Glomerular filtration rate dependence of sieving of albuminand some neutral pro - 《American Journal of Physiology》美国生理学杂志干细胞之家



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 23 下一页

返回列表

查看: 395237\|回复: 228	go Glomerular filtration rate dependence of sieving of albuminand some neutral pro [复制链接]

轻羽

新手上路

Rank: 1

积分: 0
威望: 0
包包: 0

楼主

发表于 2009-4-21 13:37 |只看该作者 |倒序浏览 |打印

作者：UllaLund, AnnaRippe, DanieleVenturoli, OlavTenstad, AndersGrubb,  BengtRippe作者单位：Departments of Nephrology and Clinical Chemistry, University Hospital, S-221 85 Lund, Sweden; and Department ofPhysiology, University of Bergen, N-5509 Bergen, Norway

      h' V, {$ n( y* E' A: s" C# W

                     ( h* J) \9 z+ d$ L$ E: H1 k
         0 [+ a& D* l- }
         ' u( q8 r; P# d+ p6 W0 K: F
         ( q) x$ l8 @+ o



      6 W, S3 T% t6 U8 T  A1 `' o

      【摘要】2 T/ J5 C; W  \- R2 P
   The size and charge-selective propertiesof the glomerular barrier are partly controversial. Glomerular sievingcoefficients ( ) for proteins have rarely been determinednoninvasively before in vivo. Therefore, was assessed vs.glomerular filtration rate (GFR; 51 Cr-EDTA clearance) inintact rats for radiolabeled myoglobin, -dimer, neutral horseradishperoxidase (nHRP), neutral human serum albumin (nHSA), and nativealbumin (HSA). To obtain, glomerular tracer clearance, assessedfrom the 7- to 8-min kidney uptake of protein, was divided by the GFR.The data were fitted with a two-pore model of glomerular permeability,where the small-pore radius was 37.35 ± 1.11 (SE) Å, and the"unrestricted pore area over diffusion path length"( A 0 / X ) 1.84 ± 0.43 · 10 6 cm. Although seeminglyhorizontal for nHRP and nHSA, the log vs. GFR curves showedslightly negative slopes for the proteins investigated in the GFRinterval of 2-4.5 ml/min. Strong negative (linear) correlationsbetween (log) and GFR were obtained for myoglobin( P = 0.002) and HSA ( P = 0.006),whereas they were relatively weak for nHRP and nHSA and nonsignificantfor -dimer. for nHSA was markedly higher than that for HSA. Inconclusion, there were no indications of increases in vs. GFR, asindicative of concentration polarization, for the proteins investigatedat high GFRs. Furthermore, the glomerular small-pore radius assessedfrom endogenous (neutral) protein sieving data was found to be smallerthan previously determined using dextran or Ficoll as test molecules. + e" R+ A3 U" Y% ^
      【关键词】 glomerular permeability macromolecules reflection coefficient transport
   INTRODUCTION: l- W$ Z- U/ m* h$ y

THE GLOMERULAR BARRIER SELECTS molecules based on their size, shape, and charge,and it almost completely prevents large macromolecules from reachingBowman's space ( 8 ). The fenestrated endothelium with itsglycocalyx, the glomerular basement membrane (GBM), and the epithelialfiltration slits are arranged in series to produce this highlyselective sieving filter. There is little agreement as to where themajor barrier function is located ( 8 ). It has beensuggested that the most size-selective portion of the glomerular barrier be represented by the podocyte slit membrane (PSM), especially by a zipper-like arrangement of structures in this membrane( 24 ), conceivably made up (partly) of nephrin molecules( 38 ). Some authors have brought attention to the fact thatthe most charge-selective barrier may be located close to the plasmacompartment, possibly in the endothelial glycocalyx ( 25 ),whereas the most size-selective barrier may be more distally located( 18 ). That the charge selectivity may be located in theendothelial glycocalyx has become even more evident after measurementsof the charge-barrier properties of isolated GBMs, which were similarfor neutral and negatively charged Ficoll molecules ( 2 ) orfor native (anionic) and cationized albumin ( 1 ).! }8 {  f! ^2 c: u2 J$ H/ m6 }: ?; v

If the PSM were the major sieving barrier of the glomerular filter,this arrangement would result in concentration polarization of proteinsin the GBM at high glomerular filtration rates (GFRs) ( 9 ).According to the fact that the relative contribution of diffusionaltransport decreases with increasing GFRs, high filtration rates willnormally lead to reductions in the glomerular sieving coefficients( ) for (small) macromolecules ( 4, 20, 21, 23, 33 ). Bycontrast, if concentration polarization occurs, then increases in may instead be expected for the highest filtration rates( 9 ). However, only very few studies have been performed, particularly in vivo, in which the GFR dependence of formacromolecules has been systematically investigated.; h( A) G3 U4 @1 y* W
5 M( m" x" m6 ~+ y! ?
In view of the paucity of data on fractional clearances ofmacromolecules, especially of proteins, as a function of GFR, we assessed the glomerular for a number of neutral proteins and albumin at normal and high GFRs using a noninvasive technique in intactrats ( 35, 36 ). Measured values were consistent with atwo-pore model of glomerular permselectivity ( 23, 33 ), inwhich the small-pore radius was ~37.4 Å, when the(negatively charged) large-pore radius was set at 110 Å. Whereas thesmall-pore radius was smaller than that usually obtained using Ficollor dextran as test molecules, measured diffusional small-solutecapacities, i.e., the effective area for diffusion over unit pathlength ( A 0 / X ), were largelyconsistent with the calculated glomerular filtration coefficient( L p S ). Furthermore, A 0 / X (and L p S ) remained stable as a function of GFR. Inaddition, there were no indications of concentration polarization(increases in ) occurring at the highest filtration rates.8 e) V% P: K& z9 p/ H. B8 L% ]" D2 r2 N
* A2 W0 N" o0 J6 `* k  J
METHODS

Experiments were performed in 85 male Wistar rats (Møllegaard,Stensved, Denmark) weighing 270 ± 8 (SE) g. The rats were kept onstandard chow and had free access to water before the start of theexperiments. Experiments were approved by the Animal Ethics Committeeat Lund University.! ]0 [# x. Q& ?0 M1 F) f, N

Anesthesia was induced using pentobarbital sodium (50 mg/kg ip), and athermostatically controlled heating pad maintained the body temperatureat 37°C. A tracheotomy was performed to ensure free airways. The tailartery was cannulated for recording the arterial pressure(P A ) and for subsequent administration of drugs. The rightjugular artery and the left jugular vein were cannulated for infusionand sampling purposes. Via an abdominal incision, a catheter was placedin the urinary bladder for continuous urine sampling, the abdominalinsertion being sealed with Histoacryl (Melsungen, Germany).

