- Email: [email protected]
Concentrative transport in human renal biopsy specimens
Concentrative Transport in Human Renal Biopsy Specimens
RICHARD P. WEDEEN, M.D. Jersey City, New Jersey
From the Department of Medicine, Jersey City Medical Center, Jersey City, New Jersey; and The New Jersey Medical School, College of Medicine and Dentistry of New Jersey, Newark, New Jersey. This study was presented in part at the Vllth International Congress of Nephrology, June 8-14, 1975. Florence, Italy. It was supported by National Institutes of Health Research Project Grant AM16266. Requests for reprints should be addressed to Dr. Richard P. Wedeen, Department of Medicine, Jersey City Medical Center, Jersey City, New Jersey 07304. Manuscript accepted June 10, 1976.
In vitro concentrative transport of tritiated p-aminohippuric acid (3H-PAH) was evaluated in renal biopsy specimens from human subjects by section freeze-dry autoradiography. Renal cortical tissue obtained by percutaneous needle biopsy (23 of 27 tissue specimens) was incubated in vitro using a modified renal cortical slice technic. Following incubation in medium containing 3H-PAH, the biopsy fragment was snap-frozen, and section freeze-dry autoradiographs were prepared. The autoradiographs demonstrated cellular accumulation of 3H-PAH by proximal tubules of human kkiney in vitro and suggest that PAH transport is uphill across the luminal as well as the antiluminal membrane. The number of proximal tubules which accumulate 3H-PAH in vitro diminishes as renal failure progresses. Cellular accumulation of PAH, however, was not impaired by heavy proteinuria, and appeared not to be related to the underlying glomerular histologic diagnosis. The frequency of proximal tubules showing 3H-PAH uptake in the cortex varied directly with the glomerular filtration rate and inversely with the degree of interstitial involvement. Proximal tubules whtch maintain the ability to accumulate 3H-PAH are clustered in fewer regions of the cortex in patients with reduced glomerular filtration rates. This observation is consistent with the view that residual renal function in renal failure arises from a few relatively normal nephrons rather than from a more uniformly damaged population of nephrons. Relatively little attention has been paid to renal tubules in evaluating the pathogenesis of acquired kidney diseases. Recently, immunofluorescent studies suggested that proximal tubules participate in immunologically-mediated renal disease [I], but the functional implications of the immunoglobulin deposition remain unclear. Understanding of the tubular contribution to clinical renal disease would be extended if specific tubular epithelial transport functions could be related to histologic abnormalities [2]. In order to obtain correlations between function and morphologic findings, autoradiography of tritiated p-aminohippuric acid (3H-PAH) uptake in renal biopsy specimens from human subjects was undertaken. These studies show that the frequency of proximal tubules demonstrating in vitro uptake of 3H-PAH diminishes as the glomerular filtration rate decreases but that it appears to be independent of the underlying glomerular histopatholoSY.
February 1977
The American Journal of Medicine
Volume 62
243
CONCENTRATIVE
TABLE
I
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
3H-PAH Uptake in Slices and Biopsy-Size
Fragments of Rat Kidney
Biopsy Incubation Medium
Time (min)
No.
Weight (mg)
C&T KRB KRB
180 60 180
24 13 6
13.0 f 6.5 10.4 * 5.5 9.9 +3.8
Slice
S:M,
S: M,,
Weight (ms)
S: M,
S: M,
7.3 + 2.3 7.1 + 1.7”
8.0 ? 2.9 6.5 f 1.2 12.8 + 2.3
87.9 + 45.0 178.1 i 39.3 142.7 ?;20.6
15.9 + 2.6 10.2 * 2.0 19.5 + 4.9
18.0 ? 3.8 10.5 * 2.2 19.8 k4.7
NOTE: C&T = Cross and Taggart phosphate buffer, 100 per cent oxygen in gas phase. KRB = Krebs-Ringer bicarbonate buffer with 95 per cent oxygen and 5 per cent carbon dioxide in gas phase. Comparison of methods for measuring S:M ratios in cortical slices (four per vial) and biopsy-size fragments (one per vial) of rat kidney. The drop in counts method (S:Md) results in values comparable to the tissue extraction method (S:M,). S:M, values have been corrected for recoveries less than 100 per cent by adding the counts lost to the measured tissue counts. Single cortical biopsy fragments consistently give lower S:M ratios than do slices from the same kidneys. Incubations were conducted in Cross and Taggart phosphate (C&T) or Krebs-Ringer bicarbonate (KRB) buffer with IO mM acetate and 1.5 mg/lOO ml ‘H-PAH at 25°C. Data are expressed as mean f standard deviation. *Seven tissues prepared for autoradiography unavailable for direct measurement of tissue 3H-PAH concentration; No. = 6.
