactive probes, to evaluate single-nephron glomerular filtration rate (SNGFR) and tubular reabsorption ( 16 ). The application of fluorescence microscopy to determine the FITC-inulin signal in micropuncture studies has the advantages of simplicity, precision, accuracy, and cost effectiveness without consumption of the fluid sample ( 16 ). These long-established, demanding, and invasive methods provide critical snapshots of renal physiology and have directed countless current methods to discover precise details about the players, interactions, and control mechanisms involved.
Insight into the interplay between different parts of the nephron provides a necessary order of complexity to evaluate and have a comprehensive understanding of renal function. Many critical physiological processes in the kidney like the regulation of glomerular filtration, hemodynamics, concentration, and dilution involve complex interactions between multiple cell types and customarily inaccessible structures. For example, rat experiments in the 1980s revealed that variables of nephron flow exhibited tubuloglomerular feedback (TGF)-mediated regular oscillations ( 11, 15 ). Spontaneously hypertensive rats (SHR) have been characterized as displaying irregular TGF-mediated oscillations ( 11, 31 ), and recent studies have attempted to distinguish possible factors that may contribute to the spectral complexity of the observed oscillations ( 14 ). The existence of TGF-mediated oscillations in nephron flow serves as one example of a complex functional parameter that could be better examined and also quantified by microscopy.
Multiphoton excitation fluorescence microscopy offers a state-of-the-art imaging technique superior for deep optical sectioning of living tissue samples. The higher resolution and minimal phototoxicity of this method permit longer time periods of continuous tissue scanning with uses in real-time imaging of intact organs. Using this technique, dynamic processes such as glomerular filtration ( 9, 32 ), proximal tubule endocytosis ( 23 ), apoptosis ( 9 ), microvascular function ( 9, 18 ), protein expression ( 27 ), renal cysts ( 26 ), and major functions of the juxtaglomerular apparatus, including TGF ( 20, 22 ) and renin release ( 21, 22 ), have been visualized and studied both in vivo and in vitro down to the subcellular level. The capacity to simultaneously visualize both proximal and distal segments of the nephron permits observation of the dynamic processes within the living kidney and a quantitative assessment of the various operations. In fact, a ratiometric intravital two-photon microscopic technique based on the generalized polarity concept has been recently applied to quantify glomerular filtration and tubular reabsorption ( 32 ). The rapidly developing field of fluorescence optics and ultrasensitive detection will fuel further developments. For example, the construction of novel photonic crystal fibers, and hence the advent of multiphoton fluorescence endoscopy ( 1 ), demonstrates the potential of this technology for developing noninvasive therapeutic and real-time diagnostic tools for both clinical and biomedical research applications ( 33 ). Consequently, one of the next steps for multiphoton microscopy is to provide real-time, quantitative imaging and rapid evaluation of basic organ functions.
The aim of the present study was to extend existing methods of multiphoton fluorescence imaging to directly observe and quantify basic functional parameters of the kidney. These include the noninvasive measurement of glomerular filtration rate, blood flow, urinary concentration/dilution through the course of the nephron, and renin content and release as well as more integrated and complex functions like TGF-mediated oscillations in filtration and tubular flow. These innovative technologies provide the most complex, immediate, and dynamic portrayal of renal function, clearly depicting and analyzing the components and mechanisms involved in normal physiology and pathophysiology.
MATERIALS AND METHODS
Multiphoton excitation laser-scanning fluorescence microscopy. The multiphoton microscope used in these studies consists of a Leica TCS SP2 AOBS MP confocal microscope system (Leica Microsystems, Heidelberg, Germany). A Leica DM IRE2 inverted microscope was powered by a wide-band, fully automated, infrared (710-920 nm) combined photo-diode pump laser and mode-locked titanium:sapphire laser (Mai-Tai, Spectra-Physics, Mountain View, CA). Images were collected in a time ( xyt, 20 Hz) or line ( xt, 1,000 Hz) series, depending on the purpose of study, with Leica confocal software (LCS 2.61.1537) and analyzed with the LCS 3D, Process, and Quantify packages. The principles and advantages of conventional confocal and two-photon microscopy and their application in imaging renal tissues have recently been reviewed ( 18, 22 ).