Tissue uptake technique. The technique has been described in detail and validated by Tenstad etal. ( 35, 36 ). When a tracer protein is added to the plasmacompartment, it will mix with the plasma, dissipate within theextracellular space, and filter across the glomerular barrier. Afterappearing in Bowman's space, it will be reabsorbed, more or lesscompletely, by the renal proximal tubules to be processed by thetubular cells. During the first 7-9 min of protein reabsorption, the breakdown of the protein and the subsequent reabsorption to theplasma of split products will be negligible, whereas a tiny fraction ofthe tracer will appear in the urine. This is the principle utilized inthe present experiments. Glomerular protein clearance was assessed asthe timed total (cortical) kidney uptake plus the (precipitable) urineexcretion of protein tracer divided by the average plasma tracerprotein concentration. Was calculated from the protein clearancedivided by GFR, determined by the simultaneous assessment of theplasma-to-urine clearance of 51 Cr-EDTA.7 _9 E! v. L# r% @8 u+ y
- f1 e6 s% Y, b! a/ ?$ [7 b3 p
For GFR measurements, 51 Cr-EDTA (Amersham, Biosciences,Buckinghamshire, UK) was given in a priming dose (0.09 MBq in 0.2 ml iv), followed by a constant infusion (0.005 MBq/min) for repeated measurements of the plasma-to-urine 51 Cr-EDTA clearanceduring 20-min intervals throughout the study. During the infusion,blood sampling (20 µl at a time) was performed approximately every 10 min using microcapillaries. Urine was also sampled approximately every10 min. After at least one measurement of 51 Cr-EDTAclearance, a constant infusion of tracer protein was performed for7-8 min, concomitant with repeated sampling of plasma (20 µlevery 2 min) and urine for the entire infusion period (7-8 min),after which the animals were killed using saturated KCl (iv). Bothkidneys were then removed, blotted, weighed, and assessed forradioactivity. TCA (10%)-precipitable urine radioactivity was assessedand included in the clearance measurements. All radioactivity measurements were performed in a gamma scintillation counter (Wizard 1480, LKP Wallac, Turku, Finland). Appropriate corrections for radioactive decay and spillover from the 51 Cr to the 125 I channel were performed.

A modified protocol was used for native and neutralized albumin. After8 min of tracer infusion, a whole body vascular washout was started byrapidly infusing an equal mixture of 0.9% saline and heparinized horseserum (SVA, Uppsala, Sweden) containing 1 mg/l papaverine (vasodilator,P 3510, Sigma, St. Louis, MO) via the jugular vein (or sometimes viathe carotid artery) at a rate of 20 ml/min, after the inferior venacava was opened via a laparotomy. This usually occurred at ~10 minafter the start of the tracer infusion, because the laparotomy usuallylasted 2-3 min. During the subsequent 8 min of washout, theanimals usually expired within the first 2 min. In the albuminexperiments, the inner renal medulla (rich in interstitial tissue) wasdissected away from the rest of the kidney and not included in theradioactivity measurements. TCA-precipitable urine radioactivity,however, was included.

Experiments were performed at either the prevailing GFR or at elevatedGFRs. Increases in GFR were induced by volume loading the animals viaan infusion (iv) of horse serum and by infusing glucagon (iv). Fivemilliliters of horse serum were given for ~1 min, starting 5 minbefore the protein tracer infusion period, and 2 ml were given,starting 2 min before the test period. Furthermore, to further increaserenal blood flow, glucagon (1 mg/ml iv, Novo Nordisk, Copenhagen,Denmark) was infused at 3 µg/min, starting 2 min before andcontinuing throughout the test period.# d- G* w' S7 D1 ]. [
# ?9 H4 L6 I% Q1 c/ r8 S9 T/ O
Tracers and labeling procedures. The protein probes were labeled with 125 I by using1,3,4,6-tetrachloro-3,6 -diphenylglycouril (Iodo-Gen)( 10 ). Briefly, 0.1 mg Iodo-Gen (T0656, Sigma) dissolved in0.1 ml chloroform was dispersed in a 1.8-ml Nunc vial (Nunc-Kamstrup,Roskilde, Denmark). A film of the virtually water-insoluble Iodo-Genwas formed in the Nunc vial by allowing the chloroform to evaporate todryness under nitrogen. Then, 1 ml 0.05 M PBS solution, pH 7.5, containing 1-2 mg protein to be labeled, 5 MBq 125 I(Institute for Energy Technique, Kjeller, Norway), and 15 µl 0.01 MNaI were added, and the iodinating tube was gently agitated for 10 minbefore the reaction was terminated by removing the solution from theIodo-Gen tube. Unincorporated iodine isotope accounting for TCA precipitation, was removedby dialyzing the tracer against 1,000 ml 0.9% saline containing 0.02%azide. The stock solution was stored in the dark at 4°C and dialyzedfor at least 24 h before use.
0 S. }, ^5 o: y6 Q4 R
The following proteins were tested: myoglobin (M 0630, Sigma),horseradish peroxidase (HRP; type XII; P8415, Sigma), a human myelomadimeric -chain (a gift from Prof. Anders Grubb, Dept. of ClinicalChemistry, University Hospital, Lund, Sweden), neutral human serumalbumin (nHSA; prepared by Olav Tenstad according to the techniquedescribed below), and prelabeled native albumin ( 125 I-HSA)purchased from Kjeller, Norway (Institute for Energy Technique, Horten,Norway). Tracer characteristics (molecular weight, Stokes-Einstein radius, and isoelectric point), as determined using HPLC (Superdex 75HPand Superose 12 HR columns) and isoelectric focusing, respectively (seebelow), are shown in Table 1. Calibrationstandard curves used for the Superdex 75HP gel filtrationdeterminations were based on BSA, egg albumin, chymotrypsinogen, andRNase. The level of free (unbound) 125 I was always checkedbefore use by TCA precipitation and was kept below 1.5% (usually
5 b: e9 r5 w2 m7 a' R9 @/ }
Table 1. Physical characteristics of the molecules investigated