METHODS Kidney 3H-PAH uptake studies were performed on 27 biopsy specimens obtained from 25 patients with chronic renal disease. In two patients, renal cortical tissue was obtained by open surgical biopsy, and in two additional patients with end-stage renal disease by nephrectomy. The remaining 23 biopsy specimens were obtained by the percutaneous needle technic. Fifteen of these specimens were obtained during infusion of 50 to 200 ml 50 per cent sodium iothalamate (Conray@, Mallinckrodt, Inc., Montreal, Canada) under image-intensification fluoroscopy. Renal biopsy specimens were obtained only for routine clinical indications after informed consent had been obtained from each patient. The inner fragment of needle biopsy specimens (2 to 10 mg) was used for in vitro incubation. The outer portion of the specimens-that part most likely to contain glomeruli-was preserved for diagnostic microscopy. The glomerular filtration rate was determined by 24-hour creatinine clearances using a Technicon Autoanalyzer@, except in four cases (Cases 11, 12A, 128, 21) in which the glomerular filtration rate was determined by standard inulin and/or iothalamate-‘251 clearances, as previously described [2]. Twenty-four hour urinary protein excretion was determined by the sulfosalicylic acid method. Tissue Incubation. Renal cortical slice incubation technics were modified so that tissue to medium 3H-PAH concentration ratios could be measured without destruction of the biopsy fragment. Instead of using four outer cortical slices as previously reported from this laboratory [3], only a single fragment, approximately 5 mm in length and 1 mm in diameter, was available for incubation from needle biopsy specimens. A 50 by 15 mm test tube containing a 2 mm diameter glass mixing bead, 0.5 ml of buffer and 3H-PAH was usedfor each biopsy fragment incubation. After completion of the incubation, the tissue was weighed on buffer-soaked filter paper in a Roller-Smith balance (50 mg scale), snap-frozen and prepared for section freeze-dry autoradiography. Initially, the medium selected by Cross and Taggart [4] to maximize PAH uptake was used for incubations. Investigations conducted in this laboratory on rat kidney cortex slices indicated that the Cross and Taggarl phosphate buffer had
244
February 1977
The American Journal of Medicine
several distinct disadvantages [5]. However, Krebs-Ringer bicarbonate buffer (KRB) sustained optimal PAH uptake when 10 mM acetate was added (Table I). Consequently, the acetate KRB medium replaced Cross and Taggart buffer in later studies. Either 1.5 or 6 mg/lOO ml 3H-PAH (p-aminohippuric-acid [glycyl-2-3H] specific activity 91 to 233 mCi/mmol, New England Nuclear, Boston, Mass.) was added to the medium, and incubations were conducted for 60 or 180 minutes at 25OC. No central anoxic region free of 3H-PAH uptake [6] was encountered under these incubation conditions in rat or human cortical fragments. The longer incubation and lower medium PAH concentration were selected as a standard technic in order to approximate the plasma PAH concentrations used in clearance measurements in vivo and to maximize PAH uptake ratios. Because only a single small fragment of kidney tissue was available from needle biopsy specimens, traditional slice methods for determining PAH uptake had to be modified to preserve the same tissue for autoradiography. In renal biopsy fragments, uptake of 3H-PAH was calculated by assuming that the decrease in medium tritium concentration was caused solely by tissue accumulation of 3H-PAH. The uptake ratio determined by this drop in counts method (designated S:Md) was calculated only when the medium PAH concentration was 1.5 mg/lOO ml. Tritium was measured in a Packard liquid scintillation spectrophotometer (Model No. 3244), as previously described [3]. The validity of determining S:M ratios without destruction of tissue during analysis and the reliability of biopsy-size cortical fragments for measurements of 3H-PAH uptake were determined in rat kidneys. The S:Md calculated by the drop in counts method (Table I) closely approximattid the slice to medium concentration ratio obtained by the usual tissue extraction method (S:M& However, renal cortical fragments (approximately 10 mg) consistently gave lower S:Md ratios than four slices (approximately 125 mg) from the same rat kidneys. Whether the reduced uptake of small single cortical fragments was caused by physical (e.g., mixing, surface damage) or metabolic (e.g., endogenous inhibitor production) factors is not apparent from these studies. Table I demonstrates that the biopsy fragment incubation technic is ade-
Volume 62
CONCENTRATIVE