Three easy-to-use, water-soluble fluorophores were used to label specific structures in the living kidney. A 70-kDa dextran-rhodamine B conjugate was used (100 µl of a 10 mg/ml stock in iv bolus, Invitrogen) to label the circulating plasma or intravascular space. Tubular segments and, more specifically, the content of individual renin granules were visualized using quinacrine (100 µl of a 25 mg/ml stock in iv bolus, Sigma) in a manner similar to in vitro applications previously described ( 21, 22 ). In some experiments, the extracellular fluid marker lucifer yellow (LY) was used (10 µl of a 10 mg/ml stock in iv bolus, Invitrogen). Because the ionic charge of fluorophores affects glomerular filtration characteristics, the FITC-conjugate of the gold-standard glomerular filtration rate (GFR) marker inulin (5 kDa), and the similar neutral rhodamine B-dextran (70 kDa) were used. Although the fluid-marker LY is an anionic compound, it is freely filtered due to its small size (0.45 kDa). All three fluorescent probes were excited using the same, single excitation wavelength of 860 nm (Mai-Tai), and the emitted, nondescanned fluorescent light was detected by two external photomultipliers (green and red channels) with the help of a FITC/TRITC filter block (Leica).
Animals. Munich-Wistar male rats (200 g, Harlan, Madison, WI) and C57/BL6 mice (20 g, inbred) were anesthetized with thiobutabarbital (Inactin, 130 mg/kg body wt) alone (for rats) or in combination with 50 mg/kg body wt ketamine (for mice). After adequate anesthesia was ensured, the trachea was cannulated to facilitate breathing. The left femoral vein and artery were cannulated for dye infusion and blood pressure measurements, respectively. Subsequently, a 10- to 15-mm dorsal incision was made under sterile conditions and the kidney was exteriorized. The animal was placed on the stage of an inverted microscope with the exposed kidney placed in a coverslip-bottomed heated chamber bathed in normal saline, and the kidney was visualized from below as described by Dunn et al. ( 9 ) using a HCX PL APO 63X/1.4NA oil CS objective (Leica). High-quality images from the renal cortex were acquired up to 150 µm deep below the surface. During all procedures and imaging, core body temperature was maintained with a homeothermic table. All animal protocols were approved by the Institutional Animal Care and Use Committee at the University of Southern California.
In some rats, diabetes was induced by using a single dose of streptozotocin (STZ; 50 mg/kg ip). Blood glucose levels were measured following STZ administration by using test strips on blood samples from a tail clip (Freestyle blood glucose monitoring system, Abbott Laboratories) to verify induction of diabetes. Two groups of animals with blood glucose levels of 400 mg/ml or higher were used: one between days 4 and 6 after STZ injection for SNGFR calculations, and the other within 4 wk for glomerular permeability studies.
During in vivo imaging, the systemic blood pressure of animals was monitored through a cannula inserted into the left femoral artery and the use of an analog single-channel transducer signal conditioner (model BP-1, transducer model BLPR, World Precision Instruments, Sarasota, FL). Calibration was performed using a pressure manometer (model PM-015), and data were collected using data acquisition system QUAD-161.
Chemicals, if not indicated, were purchased from Sigma (St. Louis, MO).
GFR. Overall GFR was measured using the fluorescence-based FITC-inulin method as described before ( 16 ). Briefly, anesthetized rats were surgically instrumented for clearance measurements which included tracheostomy, cannulation of left femoral artery and vein as described above, as well as the two ureters using a 24-gauge intravenous (iv) catheter (Terumo). Immediately after surgery, animals were given a bolus (2 µl/g body wt) of PBS containing 0.05% FITC-inulin and 3.5% BSA. This was followed by a maintenance infusion of the same solution at 50 µl/min, and a 30-min equilibration period. Renal function was then determined over three consecutive 15-min clearance periods. At the midpoint of each urine collection, an arterial blood sample (100 µl) was obtained for determination of plasma FITC-inulin. Plasma and urine samples were diluted 1:100 in HEPES buffer, and FITC fluorescence intensity was measured at 540 nm in response to excitation at 485 nm in a cuvette-based fluorometer (Quantamaster-8, PTI). FITC-inulin concentrations were determined using FITC standards ( 16 ).
Data analysis. Data are expressed as means ± SE. Statistical significance was tested using ANOVA. Significance was accepted at P < 0.05.