Neutralization of HSA. nHSA was obtained by a graded modification of the COOH groups using aprocedure modified from that described by Hoare and Koshland( 11 ) as follows. HSA (1.5 g) was dissolved in 15 ml 0.133 M glycine methyl ester at pH 4.75 (at room temperature). A solution of5 ml of 0.04 M N -ethyl- N '-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) was then added to the mixture toinitiate the reaction. The pH was continuously recorded and kept at4.75 by addition of 0.1 M NaOH. Aliquots (1 ml) were removed every 5 min for 60 min and immediately added to 1 ml of 4.0 M acetate buffer atpH 4.75 to quench the reaction. After being kept for a few minutes atroom temperature, these solutions were dialyzed overnight against twochanges of 10 liters of distilled water, and the dialysate wasfreeze-dried and stored at 20°C. The effect of the reaction wasevaluated by isoelectric focusing using a vertical minigel system (CBSScientific) and Novex (Novel Experimental Technology, San Diego, CA)precast gels. It turned out that a 45-min reaction time producedalbumin with an average isoelectric point close to 7.4 without anysignificant change in hydrodynamic radius, as measured by HPLC(Superdex 75 HR and Superose 12 HR).7 f9 ?- \( B, ]. l0 D2 P
9 N/ M& S- j( T
Calculations. Renal tracer protein clearance was calculated from the amount of tracerradioactivity accumulated in both kidneys plus the TCA-precipitableurine tracer activity (collected during the tracer infusion period)divided by the average venous plasma tracer concentration and by thetracer infusion time until death. In washout experiments, clearance wasassessed by the amount of tracer in kidneys plus urine divided by thearea under the curve of the plasma tracer concentration vs. timefunction. Protein values were calculated by dividing the measuredprotein clearance by the simultaneously assessed GFR.

Values of calculated as described above, or corrected for the factthat plasma proteins are upconcentrated in the glomeruli due to thefiltration process, will be presented in this study. When correctionswere used, we employed the formula (cf. Ref. 21 ) =&thgr; m  ·  2[1 − FF(1 −&thgr; m )] 2 − FF(1 −&thgr; m )

( 1 )
* u+ I3 O8 R! I7 P2 @$ z/ B
where m represents the measured protein sievingcoefficient and the corrected sieving coefficient. FF is thefiltration fraction, which was determined under identical conditionsmeasuring 125 I-hippuran clearance simultaneously with 51 Cr-EDTA clearance at normal and elevated (and reduced)GFRs in a parallel study yielding the following relationship between FF and GFR = 0.47  · e −0.225  ·  GFR6 h4 q  g  e. n" ^
5 m+ ^; \2 m2 Y
( 2 )) t3 m7 N) w- k

GFR was calculated from = C u, E  ·  V u C pw
7 O% S7 M2 H/ G9 L% Q5 ^
( 3 )
; E  q1 p+ q  M& }+ b4 p
where C u, E represents the urinaryconcentration of Cr-EDTA and V u represents the urine flow(per min), respectively, and C pw is the plasma waterconcentration of Cr-EDTA. C pw was obtained from( 41 ) pw  =  C P, E 0.984 − 0.000718 · C prot
  K, B9 F0 {3 B$ J8 C6 o) N5 V
( 4 )
  g5 q8 l: Z8 x: t$ l4 z( {9 d
where C P, E represents the plasma concentration ofCr-EDTA, and C prot represents the plasma concentration oftotal protein.# z5 u+ m6 W* K; a; W' I

Data for were fitted to a two-pore model of membranepermeability ( 23, 34 ) using nonlinear least squaresregression analysis. A number of highly sophisticated glomerularsieving models have been published recently (for a review, see Ref. 8 ). We chose a pore model, because such models have beenwidely applied over the past few decades. Furthermore, a pore model isperhaps the simplest model that may adequately describe glomerulartransport data. The following parameters were estimated: the small-pore radius ( r s ), the unrestricted pore area overunit diffusion path length [ A 0 / X;from which the hydraulic conductance ( L p S ) could be calculated], and the fractional L p S accountedfor by the large pores ( L ). The large-poreradius ( r L ) was set at 110 Å based on resultsfrom a previous study from our laboratory ( 34 ). Based onthat estimate, the large-pore volume flow( J v, L ) was determined from the sievingcoefficient of native albumin. Because native albumin is negativelycharged, it should be completely excluded (see below) from thesmall-pore pathway in the glomerular filter. It can thus be predictedto be entirely dependent on convective transport across the large pores(according to Refs. 23 and 34 ) alb  =  J v, L  · (1 − &sfgr; L ) GFR

( 5 )7 P$ N: ]4 r3 S: o" P% L; d
5 Q7 m2 e  s) S* ~: n1 I, O) h5 Z
where L is the large-pore albumin reflectioncoefficient, as calculated taking the negative charge of albumin andthe large pores into account (see below), and alb is thesieving coefficient for albumin. alb Was found to be0.00066 ± 0.000054 ( alb, when corrected accordingto Eq. 1, was determined to be 0.00057 ± 0.000038). Knowing J v, L, the small-pore volume flow( J v, s ) could be calculated for any given GFR.Furthermore, L could also be assessed (from Eq.4 in Ref. 34 ), assuming r L tobe 110 Å and (the transglomerular oncotic pressure gradient) tobe 26 mmHg, setting L p S at 0.36 ml · min 1 · mmHg 1 (from corrected data), as calculated from the value of A 0 / X obtained ( 23, 34 ).
4 n' G, r: y' s& u
The degree of restricted diffusion and the magnitude of weremodeled as a function of solute radius according to the equations givenby Mason et al. ( 16 ). With respect to the glomerulartransport of native albumin (HSA), the negative solute and pore chargewere approximately accounted for by applying the Debye-Hückeltheory of ion-ion interaction by adding 8 Å to the molecular radiusand subtracting 8 Å from the pore radius ( 17 ). Althoughsomewhat crude, the Debye-Hückel theory, compared with a moreexact description given by Smith and Deen ( 28 ) of therejection of charged solutes from pores with charged walls, proved toagree excellently with the exact theory for solutes with radii rangingfrom 10 to 70 Å and pores with radii of ~100 Å for equal solute andmembrane charge ( 20 mM) ( 6 ).6 x9 B2 d6 [$ Q9 m: U2 }
; G$ {: p" R# C  `! B! ]
All calculations were performed using nonlinear flux analysis( 20 ), as described in detail previously ( 34 ),and Microsoft Excel and an incorporated analysis tool, Solver, according to a modification of the method describedin Ref. 39.