quately reproducible in rat despite the reduction in 3H-PAH uptake by small tissue samples. Autoradiography. The method of section freeze-dry autoradiography devised in this laboratory for cellular localization of diffusible molecules has been described elsewhere [7]. This method minimizes diffusion of water-soluble molecules by preventing contact with solvents prior to developing the autoradiograph. Briefly, following in vitro incubation in medium containing 3H-PAH, the tissue is snap-frozen at liquid nitrogen temperature. Sectioning and freeze-drying are accomplished in a closed cryostat operating at about -50%. The dried sections are brought to room temperature in vacua and placed against microscope slides precoated with liquid nuclear track photographic emulsion (Kodak NTB-2, Eastman Kodak Co., Rochester, New York; or Ilford L-2, Polyscience, Inc., Warrington, Penn.). Exposures were usually from seven to 14 days, depending on the cellular 3H-PAH concentration determined by intermittent developing of sample autoradiographs. Following development of the autoradiographs and gold latensification, the adherent tissues were stained by the periodic acid-Schiff and hematoxylin method. Autoradiographic and Histologic Grading. The number of tubules showing 3H-PAH accumulation in a low power field (4X objective lens) was graded by inspection of the autoradiographs (Table II). If virtually all proximal tubules in a field showed concentrative transport, the autoradiograph was graded “4+.” If only a rare cluster of proximal convolutions showed 3H-PAH uptake, the biopsy specimen was graded “1-t.” Autoradiographic grain density was not used in this grading, since this is determined by a number of variables in addition to the tritium concentration in the tubule, e.g., exposure time, tissue thickness, film thickness, gold latensification and latent image fading. The degree of tubular-interstitial disease in the tissue adherent to the autoradiograph was similarly graded from 4+ (normal) to 0 (severe) (Table II). Interstitial and tubular involvement was assessed in terms of flattening of proximal tubular epithelium, loss of brush border, tubular basement membrane thickening, inflammatory cell infiltration, interstitial fibrosis and loss of tubules. Cortex showing occasional areas of tubular-interstitial involvement was graded 3-l-. When no proximal tubules in a field appeared normal and extensive interstitial fibrosis was present, the tissue was graded “0.” Fixed tissue prepared for light microscopy was used to confirm the interstitial and tubular findings seen in the freezedried sections. The grading of interstitial involvement was made without regard to glomerular or vascular histopathology and does not imply the primacy of the tubular-interstitial process. Nine of the 27 biopsy specimens were obtained from patients with apparent nonglomerular disease, including five with nephrosclerosis associated with hypertension, two from a patient with lead nephropathy (Case 1, [2]), and one each with primary hyperparathyroidism and obstructive pyelonephritis. RESULTS
3H-PAH Uptake Ratios in Kidney Biopsy Specimens from Human Subjects. Diagnoses and clinical characteristics of all patients are presented in Table Ill. The 3H-PAH uptake ratios determined by the drop in Counts
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Histologic
Grading Criteria
TABLE
II
Grade
“H-PAH Uptake Tubular-interstitial Disease Frequency ___.___-._-.__~.____-.___~_.._~_
4+
3+
2+
All proximal tubules show 3H-PAH uptake (except at edges of biopsy) Small groups of proximal tubules show no 3H-PAH uptake Less than half of the tubules show 3HPAH uptake
I+
Rare proximal tubules show 3H-PAH uptake
0
No cellular accumulation of ‘H-PAH seen
No abnormality
Occasional areas of tubular damage and interstitial fibrosis Less than half of the renal cortical tissue on slide shows tubular damage and interstitial fibrosis; most proximal tubules appear normal More than half of renal cortical tissue on slide shows tubular damage, interstitial fibrosis and inflammatory cell infiltration Extensive tubular damage, interstitial fibrosis and inflammatory cell infiltration are present; many tubules are dilated or obsolete
method showed little relationship to autoradiographic patterns of tritium distribution. In four biopsy specimens in which two to four biopsy fragments were studied separately, considerable variability in S:Md was evident in each biopsy fragment obtained from the same kidney. In one patient (Case 11) in whom tissue was obtained by the open surgical technic, four separate biopsy-size fragments gave 3H-PAH S:Md values of 0.4, 0.8, 1.6 and 2.9. These repetitive measurements of uptake showed such variability as to render the measurement of S:Md ratios in human biopsy fragments unreliable. When carefully selected cortical slices are used from open biopsy material, more reproducible results may be obtained [ 81. Autoradiography. Section freeze-dry autoradiographs from human renal biopsy specimens demonstrated selective proximal tubule accumulation of tritium following in vitro incubation in media containing 3H-PAH. Considerable variation was evident in segmental uptake of 3H-PAH even in apparently normal human proximal tubules (Figure 1). Glomeruli, interstitium and distal tubules showed no concentrative transport of 3H-PAH. Some proximal tubule lumens in each tissue showed grain densities which were greater than that in surrounding epithelial cells (Figures 2D and 3). When the frequency of proximal tubule uptake of 3H-PAH is graded 0 to 4-I, no correlation between uptake frequency and specific diagnosis, blood pressure or amount of proteinuria was evident (Table Ill). On the other hand, definite correlations were found to exist between frequency of 3H-PAH uptake, glomerular filtration rate and the degree of interstitial involvement
February 1977
The American Journal of Medicine
Volume 62
245
Histologic
Age
Extracapillary Pyelonephritis
Nephrosclerosis Nephrosclerosis End stage
End stage Nephrosclerosis
20 21
22 23 15B
24 25
0t
.