RESULTS
Measurement of SNGFR. An example of the experimental technique is shown in Fig. 1. A superficial glomerulus was selected that had at least a 100-µm-long initial segment of the proximal tubule positioned in the same optical section. A single iv bolus of the fluid marker LY was injected into the femoral vein, which appeared in the glomerulus within 5 s and filtered into Bowman?s space and early proximal tubule (see supplementary video file showing glomerular filtration of LY, as well as all supplemental data for this article in the online version of this article). Using high temporal resolution, the fluorescence intensity changes of LY were measured as shown in Fig. 1, A - C, within two regions of interest (ROI): one at the opening of proximal tubule (ROI1), the other 100 µm downstream (ROI2). Internal diameter, length of the tubule, and transit time of filtrate (time shift shown in Fig. 1 C ) between the two areas were measured using the Quantify package of the Leica confocal software. The midpoint of the dye bolus, approximated by the maximal fluorescence intensity, travels at the same speed as the mean fluid velocity. Thus the transit time (shift between ROI1 and ROI2 intensity plots) was calculated at the peak ( Fig. 1 C ). By calculating tubular fluid volume [length x (diameter/2) 2 x ] the absolute value of SNGFR was calculated (volume/time). Figure 1 D demonstrates that SNGFR measurements provided equivalent values if the initial <500-µm segment of proximal tubule was used. There was no statistically significant correlation between SNGFR values and distance of ROIs ( r = 0.07). Also, SNGFR values obtained in the same nephrons using LY (32.4 ± 0.4 nl/min, n = 50) or the gold-standard FITC-inulin (34.2 ± 0.8 nl/min, n = 12) were not statistically different.
Fig. 1. Representative images of the technique to measure single-nephron glomerular filtration rate (SNGFR). The intravascular space is labeled with 70-kDa dextran-rhodamine B (red), renal tubules with quinacrine (green), and the glomerular filtrate (tubular fluid) with lucifer yellow (LY; yellow). Glomerulus (G) and proximal tubule (PT) are labeled. Fluorescence intensity of intravenously injected LY was measured within 2 areas of interest (ROI) in the early PT fluid: one always at the beginning of the PT (ROI1), the other 100 µm downstream (ROI2). A and B : downstream movement of filtered LY in tubular fluid. C : plot of LY fluorescence intensity changes within the 2 ROIs. The shift ( t ) indicates the duration of fluid movement from ROI1 to ROI2. Scale is 50 µm. D : effects of positioning ROI2 at different distances from ROI1 on SNGFR calculations ( n = 18).
To demonstrate the utility of this technique, SNGFR was measured in control and STZ-diabetic rats. SNGFR in control rats averaged 32.4 ± 0.4 nl/min ( n = 50 from 7 different animals), whereas significantly elevated levels (59.5 ± 2.5 nl/min, n = 15 from 7 different animals) were measured in select, significantly enlarged glomeruli from diabetic rats. Overall GFR measured with the FITC-inulin clearance method appeared to be higher in diabetes, but there was no statistically significant difference between the control (1.13 ± 0.13 ml/min, n = 6) and diabetic groups (1.34 ± 0.13 ml/min, n = 5). Systemic blood pressure was normal in all animals studied; the mean arterial blood pressure was 102.7 ± 8.6 mmHg.
Estimating renal blood flow. Red blood cell (RBC) velocity as an index of renal blood flow was measured in peritubular and intraglomerular capillaries by expanding recently established techniques ( 13, 19, 33 ). Peritubular capillary blood flow is shown in real time in a supplementary video. Using sufficiently high temporal resolution (1 ms), it was possible to image and characterize the motion of RBCs in the renal cortex. As shown in Fig. 2 A, RBCs exclude the fluorescent dye used to label the circulating plasma and consequently they appear as dark, nonfluorescent objects on the images. With the acquisition of repetitive scans along the central axis of a capillary (called a line-scan), the motion of RBCs leaves dark bands in the data set ( Fig. 2, A and B ). Importantly, the slope of the bands is inversely proportional to the velocity, which was measured as shown in Fig. 2 B using the LCS software. In most peritubular capillaries a regular, steady flow of RBCs was measured, as indicated by the constant slope of the bands ( Fig. 2, B and C ). Significantly higher speeds of RBCs were measured in intraglomerular capillaries (16.7 ± 0.4 mm/s) compared with peritubular vessels (4.7 ± 0.2 mm/s, n = 10 each from 8 different animals) with similar diameters. In addition, we observed instances of irregular flow in some peritubular capillaries in which RBC velocity showed fluctuations ( Fig. 2 D ). Systemic blood pressure was within the normal range throughout these experiments.
Fig. 2. Illustration of red blood cell (RBC) flow and velocity measurement in renal capillaries. A : circulating plasma was labeled with 70-kDa dextran-rhodamine B (red) and proximal (PT) and distal (DT) renal tubules with quinacrine (green). Note that some rhodamine was filtered through the glomerulus and became concentrated in the DT as indicated by the red color of tubular fluid. A horizontal line, 20 µm long, was positioned through the central axis of a capillary (white line in the top left corner), and a line ( xt ) scan was performed at 1-ms intervals along this line continuously for 2 s. RBCs appear as unstained dark objects ( A ) and leave dark bands on the line scan ( B - D ) as they move through the capillary (line). Xt line scans of a peritubular ( B ) and intraglomerular ( C ) capillary are shown. RBC velocity as the slope of the bands x / t was calculated as shown. Note the regular blood flow (constant slope) in B and C and higher RBC speed (smaller slope) in C vs. B. D : line scan of irregular blood flow in a peritubular capillary, indicated by the variable slope.