Statistics. Values are given as means ± SE. Differences among groups weredetected using ANOVA. Calculating the variance-covariance matrix (see APPENDIX A ) assessed the SE of the fitted parameters.# A6 E# U. _  {4 r
6 x2 p* f+ `! T$ L
RESULTS
1 a; y% W+ A) R3 Q
In Fig. 1, fractional clearances(uncorrected values or " raw ") for the neutralproteins investigated and for native albumin are plotted vs. GFR in asemilogarithmic diagram. In the GFR interval of 2.0-4.5 ml/min, correlated negatively with GFR for all proteins investigated,except for the -dimer (see below). Values for correctedaccording to Eq. 1 are plotted vs. GFR in Fig. 2, and, furthermore, the best fittingtwo-pore parameters were here fitted to the data. According to thetwo-pore model, r s was calculated to be37.35 ± 1.11 (SE) Å (for uncorrected data we obtained 37.55 ± 0.75 Å) when r L was fixed at 110 Å.Furthermore, A 0 / X was 1.84 · 10 6 ± 4.29 · 10 5 cm (2.40 · 10 6 ± 5.19 · 10 5 cm for uncorrected data), and l was 4.86 · 10 4 ± 2.34 · 10 4 (4.61 · 10 4 ± 6.25 · 10 5 for uncorrected data). Average data, corrected anduncorrected (or raw), for (at an average GFR of ~3 ml/min) and for myoglobin, -dimer, neutral HRP, and neutral and negativealbumin are shown in Table 2. L p S calculated from A 0 / X (corrected data) was 0.36 ml · min 1 · mmHg 1 (both kidneys). For a GFR of 3 ml/min, with the assumption of a nettransglomerular pressure gradient on the order of 10 mmHg in ourexperimental animals, L p S can be estimated to be0.3 ml · min 1 · mmHg 1.Thus the A 0 / X calculated fromsieving data for small and intermediate size solutes was largelyconsistent with the filtration coefficient of the glomerular filtrationbarrier. Furthermore, these two entities could be set constant (andindependent of GFR) in all simulations.
2 V& `; C8 q6 ]* i9 o5 v
Fig. 1. Fractional clearances ( ) for the neutral proteins myoglobin( ), -dimer ( ), neutral horseradishperoxidase (nHRP; ), and neutral human serum albumin(nHSA; ), together with those for native albumin (HSA; ) plotted as a function of glomerular filtration rate(GFR) in a semilog diagram. Data were not corrected according to Eq. 1.0 c8 a3 X' ]: e2 p" S

Fig. 2. Values for for the proteins investigated, corrected accordingto Eq. 1, plotted as a function of GFR in a semilog diagram.Symbols are as defined in Fig 1. Computer simulated vs. GFR curvesare shown for the best fitting 2-pore parameters, i.e., for small-poreradius ( r s ) = 37.35 Å, unrestricted porearea over diffusion path length( A 0 / X ) = 1.84 · 10 6 cm, and fractional hydraulicconductance ( L ) = 0.00049, when large-pore radius( r L ) was preset at 110 Å. The prominentdiffusive protein transport at low GFRs resulted in a negativedependence of on GFR, particularly evident for myoglobin. Fornative (negative) HSA, passing only through large pores, there is alsoa dependence of on GFR because the fractional large-pore volumeflow asymptotically falls (to approach L ) with increasesin GFR. There is no obvious indication of concentration polarizationfor any of the proteins investigated.- Z0 d/ }- X7 }0 c; P$ ?

Table 2. Sieving coefficients of the test molecules investigated
/ ]( X9 E" G' }8 @
Figures 1 and 2 indicate the presence of a negative dependence of on GFR. By applying a simple linear regression analysis of log vs.GFR, we obtained a highly significant negative correlation formyoglobin ( P = 0.002 for corrected ), HSA( P = 0.006), and nHSA ( P = 0.01), but abarely significant one for nHRP ( P = 0.04), whereas itwas nonsignificant for the -dimer. The regression coefficients withtheir 95% confidence intervals (for corrected and raw data) are listedin Table 3. The -GFR relationships arelargely consistent with models in which diffusion and convection occursimultaneously across a size (and/or charge)-selective barrier. Thusfor small proteins, the reduction in the diffusional component oftransport with increasing filtration rates will cause reductions in. However, for large proteins, and at high filtration rates, theimpact of diffusion is small. This results in an essentially flat vs. GFR curve, where approximates (1 ) at high GFRs. Note that for solutes with radii larger than the small-pore radius (37.35 Å), one would, according to the two-pore model, also expect adependency of on GFR. In a heteroporous model, this phenomenon results from the fact that the fractional large-pore volume flow ( J v, L /GFR) will asymptotically fallwith increases in GFR to approach L at high GFRs. Thisbehavior is expected for HSA, because the presence of chargeinteractions may completely prevent HSA from entering the small pores,whereas nHSA may filter through both small and large pores. Note alsothat nHSA transport was one order of magnitude higher than that ofnative (negatively charged) albumin. Finally, for all proteins,including the two largest investigated (nHSA and HSA), there were noindications of concentration polarization occurring at any filtrationrates.+ B/ z. Y# L: G6 V; K- `4 f, k$ @( L

Table 3. Statistical analysis of the measured sieving coefficients2 b) W( x/ r" g& w

Figure 3 illustrates (log) vs. bothGFR and solute radius in a three-dimensional diagram simulated usingthe present two-pore parameters (for corrected data). Here, it is againevident that, for solutes with radii is unity andcompletely independent of GFR, whereas for solutes with radiiranging between 15 and 30 Å is dependent on GFR. In the GFRinterval of 2-5 ml/min, however, solutes with radii of 30-37Å exhibit rather stable values, which are close to their (1 ) values. For solutes with radii larger than the small-poreradius, there is again a dependency of on GFR, determined by the J V, L /GFR ratio, as mentioned above.