68 93 75 & (29) (86)
160/110 160/100 170/100 150/100 170190 140190 2401140
.
9 5
0.6 0.4
40 39
. . .
3.8 0.0
0.9 0.8
ot
ot
25 15
30 25
32
170/110 loo/70
0.2
35
32
52 48 36
120170
0.1
0.0 2.4 3.6
55
65
67
92
100
101
58
43 24 44
6.5
0.0
0.0
2.9
1 .1
0.0
109
117
0.2
0.1
144 125
150 146
154
0.0 3.4
10.5 12.9
1.5
-
0 0 0
1+ 1+
0 0 0 0 0
1+ It
2+ 2t
2t
2t
0 0 2t
2+
3+
3+
2+
1+ 0
0 Ii
2+
2+
1+ 0 1+
1+
2+
2+
2+
0.6
2.7
0.0 0.4
0.8 1 .o
1.9
. . . . .
. ’
. .
. . . .
.
2.9, 1 .6, 0.8,0.4 .
. .
f GFR assumed to approximate
0 in patients
undergoing
dialysis.
__--
.
= 0.72
at 25°C).
.
.
0.0 . . . Chronic Chronic Chronic
hemodialysis hemodialysis hemodialysis
..
Heroin user Postobstruction, infection, EPAH = 0.31
. I . .
Lupus erythematosus
Poststreptococcal
Lead colic, EPAH Hyperparathyroid . . .
Postchelation therapy, EPA,_, = 0.82 Lupus erythematosus
K,) Gold nephropathy, rheumatoid arthritis Gold nephropathy, rheumatoid arthritis E PAH = 0.31
Gold nephropathy in remission, rheumatoid arthritis Eight years poststreptococcal acute GN Low serum complement
Multiple allergies Heroin user
.
.
hematuria
. .
Unexplained
..
Intermittent proteinuria Steroid resistant
Comment
. . .
.
.
2.9
‘ii
. . .
. .
. . .
.
11.3
4+
3+
. .
1.7
0
4+
.
. . . .
. . .
4.4
180 min
4+
.
.
3.1,2.9,0,8 . *.
4.4,‘d.i
60 min
Incubation Time
2+
4+
3+ 3+
3+ 3+
4+
Autoradiograph
2+
4+
4+ 2+
4+ 1+
4+
GlomerularTubularFiltration InterRate stitial (mllmin) Disease
160190
120180 200/l 10 200/100
-
Proteinuria (gldav)
NOTE: Leaders indicate study not done. S:Md was not measured when 6 mg/lOO ml 3H-PAH was used. GN = glomerulonephritis. +S:Md = the slice to medium ‘H-PAH concentration ratio determined by the drop in counts method (I .5 mg/lOO ml ‘H-PAH incubated
17 36
46 38 74
19 56
26 27
Nephrosclerosis Membranoproliferative G N
18 19
GN
11
Membranoproliferative GN
52
17
16
nephropathy Acute proliferative GN
28 54 73
12A 14 15A
57
nephropathy Lead nephropathy Nephrocalcinosis Membranous
1751115
39
Membranous
47
13
29
120180
Lead nephropathy
28
128
46
165/100
Nephrosclerosis
11
18
90160
Minimal change
10 53
36
Minimal change
9
13 13
130170
28
8
13
120180
130/80
16
7
Healed proliferative GN Focal GN
13
6
120170
38
Membranous
4 5
nephropathy
40 32
nephropathy Normal kidney Membranoproliferative GN
21 15 15 17
105/70 11 O/70
7 28
Minimal change Membranous
12
Blood Urea Nitrogen hog/ 100 ml)
Data in Patients
130180 120170
120180
-
Blood Pressure (mm Hg)
(yr)
Uptake
and ‘H-PAH
13
2 3
Diagnosis
Clinical,
Minimal change
III
1
Case No.
TABLE
CONCENTRATIVE
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Figure 1. Autoradiograph showing 3H-PAH uptake by all proximal tubules, grade 4i-, except damaged segments at edge of biopsy specimen (Case 6). Electron microscopy demonstrated thickening of giomerular basement membranes. Tubules and interstitial regions were morphologically within normal limits (grade 4+). Periodic acid-Schiff and hematoxylin stain; original magnification X 76, reduced by 50 per cent.