Imaging urinary concentration and dilution. We applied multiphoton microscopy to provide a continuous, real-time measurement of the urinary concentrating and diluting function of individual renal tubules based on recent work ( 26 ). The technique is demonstrated in Fig. 3. A single iv bolus of a relatively high-molecular-weight fluorescent plasma marker (rhodamine-conjugated 70-kDa dextran) remained in the intravascular space for several hours of imaging due to its large size. Steady-state levels of the marker in the plasma were evident by the constant level of fluorescence in the renal vasculature throughout these experiments ( Figs. 1 - 6, supplementary videos). A microscopic fraction of the marker was filtered in the glomeruli, and thus only trace amounts of rhodamine were detected in the proximal tubular fluid (see dark proximal tubule fluid in Fig. 3 A ). However, due to the concentrating mechanism, an approximately sevenfold higher level of rhodamine fluorescence was present in the collecting duct system ( Fig. 3, A and D ). Subsequently, the loop diuretic furosemide was used to demonstrate the capabilities of this quantitative imaging technique. Effects of a single iv bolus of 2 mg/kg furosemide on the morphology and function of the renal cortex are shown in Fig. 3, B - D. Within 2 min of administration, furosemide caused the enlargement of distal nephron segments ( Fig. 3 B ). At the peak of furosemide?s effect, the volume of tubular fluid in the cortical collecting tubule increased 10-fold compared with control. Obstruction of peritubular capillaries by the enlarged distal nephron segments was evident ( Fig. 3 C ). Compared with the magnitude of distal nephron enlargement, no significant morphological changes were observed in proximal tubule segments ( Fig. 3 ). Simultaneously with tubular enlargement, furosemide reduced urinary concentration 50% as measured by the distal/proximal ratio of fluorescence intensity of rhodamine in the tubular fluid ( Fig. 3 D ).
Fig. 3. Visualization of the concentrating and diluting function of the kidney. A : circulating plasma was labeled with 70-kDa dextran-rhodamine B (red) and PT and DT (*) with quinacrine (green). Note that some rhodamine-dextran was filtered through the glomerulus (G) and became concentrated in the DT vs. PT as indicated by the intense red color of tubular fluid. B and C : effects of the loop diuretic furosemide. Note the reduced red color and increased volume of DT fluid. D : reduction in the ratio of DT/PT fluid rhodamine B fluorescence in response to furosemide, injected intravenously at time 0. Scale is 20 µm.
Fig. 6. In vivo imaging of glomerular permeability in the diabetic kidney. Sclerotic glomeruli (arrows) and a hyperfiltering glomerulus (G) are shown. Note the intense ultrafiltration of the high-molecular-weight (70 kDa) dextran-rhodamine B (red) from the plasma into Bowman?s space in the sclerotic, but not in the hyperfiltering glomerulus.
Physiological oscillations in SNGFR and tubular flow. In the intact kidney, the closed loop system allows for the observation of myogenic and TGF-induced oscillations in GFR and, consequently, changes in tubular flow. The renal cortex was visualized continuously for 5 min, as illustrated in Fig. 4. A video of the same preparation is available at http://ajprenal.physiology.org/cgi/content/full/DC1. Regular periods of glomerular contraction-relaxation were observed, resulting in oscillations of filtration and tubular flow rate ( Fig. 4 B ). This was measured by the changes in fluorescence intensity of the filtered plasma marker, 70-kDa dextran-rhodamine B, in the tubular (Bowman?s space) fluid as well as by changes in tubular diameter (not shown; see supplementary video). Oscillations clearly showed two components: a faster (cycle time 10 s) and a slower (cycle time 45 s) mechanism. Tubular flow oscillations were delayed compared with oscillations in glomerular filtration: an 5- to 10-s delay was detected in the proximal tubule, whereas the delay was measured to be 25-30 s in distal nephron segments ( Fig. 4 B ). Glomerular and tubular flow oscillations were absent after furosemide treatment (not shown).