Fig. 3. Values for simulated vs. both GFR and solute radius in a3-dimensional diagram of the 2-pore model parameters obtained in thepresent study. Note the semilogarithmic scale. Solutes with radii of 1 (or close to 1) throughout the GFR interval. Soluteswith radii of 15-30 Å show marked GFR dependence of their 30 Å, but smaller than the pore radius, the GFRdependence in the GFR interval 2-5 ml/min is moderate. For soluteswith radii larger than small-pore radii, or for negatively chargedmacromolecules, which are confined to the large-pore pathway for theirpassage across the glomerular membrane, there is a marked dependence of on GFR, attributable to the fact that the ratio of the large-porevolume flow over GFR ( J v, L /GFR) ishigher than L, but asymptotically approaches L when GFR is high.% \/ C2 V8 X9 R  }- T+ }4 j

DISCUSSION

The essential result of this study is that the fractionalclearances ( ) of small endogenous neutral proteins and of albumin inrats, assessed as a function of GFR in vivo, slightly declined withincreases in GFR in the GFR interval of 2.0-4.5 ml/min. This wasparticularly evident for myoglobin. The results essentially agree withdextran sieving data obtained during isoncotic volume expansion,performed to increase GFR, in Munich-Wistar rats ( 4 ). Withincreases in GFR, dextrans with radii of 20-38 Å clearly decreased their values. In the present study, there was no evidence for concentration polarization (increasing values) occurring athigh GFRs, either for the small proteins investigated or for neutral ornative albumin. Furthermore, the values for small proteins andalbumin were lower than previously obtained using Ficoll or dextran ofequal hydrodynamic radii as probes for testing glomerularpermselectivity. Also, there was a marked charge dependency ofglomerular transport, as evidenced by the large difference in forneutralized vs. negatively charged albumin.
7 Y6 [4 h+ n) |( ~, N2 q1 ~% l
There have been very few previous analyses of the dependence of protein values on GFR, at least in vivo. In the isolated perfused ratkidney (IPK) at 8°C, however, there is one recent set of measurements of this kind for "asymmetrical" proteins ( 19 ). Although in the IPK the GFR (per kidney) was only20% of those obtained at 37°C in the intact rat in vivo, theobserved for the most permeable proteins investigated (hyaluronanand bikunin) showed a similar asymptotic reduction as a function of GFR, as found in the present study. This is indeed the expected behavior of when the diffusional component of transport is high, because this component will theoretically decrease asymptotically withincreases in GFR according to nonlinear transport formalism ( 20 ). Only at high GFRs, the impact of the diffusionalcomponent will become negligible, so that the protein will equal(1 ) ( 7, 8, 20, 21, 23, 33 ). Contrary toresults from the IPK ( 19 ), there were no indications ofincreases in microvascular permeability occurring for albumin at highintraglomerular hydrostatic pressures in the present study. A tentativeexplanation could be that the IPK, although partly protected frominflammatory mediators or ischemia-reperfusion injury by thetemperature reduction, might be more vulnerable to high intraluminalpressures than is the intact kidney under in vivo conditions.3 K$ ~! O, _2 K7 D1 A

One important consequence of the presence of a large diffusionalcomponent of small protein transport across the glomerular filter,i.e., a large A 0 / X for the renalmicrocirculation, is that assessments of for small proteins must bestandardized to rather narrow GFR ranges to be compared betweendifferent experimental conditions. For example, if anischemia-reperfusion insult per se results in a fall in GFR,then by necessity for a small protein, such as myoglobin or 2 -microglobulin, must increase, even if the glomerularpermeability is unaffected. On the other hand, if GFR is increased butthe permeability is unchanged, then for a protein will fall. Forthe largest neutral molecules investigated in the present study( -dimer, nHRP, and nHSA), however, the diffusional contribution to was expected to be low in the whole GFR range investigated. Indeed, was independent of GFR changes for the -dimer and nearly so fornHRP. Unexpectedly, there was, however, a slight decline of the log vs. GFR relationship for nHSA.& E0 l/ I- U: I" d% I3 ?) i
& I3 _0 M: V8 W
Theoretically, of a magnitude measured for HRP and nHSA wouldremain stable as a function of GFR even at high filtration rates, ifthe major barrier to solute sieving were close to the blood side of themembrane. However, if the major barrier function were instead locatedclose to Bowman's space, e.g., at the PSM, then one would expect atleast some degree of concentration polarization to occur at highfiltration rates. According to a recent modeling study of the sievingbehavior of the glomerular capillary wall ( 9 ), it wasassumed that the glomerular barrier exhibited three transportresistances arranged in series, with a major portion of the overalltransport resistance to macromolecules present at the level of the PSM.Furthermore, it was assumed that the resistance of the transport oflarge solutes was low (negligible) at the fenestrae. Under suchassumptions, a rise in single-nephron GFR from 40-45 to ~80nl/min, corresponding to a rise in whole rat (300 g) GFR from ~2.5 to5 ml/min, caused a significant rise in for (neutral) solutes havinga similar to that of albumin (nHSA and HSA) in the present study.However, because we were not able to detect any signs of concentrationpolarization occurring in the GFR interval of 2-4.5 ml/min, we areinclined to conclude that the case for a major sieving barrier locatedat the PSM is rather weak. In case the slit membrane would still be themajor filtration barrier, the present data indicate that serialbarriers proximal to the PSM must be very highly permeable tomacromolecules to prevent the buildup of concentration polarizationlayers at the PSM.
& ?, M3 d# s9 u. k; r& p0 O* Y
The present study essentially confirms and extends previousmeasurements of using micropuncture techniques. Micropuncture techniques have been criticized, because they imply exposure of andmechanical interactions with an intact kidney. Furthermore, proteinssampled from the tubules may bind to the glass pipette, andinterstitial proteins may leak into the tubules during the micropuncture. Moreover, because the tubular micropuncture procedure has to be performed at sites distally to Bowman's capsule, primary urine cannot be directly assessed ( 15 ). Indeed, tubularprotein concentration falls along the distance of the proximal tubule, because protein reabsorption is usually more avid than that of water.In an attempt to avoid all these sources of error, Tojo and Endou( 37 ) used a double-barrel pipette technique, which made itpossible to seal the punctured (rat) proximal tubule from theinterstitium. Furthermore, they assessed the tubular concentration ofprotein together with that of a filtration marker (inulin) at variousdistances from Bowman's capsule ( 37 ). With thistechnique, they were able to quite precisely estimate the urinaryalbumin protein concentration of Bowman's capsule by an extrapolation procedure. Using this careful technique, they estimated the valuefor native albumin to be 6.2 · 10 4, which is almost identical to that assessed by thepresent technique in vivo (6.6 · 10 4 ). Also, our assessments of values for myoglobin,dimeric -chain (Bence Jones proteins), and nHRP are remarkably closeto estimates previously obtained using micropuncture techniques( 14 ).
) e/ J0 A0 g9 j# q: Y2 N; c  M
All measured values of for neutral proteins in the present studyare much lower than the corresponding values previously obtainedfor neutral Ficoll, which, in turn, are much lower than values fordextran ( 15 ). The marked discrepancy between glomerular protein sieving data and glomerular dextran sieving data was discussed at some length in the classic review by Renkin and Gilmore( 21 ). It may be due to the fact that dextrans are flexiblemolecules, and thereby hyperpermeable in vivo, so that they mayactually transmigrate through pores, which are even smaller than their Stokes-Einstein radii, sometimes denoted "reptation"( 17 ). Moreover, recent data indicate that the more idealFicoll molecule, a copolymer of epichlorhydrine and sucrose, may notbehave in all aspects as an ideal ridgid sphere ( 12, 27 ),but we will return to this issue in a forthcoming publication. At anyrate, the permselectivity of the glomerular barrier, in terms of thesmall-pore radius, for example, seems to be dependent on the physicalproperties of the probe used for testing permeability. Using neutraldextran as a probe, the average glomerular r s has been determined to be on the order of 50-55 Å ( 15 ). Using Ficoll at normal ionic strength, r s has been determined to be on the order of 45 Å ( 18, 19 ), whereas at low ionic strengths r s was only 41 Å ( 30 ). Thisvalue is similar to the rather low r s estimateof the present study and to earlier estimates using proteins forprobing glomerular permeability ( 21 ).- {& M8 x; w2 _- [