(Figure 4). Fewer proximal tubules per low power field accumulated 3H-PAH in the presence of reduced glomerular filtration rate and increased interstitial disease. Proximal tubule uptake could not be predicted from tubular morphology in the freeze-dried sections since adjacent tubules with comparable structural changes could be associated with intense to absent autoradiograph grain density (Figures 1, 26, 2D, 2F and 3). COMMENTS
The most striking observation in this study is that the frequency of proximal tubular uptake of 3H-PAH correlated directly with the glomerular filtration rate and inversely with tubular-interstitial disease (Figure 4). A correlation between glomerular filtration rate and the degree of tubular-interstitial involvement irrespective of glomerular histopathology has been recognized [9]. Regardless of the etiology of the renal disease, the lower the glomerular filtration rate and the greater the tubular-interstitial involvement, the fewer the proximal tubules that demonstrated cellular accumulation of 3H-PAH in the autoradiographs. In the presence of renal failure, the persistence of 3H-PAH uptake in only isolated groups of proximal tubules is consistent with the view that a few relatively normal nephrons support residual renal function rather than a more uniformly damaged population of nephrons. In the presence of reduced glomerular filtration rate and interstitial changes, broad areas of each kidney cortex section showed no uptake of PAH. It has previously been shown that serum from uremic subjects inhibits PAH uptake in kidney cortex slices [lo]. Since some tubules showed persistence of 3H-PAH uptake even in azotemic patients (Figure 2), failure of concentrative transport was apparently not due to competitive inhibitors of PAH found in the serum of uremic patients. Inhibitors of PAH uptake in such serum would be expected to produce uniform effects on all nephrons; such inhibition of uptake should have produced a ho-
mogeneous autoradiographic pattern. On the contrary, in tissue obtained from azotemic patients, proximal segments demonstrating 3H-PAH uptake tended to be clustered in isolated areas (Figures 2C and 2E). Moreover, patients undergoing hemodialysis with end-stage renal disease did not show improved tubular uptake of 3H-PAH due to dialysis (Table Ill). Endogenous organic molecules within the biopsy fragments which act as competitive inhibitors of PAH transport are apparently greatly diluted by the 1,OOO-fold excess volume of the incubation medium. The failure of proximal tubules to accumulate 3H-PAH similarly cannot be attributed to a diffusion barrier produced by interstitial fibrosis or tubular basement membrane thickening. Concentrative transport was often demonstrable deep within a biopsy fragment despite broad areas of no uptake in adjacent but more superficial regions. If the widened interstitium had prevented 3H-PAH from reaching some proximal tubule segments, this effect would have been most apparent in the regions of the tissue that were more remote from the incubation medium. The long incubation times and relatively high 3H-PAH concentrations used in this study insured full penetration of the tissue by 3H-PAH. Moreover, proximal tubule segments with comparable basement membrane thickening and surrounding interstitial fibrosis showed variations in uptake that could not be explained by morphologic changes alone (Figures 2 and 3). Cellular damage due to the biopsy needle appeared to be limited to the surface of the specimen so that interference with concentrative transport due to gross cellular trauma was limited to the outer edges of the tissue specimen (Figure 1). Linear fractures apparent in the interior of the tissue were produced during frozen tissue sectioning. Such sectioning artifacts occur after the in vitro incubation and, thus, could not alter the concentrative transport process. The loss of PAH uptake by proximal tubules evident in the autoradiographs thus appears to be caused by cellular damage accompanying renal failure.
February 1977
The American Journal of Medicine
Volume 62
247
CONCENTRATlVE TRANSPORT IN HUMAN RENAL BlOPSY SPECIMENS-WEDEEN
Figure 2. Autoradiographs demonstrating intermediate grades of 3H-PAH uptake. A and B, Case I 7: 3H-PAH uptake grade 3+, tubular-interstitial disease grade 2+. C and D, Case 17: 3H-PAH uptake grade 2+, tubular-interstitial disease grade 2+. E and F, Case 2 I: 3H-PAH uptake grade I+. tubular-interstitial disease grade 0 (see Table II).
248
February 1977
The American Journal of Medicine
Volume 62
CONCENTRATIVE TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Figure 3. Autoradiographfrom hypertensive patient (Case 71) demonstrating cellular accumulation of 3H-PAH in proximal tubules (3+). Higher 3H-PA H concentrations are present in some lumens of the proximal tubule than in cells. Moderateinterstitial fibrosis, mononuclearcell infiltration and tubular basement membrane thickening is present (grade 2+). Periodic acid-Schiff and hematoxylin stain; original magnification X 795, reduced by 50 per cent.
The correlation of the frequency of 3H-PAH uptake with the glomerular filtration rate and tubular-interstitial disease contrasts with the independence of 3H-PAH uptake from proteinuria. Although tubular reabsorption and catabolism of proteins is an important proximal tubular function [ Ill, an increased tubular protein load in vivo does not appear to interfere with accumulation of 3H-PAH in vitro. As renal disease progresses, fewer tubules participate in concentrative transport of PAH, whatever the underlying etiology. Since the nontransporting nephron segments are not necrotic, they must continue to receive some blood supply. PAH contained in the blood reaching nephrons which have lost the ability to accumulate 3H-PAH is presumably not extracted. The failure to secrete PAH undoubtedly contributes to the decrease in PAH extraction associated with renal failure [ 121.