Fig. 4. Fluorescence labeling of the intact living rat kidney. A : rhodamine B-conjugated 70-kDa dextran (red) was given intravenously to label the cortical vasculature (plasma). Quinacrine (green) stained renal tubules [PT and cortical collecting duct (CCD)]. A small fraction of plasma dextran is filtered in the glomerulus (G) and appears in the concentrated tubular fluid of the CCD (red) more or less intensely depending on the high or low phase of flow oscillations (see supplementary video). B : spontaneous oscillations in glomerular filtration and PT and DT flow detected by changes in rhodamine B fluorescence in the tubular fluid (or Bowman?s space). Magnitude of changes are not drawn to scale. Simultaneous measurements are shown (vertical dotted line represents time 0 ), so the delay in PT and CCD oscillations compared with those in the parent glomerulus (Glom) is easy to read. Scale is 20 µm.
Renin content and release. Renin content of juxtaglomerular granular cells was visualized in vivo in the rat ( Fig. 5 A ) and mouse ( Fig. 5 B ) using quinacrine, a selective marker of acidic and secretory granules. Quinacrine staining of individual renin granules in the terminal afferent arteriole was highly specific. However, weak background staining was observed in all tubule cell types, most likely due to the accumulation of quinacrine in lysosomes. In addition to the quantification of renin content (based on the length of the renin-positive segment of afferent arteriole), we measured the dynamics of renin release. Furosemide was used to trigger juxtaglomerular renin release. Figure 5 C shows a representative graph of four separate experiments. Minutes after iv bolus administration of 2 mg/kg furosemide, the renin content of the afferent arteriole was reduced by 20%, as measured by the reduction in quinacrine fluorescence ( Fig. 5 C ).
Fig. 5. In vivo imaging of the juxtaglomerular renin content in the rat ( A ) and mouse ( B ). Circulating plasma is labeled with rhodamine B-conjugated 70-kDa dextran (red), content of individual renin granules with quinacrine (green), and nuclei with Hoechst 33342 (blue). Note the significant renin content around the juxtaglomerular portion of the afferent (AA), but not the efferent arteriole (EA). G, glomerulus. Scale is 20 µm. C : reduction in quinacrine fluorescence intensity (index of renin release) in response to furosemide, injected intravenously at time 0.
Glomerular permeability. In vivo real-time imaging allowed us to observe functional differences between hyperfiltering and sclerotic glomeruli in the early phase of STZ-diabetes (within 4 wk after STZ injection). Sclerotic glomeruli were evident by their reduced size and the small number of perfusing glomerular capillary loops ( Fig. 6 ). More importantly, increased glomerular permeability was readily apparent in the sclerotic glomeruli ( 20% of all glomeruli), whereas rhodamine leakage was significantly less noticeable in the hyperfiltering, enlarged glomeruli as illustrated in Fig. 6. The behavior of a 70-kDa dextran-rhodamine conjugate in the vascular space was compared. This large molecule (comparable to the size of albumin) was more freely filtered into Bowman?s space in the sclerotic than in the hyperfiltering glomeruli. The ratio of Bowman?s space/intravascular rhodamine fluorescence was 0.1 ± 0.01 in hyperfiltering glomeruli, whereas it was 0.8 ± 0.1 in sclerotic glomeruli ( n = 8 each, P < 0.05). This finding strongly suggests that the significant proteinuria present in diabetic nephropathy is predominantly attributable to the sclerotic rather than the hyperfiltering glomeruli. In most sclerotic glomeruli, the free downstream passage of the "red" proximal tubular fluid was persistent and characteristic. Furthermore, LY given in iv bolus (as shown in Fig. 1 ) still cleared from Bowman?s space and proximal tubule, evidence that some blood supply and glomerular filtration still remained.
DISCUSSION
Multiphoton microscopy is a state-of-the-art imaging technique which can provide high-resolution images from deep optical sections of the living tissue, including the kidney ( 18 ). However, multiphoton microscopy has applications far beyond the generation of superior images: dynamic processes, like the multiple functions of an intact organ, can be visualized by real-time imaging in a noninvasive way and quantified. This study provides further examples of how basic parameters of renal (patho)physiology can be continuously monitored in the intact rat kidney.