It has been well established since the 1970s and 1980s that theglomerular filter discriminates among macromolecules based on boththeir net charge as well as their size ( 15 ). Much of theevidence in favor of charge selectivity of the glomerular filter hasbeen based on comparisons between sieving data for uncharged andanionic dextran (dextran sulfate) ( 3 ). Although vividarguments against glomerular charge selectivity have been raised duringthe last decade ( 5, 26, 32, 42 ), strong evidencesupporting the classic view was recently given by comparing neutral andanionic lactate dehydrogenase or neutral and anionic HRP in the IPK( 13, 31 ). These studies largely confirm the classicstudies by Rennke et al. ( 22 ) for differently charged HRP( 22 ). The present data are entirely consistent with the glomerular filter as a charge-selective barrier, producing a near 10-fold difference in for neutralized vs. negatively charged (native) albumin.
( i# I( L+ A" G' N2 n& V/ a
The present tissue uptake technique has been validated for smallproteins in previous publications ( 35, 36 ). For proteins with very low renal clearances, such as albumin, it is crucial that thekidneys are completely washed free of intravascular tracer and that thebulk of interstitially accumulated tracer (and free iodine) is to alarge extent cleared by back-diffusion to the rinse fluid. The washoutprocedure is thus crucial to the success of the technique. Even thoughwe consider the washout to have been more or less complete, we cannotcompletely rule out that some tracer remained either intravascularly orextracellularly after tracer washout. From that point of view, thepresent for native albumin of 6.6 · 10 4 may represent an overestimate. Still, the valueobtained is in agreement with the recent careful micropuncture study byTojo and Endou ( 37 ) referred to above. Therefore, we feelconfident that the degree of overestimation of for native albuminwas, after all, rather moderate.

In conclusion, there was a dependence of glomerular small-protein on GFR for neutral molecules with molecular radii ranging between 15 and 30 Å and also for native albumin. The data were readily fitted toa two-pore model of glomerular permeability where r s was found to be ~37-38 Å. Neither forsmall proteins nor for albumin was there any evidence for concentrationpolarization present at high GFRs. Furthermore, the glomerular filtershowed properties of a negative charge barrier. Taken together, thepresent in vivo data may be interpreted to indicate that theendothelial glycocalyx-filled fenestrae are playing a greater role thanpreviously thought, and the epithelial slit diaphragms a lesser role,in determining the sieving properties of the glomerular filtration barrier.

APPENDIX A

The fractional clearance data ( ) were fitted to a two-poremodel of membrane permeability ( 23, 34 ) using a weightednonlinear least squares regression analysis. In detail, the function to be minimized was w  = 

∑ 1 n s W 2 s  ·  &thgr; exp s,  i  − &thgr; th s,  i &thgr; exp s,  i 24 N; ~3 _% r! e/ E+ h& j4 J
' g% o7 V2 F  Y) j1 Z5 r6 E
( A1 )
, \  }4 r$ U$ ?
where exp and th areexperimentally and theoretically calculated, respectively, and thesum is extended to all the experimental points( n s ) collected for each solute (s) considered.To correct for the different number of experimental points among thedifferent solutes considered, the weight W s wasdefined as s  = 1 −  n s N' M, `' h- e, a) y: Z3 C
# F, `- D( U  r0 H. T; i
( A2 )
. C4 K% i, G. m+ H6 i
where N is the total number of measurements. Tocompensate for the extremely large range of experimental values (3 orders of magnitude from myoglobin to native albumin values), therelative difference with respect to exp was introducedin place of the usual squared difference.

The estimated parameters were r s, A 0 / X, and the fractional L p S accounted for by the large pores( L ). However, because the numerical values of theseparameters differ by several orders of magnitude (9 from A 0 / X to L ), a setof scaling multipliers was introduced, so that the minimizationalgorithm had to deal with parameters near to unity.
" j% M" W7 j! ]
SE of the fitted parameters was assessed by calculating thevariance-covariance matrix according to the method described by Smithet al. ( 29 ).

ACKNOWLEDGEMENTS
1 m3 J7 f. K4 _4 w' z
We are grateful to Kerstin Wihlborg for skillful typing and editingof the manuscript. The expert technical assistance by VeronicaLindström (Dept. of Clinical Chemistry, University Hospital, Lund, Sweden) is acknowledged.; y6 g6 ?2 w* l: n0 K$ s
      【参考文献】
1. Bertolatus, JA,andKlinzman D. Macromolecular sieving by glomerular basement membrane in vitro: effect of polycation or biochemical modifications. Microvasc Res 41:311-327,1991  .