PAH extraction ratios (E~AH)obtained at the time of renal biopsy in four of our cases (Table III) failed, however, to show any consistent correlation with glomerular filtration rate, grade of tubular interstitial disease or autoradiographic uptake pattern. These few measurements of EPAHdo not suggest any simple correlation with 3H-PAH uptake in vitro. This is, perhaps, not surprising since EPAH appears to be determined by a number of factors other than cellular transport processes and glomerular filtration rate [ 131. The finding that cellular accumulation of 3H-PAH in vitro correlated with changes in PAH extraction &d maximal transport rate (TmpAH) in vivo in a case of occupational lead nephropathy, reported in detail elsewhere [ 21, remains unique and is presumably related to specific proximal tubule effects of lead. The relationship between cellular accumulation of PAH and transepithelial transport remains to be defined.
-
PAH-3
H Uptake
C-4Intrrstitial
Frequency
Disease
160 .
140
Figure 4. Relationship of 3H-PAH uptake frequency and interstitial disease to glomerular filtration rate. Uptake frequency (circles) and tubular-interstitial disease (triangles) were graded0 to 4+ ( Table IIf). The glomerular filtration rate was considered to be 0 in three chronic hemodialysis patients. Both 3H-PAH uptake frequency and tubular-interstitial disease correlate with the glomerular filtration rate. The calculated linear regression for uptake frequency is plotted as a solid line and that for tubular-interstitial disease as a broken line. Both regressions had correlation coefficients of 0.8 and T values of 6.
120
-;
100
.
-i a IL
60 60
(3 40
0
0
I+
AUTORADlOGRAPHlC
February 1977
2+
3+
AND HISTOLOGIC
The American Journal of Medicine
4t GRADE
Volume 62
249
CONCENTRATIVE TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Precise measurements of 3H-PAH concentrations in proximal tubule cells would require a more quantitative approach to autoradiography than has been feasible with freeze-dried sections. Grain counting has not proved useful in these preparations. In addition to the well known technical difficulties inherent in quantitative autoradiography [ 141, the histologic preparations used in this study present formidable problems for grain counting. The usual variables (tissue-film contact, latent image fading, silver halide saturation, depth of focus and background correction) are compounded by problems introduced by the freeze-dried sections. Freeze-drying sections minimize diffusion of water-soluble labeled compounds at the expense of introducing variability of tissue section thickness, cutting artifacts and possible physical displacement of 3H-PAH during sectioning and freeze-drying [7]. Quantitative autoradiography of tubular cell uptake is further handicapped by the obvious segmental variation in 3H-PAH uptake within even the normal proximal tubule (Figure 1). Some of the factors which interfere with grain counting can be overcome by using embedded freeze-dried tissue blocks [ 151 but only at the risk of permitting redistribution of labeled compounds due to solubility and water permeability of the embedding medium. Assessment of 3H-PAH uptake in human renal biopsy specimens by determining S:M ratios has proved equally futile. Minor modification of traditional slice incubation technics permitted reproducible measurement of tissue uptake in biopsy-size fragments of rat renal cortex. Comparable fragments from human renal biopsy specimens did not, however, show reproducible S:M ratios. The inability to obtain reliable 3H-PAH uptake data on human biopsy specimens in vitro appears to be caused by the sampling problems inherent in the needle biopsy technic. Portions of this minute biopsy material often contain capsule, medulla, perirenal fat or excretory tract epithelium. The diseased kidney, in addition, has regions of scarring adjacent to regions of relative preservation of morphologic integrity. The range of variation in S:M uptake data from diseased kidneys proved too great to permit reliable uptake measurements in single human biopsy fragments. The sampling errors inherent in needle biopsy material interfere with autoradiographic evaluation far less than with the determination of S:M ratios. Noncortical tissue can readily be excluded from analysis on morphologic grounds in the autoradiographs, thereby greatly reducing observational variability. Biopsy material containing no renal cortex can easily be excluded. On the other hand, a tissue fragment containing only a minute portion of cortex and therefore providing invalid S:M ratios may, nevertheless, contain enough proximal tubules to evaluate concentrative transport by autoradiography. Errors due to sampling would be expected to contribute
250
February 1977
The American Journal ol Medicine