In the pioneering work of Molitoris et al. ( 9, 18, 32 ), the process of glomerular filtration has been recently observed in real-time using intravital two-photon microscopy and a number of fluorescent probes ( 9, 32 ). The presently described methods to quantify the rate of glomerular filtration are based on previous fluorescent techniques that measured SNGFR using conventional micropuncture ( 16 ) and tubular flow rate with a nonobstructing optical method ( 7 ). In our experience, LY (a widely used extracellular fluid marker) ( 8 ) was a better tool to measure SNGFR with this optical technique than FITC-inulin due to its low molecular weight, excellent water solubility, and high fluorescent quantum yield. It should be mentioned, however, that LY is not a good GFR marker in the classic physiological sense (i.e., for clearance studies) because its clearance is most likely greater than that of inulin. Compared with the demanding conventional micropuncture methods ( 2, 16, 28 ), a single iv bolus of LY was filtered into the glomeruli and provided a convenient measure of SNGFR within 5 s. Another advantage of this noninvasive technique is that tubular flow and macula densa functions remain intact and undisrupted. Tubular dimensions and the transit time of LY in the early proximal tubule were easily measured using the LCS software ( Fig. 1 C ) from which SNGFR was calculated. Combined with monitoring tubular flow rate ( Fig. 4 B ), this technique permits real-time measurement of SNGFR in the intact kidney. Taking advantage of the high temporal resolution, it was possible to accurately measure LY transit time within the initial 500 µm of proximal tubule. To validate this technique, glomerular filtration was observed and quantified under normal physiological conditions and in diabetic hyperglycemia. Consistent with the well-established glomerular hyperfiltration in diabetes ( 4, 28 ), we measured significantly elevated SNGFR levels in select, significantly enlarged glomeruli in STZ-treated animals. The values of SNGFR are within the same range as measured by micropuncture techniques in both control and diabetes ( 2, 28 ).
In vivo imaging methods permit the observation and measurement of regional differences in blood flow to the nephron. Cortical blood flow was evaluated by measuring RBC velocity in peritubular and intraglomerular capillaries. Even with the resistance of the afferent arteriole, glomerular capillary pressure is high and the postglomerular resistance produces a significant pressure drop from the glomerular to peritubular capillaries. Thus as expected, RBC velocity was slower in peritubular capillaries than in glomerular capillaries. The occasionally observed irregular blood flow ( Fig. 2 D ) is probably due to intermittent circulation in branches of the peritubular capillary plexus and the dynamic control of vascular resistance upstream. Consistent with the higher blood flow to the kidneys compared with other organs, the measured RBC speed is higher than in other vascular beds ( 6, 13 ).
As a noninvasive alternative to existing methods, in vivo imaging allows direct and continuous visualization of all cortical segments of the nephron, often in the same visual field. Such a technique thus permits observation of the ways in which upstream changes exert effects downstream. For example, fluorescence intensity of 70-kDa dextran-rhodamine was more pronounced in distal segments of the nephron, illustrating the concentrating mechanisms present there regionally ( Fig. 3 ). Furosemide, one of the most potent loop diuretics via its ability to block the Na-K-2Cl cotransporter, elicited a massive fluid loss on its acute administration ( Fig. 3, B and C ). On treatment with furosemide, the increased distal fluid load inflated collecting ducts and diluted the distal tubular fluid ( Fig. 3 D ). The enlarged distal tubular segments appeared to compress peritubular capillaries consistent with preliminary data on the effect of furosemide to reduce renal blood flow ( 19 ). Further studies with furosemide would hold the potential to visualize morphological and functional changes occurring in the juxtaglomerular apparatus and glomerulus as well, directly responding to recent inquiries into the functional significance of furosemide?s actions on the Na-K-2Cl cotransporter 1 isoform at these sites ( 12 ). Consistent with recent in vivo and in vitro data on the effect of furosemide on renin release ( 5 ), the present studies detected a 20% release of renin from juxtaglomerular afferent arterioles in response to acute furosemide treatment ( Fig. 5 C ).
The TGF system is a key regulator of filtration rate and of water and electrolyte delivery to the distal nephron ( 25 ). Earlier experiments demonstrated that renal blood flow and consequently, tubular fluid flow, exhibit regular oscillations due to the myogenic mechanism (6- to 10-s periods) and TGF (20- to 50-s periods) ( 10, 11, 17 ). Mathematical models as well indicated that these mechanisms are coupled ( 17 ). The present experimental technique allows noninvasive, real-time, and in vivo observation and measurement of oscillations in glomerular filtration and tubular flow simultaneously in both proximal and distal nephron segments ( Fig. 4 B ). Future studies can directly visualize alterations in this system, for example, the irregular oscillations in hypertensive rats ( 11, 14, 31 ) as well as nephron-nephron interactions ( 10 ).
The renin-angiotensin system is one of the most important regulatory mechanisms of renal tubular salt and water conservation, and consequent blood pressure equilibrium. The major structural component of the renin-angiotensin system in the kidney is the juxtaglomerular apparatus located at the vascular pole of the glomerulus. Renin-producing cells of the juxtaglomerular apparatus reside in the wall of the terminal afferent arteriole and can be visualized ( Fig. 5, A and B ) using multiphoton imaging and quinacrine, a fluorescent probe selective for individual renin granules ( 21, 22 ). This model is a direct continuation of our recent in vitro work ( 21 ), and it is now possible to study the dynamics of renin content and release in vivo, with high spatial and temporal resolution down to the individual granule level.