! S& h+ }* r  O3 s" N* D

2. Bolton, GR,Deen WM,andDaniels BS. Assessment of the charge selectivity of glomerular basement membrane using Ficoll sulfate. Am J Physiol Renal Physiol 274:F889-F896,1998 .

( R7 w) y) H/ u7 L

3. Chang, RL,Deen WM,Robertson CR,andBrenner BM. Permselectivity of the glomerular capillary wall. III. Restricted transport of polyanions. Kidney Int 8:212-218,1975  .& r8 v+ t3 V  T5 [
3 R7 i5 m: W5 b1 Y+ H7 x- Z

% r$ s( j% _3 w! {5 s3 V/ K( ]
4. Chang, RL,Ueki IR,Troy JL,Deen WM,Robertson CR,andBrenner BM. Permselectivity of the glomerular capillary wall to macromolecules. Biophys J 15:887-906,1975 .
; e6 u- b2 z" w
, x7 K+ l5 j+ c* O. t- j3 g

5. Comper, WD,andGlasgow EF. Charge selectivity in kidney ultrafiltration. Kidney Int 47:1242-1251,1995  .

6. Curry, FE. Mechanics and thermodynamics of transcapillary exchange.In: Handbook of Physiology. The Cardiovascular System. Bethesda, MD: Am. Physiol. Soc, 1984, sect. 2, vol. IV, pt. 1, chapt. 8, p. 309-374.  f  T  u; ^* `2 G8 }# z# w) S
7 r* Y0 q  D! `9 B. z1 G- o6 K) m$ e6 I

7. Deen, WM,Bohrer MP,andBrenner BM. Macromolecule transport across glomerular capillaries: application of pore theory. Kidney Int 16:353-365,1979  .
0 s6 ]* b6 S# m! ~; ?  r
$ u! u( [) A( b! Y! [

8. Deen, WM,Lazzara MJ,andMyers BD. Structural determinants of glomerular permeability. Am J Physiol Renal Physiol 281:F579-F596,2001 .7 ~3 P) t& G8 z

  R) l' N5 @( r! v
9. Edwards, A,Daniels BS,andDeen WM. Ultrastructural model for size selectivity in glomerular filtration. Am J Physiol Renal Physiol 276:F892-F902,1999 .

9 X8 p. C4 @+ z) |, T

10. Fraker, PJ,andSpeck JC, Jr. Protein and cell membrane iodinations with a sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a,6a-diphrenylglycoluril. Biochem Biophys Res Commun 80:849-857,1978  .% y0 S* J. H( S

# i4 K/ F, c% l! Z* L+ \0 R0 H
11. Hoare, DG,andKoshland DE, Jr. A method for the quantitative modification and estimation of carboxylic acid groups in proteins. J Biol Chem 242:2447-2453,1967 .* a4 `* {5 B; e0 V9 b4 ?& X$ u9 X
" J& {' l" N) C

12. Lavrenko, PN,Mikriukova OI,andOkatova OV. On the separation ability of various Ficoll gradient solutions in zonal centrifugation. Anal Biochem 166:287-297,1987  .
' P8 y6 G, F* f

3 J* x1 V  ~* n. b. T% j
13. Lindström, KE,Johnsson E,andHaraldsson B. Glomerular charge selectivity for proteins larger than serum albumin as revealed by lactate dehydrogenase isoforms. Acta Physiol Scand 162:481-488,1998  .
% ?' ^6 r4 B% ]  b* e
8 {; Z8 k. n/ o2 T5 v% {: R+ R+ [
+ u' A) M* D3 r, q" @. J
14. Maack, T,Hyung Park C,andCamargo MJF Renal filtration, transport and metabolism of proteins.In: The Kidney: Physiology and Pathophysiology (2nd ed), edited by Seldin DW,and Giebisch G.. New York: Raven, 1992, p. 3005-3038.& I9 v( j* Q- `# M: i
' i" d" }. G* R, L: W
$ w; u; o; Y# k( k0 i5 f
$ n/ O) V( E  E! C
15. Maddox, DA,Deen WM,andBrenner BM. Glomerular filtration.In: Handbook of Physiology. Renal Physiology. Bethesda, MD: Am. Physiol. Soc, 1992, sect. 8, vol. I, chapt. 13, p. 545-638.: `8 g$ _1 [1 h, I7 g: J
  n1 w; j% F4 G% l
7 @. y8 b1 c9 q: j( v# {

16. Mason, EA,Wendt RP,andBresler EH. Similarity relations (dimensional analysis) for membrane transport. J Membr Sci 6:283-298,1980.$ G- D% k' ?$ ?0 }: O
9 h: o1 t: S1 r
+ x5 O3 p. p6 v/ n7 T

17. Munch, WD,Zestar LP,andAnderson JL. Rejection of polyelectrolytes from microporous membranes. J Membr Sci 5:77-102,1979.9 \5 ~( }1 p" H0 P9 l% ^+ `; l+ E& x
! ^8 k7 x7 B! e

18. Ohlson, M,Sörensson J,andHaraldsson B. A gel-membrane model of glomerular charge and size selectivity in series. Am J Physiol Renal Physiol 280:F396-F405,2001 .
+ w7 @. j2 @+ ^6 C

" I* V9 e" {3 E# h0 b: N3 [
19. Ohlson, M,Sörensson J,Lindström K,Blom AM,Fries E,andHaraldsson B. Effects of filtration rate on the glomerular barrier and clearance of four differently shaped molecules. Am J Physiol Renal Physiol 281:F103-F113,2001 .. N" j  y- R! z4 `
$ a8 r' ], n3 B% X# `
2 M1 ~/ d" B/ R  i

20. Patlak, CS,Goldstein DA,andHoffman JF. The flow of solute and solvent across a two-membrane system. J Theor Biol 5:426-442,1963  .  F' n' T: Y! e: A. A+ H8 r
/ ]& O% `/ P& n0 ^4 w, X
9 B* b" L% j- [4 E1 y
9 w0 ]/ n% i$ s/ ?% G5 @+ \8 ], j
21. Renkin, EM,andGilmore JP. Glomerular filtration.In: Handbook of Physiology. Renal Physiology. Bethesda, MD: Am. Physiol. Soc, 1973, sect. 8, chapt. 9, p. 185-248.