to the considerable variation in individual autoradiograph grades when compared to whole kidney glomerular filtration rate (Figure 4). The 23 percutaneous renal biopsy specimens discussed here were selected from a consecutive series of 30 in vitro biopsy incubations because renal cortical tissue was present for autoradiographic evaluation. Of the seven biopsy specimens excluded, two were lost because of early technical problems in specimen handling, four contained only renal medulla, and one contained only fibrous tissue. In addition to demonstrating a correlation between the frequency of proximal tubular uptake of 3H-PAH with glomerular filtration rate and tubular interstitial disease, these autoradiographic studies provide insight into the function of normal human proximal tubules. They demonstrate that the human proximal tubule accumulates 3H-PAH intracellularly in vitro. The finding that 3H-PAH concentrations in the lumens are occasionally higher than that in adjacent cells, as judged by inspection of autoradiograph grain density, is similar to previous observations in the rat [ 161. Luminal accumulation of 3H-PAH was present in both percutaneous needle biopsy fragments and renal cortical slices obtained by the open surgical biopsy technic. Assuming that the renal cortex in man in vitro behaves like that of the rat, the luminal channel of proximal tubules is not in open communication with the incubation medium. In the rat, neither labeled inulin nor labeled sucrose is detectable in the lumens of the proximal tubule following in vitro slice incubation [ 161. 3H-PAH thus reaches the lumen only by transport across the epithelial cell. Since the autoradiographic grain density is greater in the lumen than in the ,cell in this freeze-dried tissue (Figures 2D and 3) 3H-PAH transport appears to be uphill across the luminal as well as the peritubular membrane. The failure to demonstrate comparable uphill transport of PAH at the luminal surface in isolated perfused rabbit tubules [ 171 may relate to species differences or to the method of isolating and perfusing individual nephron segments. Moreover, apparent uphill luminal secretion in vitro is by no means universal in proximal tubules in the autoradiographs and may therefore be missed in individual perfused tubular segments. The high luminal grain density arising from 3H-PAH does not appear to be a technical artifact. Differences in tissue absorption of the beta emissions from tritium cannot explain the autoradiographic grain distribution pattern. Freeze-dried unembedded 3 p thick cells have too little density to account for the manifold reduction in autoradiograph grains underlying cytoplasm as compared to lumens (Figure 3). Even when the lumens of the proximal tubule appear to be closed or filled with proteinaceous material, increased luminal grain density is evident in apposition to the brush borders. The autoradiographic pattern, suggesting uphill transport of
Volume 62
CONCENTRATIVE
3H-PAH from cell to lumen, has consistently been evident in renal autoradiographs from man, rat and dog in this laboratory. Moreover, high luminal radioactivity has never been seen in renal cortical slices incubated with a variety of other tritium or carbon-14 labeled compounds including inulin, sucrose, iodoacetamide, dinitrophenol, ouabain, cY-aminoisobutyric acid, cycloleutine, lysine or methionine. These initial autoradiographic investigations of human tubular function have been confined to PAH, a nonmetabolized molecule, which is avidly secreted and accumulated by the proximal tubule cell. Similar morphologic-functional correlations could be undertaken with other labeled diffusible compounds which undergo
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
concentrative transport, such as nonmetabolized amino acids, sugars and antibiotics. If specific patterns of tubular cell dysfunction can be identified by section freeze-dry autoradiography, a clearer understanding of the role of interstitial and tubular changes in renal disease may emerge. ACKNOWLEDGMENT These studies were performed with the technical assistance of Ms. Claffertene Cheeks and Mr. Bhagirath Vyas. Clinical microscopy was performed by Drs. Michael Lyons, Norman Ende and Gabriel Mulcahy. Dr. Jacob Churg generously provided guidance in the grading of tubular-interstitial disease.
REFERENCES 1.
2. 3.
4.
5.
6.
7.
8.
Andres GA, McCluskey RT: Tubular and interstitial renal disease due to immunologic mechanisms. Kidney Int 7: 271, 1975. Wedeen RP, Maesaka JK, Weiner B, et al.: Occupational lead nephropathy. Am J Med 59: 630, 1975. Wedeen RP, Thier SO: The intrarenal distribution of nonmetabolized amino acids in vivo. Am J Physiol 220: 507, 1971. Cross RJ, Taggart JV: Renal tubular transport: accumulation of p-aminohippurate by rabbit kidney slices. Am J Physiol 161: 181, 1950. Wedeen RP, Weiner B: The effect of acetate on p-aminohippuric acid uptake in kidney slices. Fed Proc 33: 1103, 1974. Wedeen RP, Weiner B: The distribution of p-aminohippuric acid in rat kidney slices. II. Depth of uptake. Kidney Int 3: 214, 1973. Bordier B, Ornstein L, Wedeen RP: The intrarenal distribution of para-aminohippuric acid determined by a modified technique of section freeze-dry autoradiography. J Cell Biol 46: 518, 1970. Thier SO: Amino acid transport i:r the renal tubule cell. Recent Advances in Renal Physiology and Pharmacology (Wesson LG. Fanelli GM Jr, eds), Baltimore, University Park Press, 1974, p 39.
February
9.
10.
11. 12. 13.
14.
15.
16.
17.