The current understanding of diabetes recognizes the disease as a pathology that begins with hyperfiltration that eventually progresses to loss of filtration function. The proteinuria characteristically associated with the disease is considered telltale evidence of concomitant pathology, but the mechanism of its inception was unknown ( 4 ). Some theories implicated the hyperfunctioning glomerulus as the culprit: the increased blood flow and vascular damage allowed proteins to leak through ( 4 ). Our images and recordings revealed that sclerotizing glomeruli actually leak into Bowman?s space and distal collecting segments, whereas the high-molecular-weight, fluorescence-tagged dextran in the circulating plasma often remains neatly confined within the hyperfiltering glomerular vasculature ( Fig. 6 ). Diabetes has already been implicated in epithelial foot process damage ( 30 ), a pathological milestone that would permit proteinuria from the nonfunctional, and hence nondiscriminate, sclerotic glomeruli may be the source of the proteinuria.
In summary, quantitative imaging of the intact kidney with multiphoton microscopy provides an excellent noninvasive tool for visualizing and studying renal function. With the application of only three fluorescent probes, several basic parameters of renal physiology can be measured in real time, including glomerular filtration and permeability, tubular fluid and blood flow, urinary concentration/dilution, and renin content and release. Further studies using this technique will allow investigations of highly complex and integrative questions about the processes of renal (patho)physiology. Quantitative imaging of kidney function with multiphoton microscopy may eventually provide a novel noninvasive diagnostic tool for future clinical applications.
GRANTS
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-064324 and American Heart Association Grant EIA 0640056N.
ACKNOWLEDGMENTS
The authors thank Donald J. Marsh for expert advice and help with the quantitative analysis of SNGFR.
【参考文献】
Bird D and Gu M. Two-photon fluorescence endoscopy with a mirco-optic scanning head. Opt Lett 28: 1552-1554, 2003.
Brown R, Ollerstam A, and Persson AE. Neuronal nitric oxide synthase inhibition sensitizes the tubuloglomerular feedback mechanism after volume expansion. Kidney Int 65: 1349-1356, 2004.
Burg M, Grantham J, Abramow M, and Orloff J. Preparation and study of fragments of single rabbit nephrons. Am J Physiol 210: 1293-1298, 1966.
Castellino P, Shohat J, and DeFronzo RA. Hyperfiltration and diabetic nephropathy: is it the beginning? Or is it the end? Semin Nephrol 10: 228-241, 1990.
Castrop H, Lorenz JN, Hansen PB, Friis U, Mizel D, Oppermann M, Jensen BL, Briggs J, Skøtt O, and Schnermann J. Contribution of the basolateral isoform of the Na-K-2Cl - cotransporter (NKCC1/BSC2) to renin secretion. Am J Physiol Renal Physiol 289: F1185-F1192, 2005.
Chaigneau E, Oheim M, Audinat E, and Charpak S. Two-photon imaging of capillary blood flow in olfactory bulb glomeruli. Proc Natl Acad Sci USA 100: 13081-13086, 2003.
Chou CL and Marsh DJ. Measurement of flow rate in rat proximal tubules with a nonobstructing optical method. Am J Physiol Renal Fluid Electrolyte Physiol 253: F366-F371, 1987.
Chu J, Chu S, and Montrose MH. Apical Na + /H + exchange near the base of mouse colonic crypts. Am J Physiol Cell Physiol 283: C358-C372, 2002.
Dunn KW, Sandoval RM, Kelly KJ, Dagher PC, Tanner GA, Atkinson SJ, Bacallao RL, and Molitoris BA. Functional studies of the kidney of living animals using multicolor two-photon microscopy. Am J Physiol Cell Physiol 283: C905-C916, 2002.
Holstein-Rathlou NH. Synchronization of proximal intratubular pressure oscillations: evidence for interaction between nephrons. Pflügers Arch 408: 438-443, 1987.
Holstein-Rathlou NH and Leyssac PP. TGF-mediated oscillations in proximal intratubular pressure: differences between spontaneously hypertensive rats and Wistar-Kyoto rats. Acta Physiol Scand 126: 333-339, 1986.
Kaplan MR, Plotkin MD, Brown D, Hebert SC, and Delpire E. Expression of the mouse Na-K-2Cl cotransporter, mBSC2, in the terminal inner medullary collecting duct, the glomerular and extraglomerular mesangium, and the glomerular afferent arteriole. J Clin Invest 98: 723-730, 1996.