, T0 P# X  a& ~

22. Rennke, HG,Patel Y,andVenkatachalam MA. Glomerular filtration of proteins: clearance of anionic, neutral, and cationic horseradish peroxidase in the rat. Kidney Int 13:278-288,1978  .# n9 F  v& w  n. S2 L/ q

* A6 \2 Y* r: a. P  _# k# p6 a5 q
/ T/ B8 h: F4 H. s) v7 a. b
23. Rippe, B,andHaraldsson B. Transport of macromolecules across microvascular walls. The two-pore theory. Physiol Rev 74:163-219,1994 .
- f, _/ R; \$ ^  k
3 w0 `- ]# F4 L) s: v

24. Rodewald, R,andKarnovsky MJ. Porous substructure of the glomerular slit diaphragm in the rat and mouse. J Cell Biol 60:423-433,1974 .
7 S: U/ {1 p+ [6 I

25. Rostgaard, J,andQvortrup K. Sieve plugs in fenestrae of glomerular capillaries site of the filtration barrier? Cells Tissues Organs 170:132-138,2002  .
/ D/ }# V! C! J& i' N( R4 B4 v7 R/ `  \

! B/ {5 R: m# J
26. Russo, LM,Bakris GL,andComper WD. Renal handling of albumin: a critical review of basic concepts and perspective. Am J Kidney Dis 39:899-919,2002  ./ B: m+ `9 U. D2 E5 S3 t

8 y  |1 f# p( l) ?

27. Shah, G,andDubin PI. Adsorptive interaction of ficoll standards with porous glass size-exclusion chromatography columns. J Chromatogr A 693:197-203,1995.6 k, Q: f4 E! G0 E

$ r9 d  ^6 e2 i2 p& X
7 A0 Q3 n2 ?6 E$ l# l$ S" B; x
28. Smith, FG,andDeen WM. Electrostatic double-layer interactions for spherical colloids in cylindrical pores. J Colloid Interface Sci 78:444-465,1980.3 t" ]0 K+ {" ^; O% _4 F" }0 O9 [/ f
( q/ {( ~0 d1 l& Z. V
0 u5 V) W( F) U( H1 \1 U
( V3 d& e$ b6 v+ E
29. Smith, LH,McCarty PL,andKitanidis PK. Spreadsheet method for evaluation of biochemical reaction rate coefficients and their uncertainties by weighted nonlinear least-squares analysis of the integrated Monod equation. Appl Environ Microbiol 64:2044-2050,1998 .0 E1 N& C& V: C0 ~

2 j) Z. r3 V  b7 |
30. Sörensson, J,Ohlson M,andHaraldsson B. A quantitative analysis of the glomerular charge barrier in the rat. Am J Physiol Renal Physiol 280:F646-F656,2001 .
5 Z7 U% ]$ I# V& K

% L2 l# z, C* n
31. Sörensson, J,Ohlson M,Lindström K,andHaraldsson B. Glomerular charge selectivity for horseradish peroxidase and albumin at low and normal ionic strengths. Acta Physiol Scand 163:83-91,1998  .: ?, M$ H* X, S* E; Q, y

3 b  d& \1 S: e+ ?; i4 ^! w

32. Tay, M,Comper WD,andSingh AK. Charge selectivity in kidney ultrafiltration is associated with glomerular uptake of transport probes. Am J Physiol Renal Fluid Electrolyte Physiol 260:F549-F554,1991 .
2 Z& ]; S1 Y* r) I) W! B7 R
4 q7 d2 d) f0 k! W, l3 ^! t
  q2 z1 B! R' N. e4 |3 c
33. Taylor, AE,andGranger DN. Exchange of macromolecules across the microcirculation.In: Handbook of Physiology. The Cardiovascular System. Microcirculation. Bethesda, MD: Am. Physiol. Soc, 1984, sect. 2, vol. IV, pt. 1, chapt. 11, p. 467-520., u* F4 B" J( _  b( U
/ q: e) L' z# y- ?2 z2 u: i) `0 O) a
4 l8 ]; T$ E* P% g8 d3 u

34. Tencer, J,Frick IM,Öqvist BW,Alm P,andRippe B. Size-selectivity of the glomerular barrier to high molecular weight proteins: upper size limitations of shunt pathways. Kidney Int 53:709-715,1998  .
$ Q. \2 u* M/ L/ R: Y; j# w3 y

35. Tenstad, O,Roald AB,Grubb A,andAukland K. Renal handling of radiolabelled human cystatin C in the rat. Scand J Clin Lab Invest 56:409-414,1996  .; ^# k5 s( x- l. s
9 X( r2 J% c& |9 x
$ X7 B" b! N: c1 G' p

36. Tenstad, O,Williamson HE,Clausen G,Øien AH,andAukland K. Glomerular filtration and tubular absorption of the basic polypeptide aprotinin. Acta Physiol Scand 152:33-50,1994  .

37. Tojo, A,andEndou H. Intrarenal handling of proteins in rats using fractional micropuncture technique. Am J Physiol Renal Fluid Electrolyte Physiol 263:F601-F606,1992 .0 W) e- n  R0 M0 O# p
, u: `, x8 t) [
$ _: m8 K% @0 g% ~2 T5 E' _& e: P! i
, t) n6 N8 }# T% d2 s2 Z
38. Tryggvason, K. Unraveling the mechanisms of glomerular ultrafiltration: nephrin, a key component of the slit diaphragm. J Am Soc Nephrol 10:2440-2445,1999 .

39. Walsh, S,andDiamond D. Non-linear curve fitting using microsoft excel solver. Talanta 42:561-572,1995.
; ^$ z0 V5 e" ]" `; O) J7 c4 ~2 Q

41. Waniewski, J,Heimbürger O,Werynski A,andLindholm B. Aqueous solute concentrations and evaluation of mass transport coefficients in peritoneal dialysis. Nephrol Dial Transplant 7:50-56,1992 .- @3 c& w% m4 C8 _" f
1 F* p0 |9 K6 c1 D! k7 X- z

8 `/ d0 T) _( s4 [
42. Vyas, SV,Parker JA,andComper WD. Uptake of dextran sulphate by glomerular intracellular vesicles during kidney ultrafiltration. Kidney Int 47:945-950,1995  .

收藏0 分享0 顶0 踩0