Kleinfeld D, Mitra PP, Helmchen F, and Denk W. Fluctuations and stimulus-induced changes in blood flow observed in individual capillaries in layers 2 through 4 of rat neocortex. Proc Natl Acad Sci USA 95: 15741-15746, 1998.
Layton AT, Moore LC, and Layton HE. Multistability in tubuloglomerular feedback and spectral complexity in spontaneously hypertensive rats. Am J Physiol Renal Physiol. In press.
Leyssac PP and Baumbach L. An oscillating intratubular pressure response to alterations in Henle loop flow in the rat kidney. Acta Physiol Scand 117: 415-419, 1983.
Lorenz JN and Gruenstein E. A simple nonradioactive method for evaluating single-nephron filtration rate using FITC-inulin. Am J Physiol Renal Physiol 276: F172-F177, 1999.
Marsh DJ, Sosnovtseva OV, Chon KH, and Holstein-Rathlou NH. Nonlinear interactions in renal blood flow regulation. Am J Physiol Regul Integr Comp Physiol 288: R1143-R1159, 2005.
Molitoris BA and Sandoval RM. Intravital multiphoton microscopy of dynamic renal processes. Am J Physiol Renal Physiol 288: F1084-F1089, 2005.
Oppermann M, Castrop H, Briggs J, and Schnermann J. Influence of furosemide on renal hemodynamics in mice (Abstract). J Am Soc Nephrol 16: 393A, 2005.
Peti-Peterdi J, Morishima S, Bell PD, and Okada Y. Two-photon excitation fluorescence imaging of the living juxtaglomerular apparatus. Am J Physiol Renal Physiol 283: F197-F201, 2002.
Peti-Peterdi J, Fintha A, Fuson AL, Tousson A, and Chow RH. Real-time imaging of renin release in vitro. Am J Physiol Renal Physiol 287: F329-F335, 2004.
Peti-Peterdi J. Multiphoton imaging of renal tissues in vitro. Am J Physiol Renal Physiol 288: F1079-F1083, 2005.
Sandoval RM, Kennedy MD, Low PS, and Molitoris BA. Uptake and trafficking of fluorescent conjugates of folic acid in intact kidney determined using intravital two-photon microscopy. Am J Physiol Cell Physiol 287: C517-C526, 2004.
Sands JM. Micropuncture: unlocking the secrets of renal function. Am J Physiol Renal Physiol 287: F866-F867, 2004.
Schnermann J and Briggs J. Function of the juxtaglomerular apparatus: local control of glomerular hemodynamics. In: The Kidney, edited by Seldin DW and Giebisch G. New York: Raven, 1985, p. 669-697.
Tanner GA, Sandoval RM, and Dunn KW. Two-photon in vivo microscopy of sulfonefluorescein secretion in normal and cystic rat kidneys. Am J Physiol Renal Physiol 286: F152-F160, 2004.
Tanner GA, Sandoval RM, Molitoris BA, Bamburg JR, and Ashworth SL. Micropuncture gene delivery and intravital two-photon visualization of protein expression in rat kidney. Am J Physiol Renal Physiol 289: F638-F643, 2005.
Thomson SC, Deng A, Komine N, Hammes JS, Blantz RC, and Gabbai FB. Early diabetes as a model for testing the regulation of juxtaglomerular NOS I. Am J Physiol Renal Physiol 287: F732-F738, 2004.
Wearn JT and Richards AN. Observations on the composition of glomerular urine, with particular reference to the problem of reabsorption in the renal tubules. Am J Physiol 71: 209-227, 1924.
Wolf G, Chen S, and Ziyadeh FN. From the periphery of the glomerular capillary wall toward the center of disease: podocyte injury comes of age in diabetic nephropathy. Diabetes 54: 1626-1634, 2005.
Yip KP, Holstein-Rathlou NH, and Marsh DJ. Chaos in blood flow control in genetic and renovascular hypertensive rats. Am J Physiol Renal Fluid Electrolyte Physiol 261: F400-F408, 1991.
Yu W, Sandoval RM, and Molitoris BA. Quantitative intravital microscopy using a Generalized Polarity concept for kidney studies. Am J Physiol Cell Physiol 289: C1197-C1208, 2005.
Zipfel WR, Williams RM, and Webb WW. Nonlinear magic: multiphoton microscopy in the biosciences. Nat Biotechnol 11: 1369-1377, 2003.
作者单位：Departments of Physiology and Biophysics and Medicine, Zilkha Neurogenetic Institute, University of Southern California, Los Angeles, California