1977
Schainuck LI, Striker GE, Cutler RE, et al.: Structural-functional correlations in renal disease. Il. The correlations. Human Pathol 1: 631, 1970. Preuss HG, Massry SG, Maher JF, et al.: Effects of uremic sera on renal tubular p-aminohippurate transport. Nephron 3: 265, 1966. Maack T: Renal handling of low molecular weight proteins. Am J Med 58: 57, 1975. Wesson LG, ed: Physiology of the Human Kidney, New York, Grune & Stratton, 1969, p 165. Haberle D: Influence of glomerular filtration rate on the rate of para-aminohippurate secretion by the rat kidney. Kidney Int 7: 385, 1975. Perry RP: Quantitative autoradiography. Methods in Cell Physiology, vol 1 (Prescott DM, ed), New York, Academic Press, 1964, p 305. Kinter WB, Wong M: Peritubular membrane transport of PAH into proximal cells in necturus kidney slices. Am J Physiol 227: 50, 1974. Wedeen RP, Weiner B: The distribution of p-aminohippuric acid in rat kidney slices. I. Cellular localization. Kidney Int 3: 205, 1973. Tune BM, Burg MB, Patlak CS: Characteristics of p-aminohippurate transport in proximal renal tubules. Am J Physiol 217: 1057, 1969.
The American
Journal
of Medicine
Volume 62
251
RICHARD P. WEDEEN, M.D. Jersey City, New Jersey
From the Department of Medicine, Jersey City Medical Center, Jersey City, New Jersey; and The New Jersey Medical School, College of Medicine and Dentistry of New Jersey, Newark, New Jersey. This study was presented in part at the Vllth International Congress of Nephrology, June 8-14, 1975. Florence, Italy. It was supported by National Institutes of Health Research Project Grant AM16266. Requests for reprints should be addressed to Dr. Richard P. Wedeen, Department of Medicine, Jersey City Medical Center, Jersey City, New Jersey 07304. Manuscript accepted June 10, 1976.
In vitro concentrative transport of tritiated p-aminohippuric acid (3H-PAH) was evaluated in renal biopsy specimens from human subjects by section freeze-dry autoradiography. Renal cortical tissue obtained by percutaneous needle biopsy (23 of 27 tissue specimens) was incubated in vitro using a modified renal cortical slice technic. Following incubation in medium containing 3H-PAH, the biopsy fragment was snap-frozen, and section freeze-dry autoradiographs were prepared. The autoradiographs demonstrated cellular accumulation of 3H-PAH by proximal tubules of human kkiney in vitro and suggest that PAH transport is uphill across the luminal as well as the antiluminal membrane. The number of proximal tubules which accumulate 3H-PAH in vitro diminishes as renal failure progresses. Cellular accumulation of PAH, however, was not impaired by heavy proteinuria, and appeared not to be related to the underlying glomerular histologic diagnosis. The frequency of proximal tubules showing 3H-PAH uptake in the cortex varied directly with the glomerular filtration rate and inversely with the degree of interstitial involvement. Proximal tubules whtch maintain the ability to accumulate 3H-PAH are clustered in fewer regions of the cortex in patients with reduced glomerular filtration rates. This observation is consistent with the view that residual renal function in renal failure arises from a few relatively normal nephrons rather than from a more uniformly damaged population of nephrons. Relatively little attention has been paid to renal tubules in evaluating the pathogenesis of acquired kidney diseases. Recently, immunofluorescent studies suggested that proximal tubules participate in immunologically-mediated renal disease [I], but the functional implications of the immunoglobulin deposition remain unclear. Understanding of the tubular contribution to clinical renal disease would be extended if specific tubular epithelial transport functions could be related to histologic abnormalities [2]. In order to obtain correlations between function and morphologic findings, autoradiography of tritiated p-aminohippuric acid (3H-PAH) uptake in renal biopsy specimens from human subjects was undertaken. These studies show that the frequency of proximal tubules demonstrating in vitro uptake of 3H-PAH diminishes as the glomerular filtration rate decreases but that it appears to be independent of the underlying glomerular histopatholoSY.
February 1977
The American Journal of Medicine
Volume 62
243
CONCENTRATIVE
TABLE
I
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
3H-PAH Uptake in Slices and Biopsy-Size
Fragments of Rat Kidney
Biopsy Incubation Medium
Time (min)
No.
Weight (mg)
C&T KRB KRB
180 60 180
24 13 6
13.0 f 6.5 10.4 * 5.5 9.9 +3.8
Slice
S:M,
S: M,,
Weight (ms)
S: M,
S: M,
7.3 + 2.3 7.1 + 1.7”
8.0 ? 2.9 6.5 f 1.2 12.8 + 2.3
87.9 + 45.0 178.1 i 39.3 142.7 ?;20.6
15.9 + 2.6 10.2 * 2.0 19.5 + 4.9
18.0 ? 3.8 10.5 * 2.2 19.8 k4.7
NOTE: C&T = Cross and Taggart phosphate buffer, 100 per cent oxygen in gas phase. KRB = Krebs-Ringer bicarbonate buffer with 95 per cent oxygen and 5 per cent carbon dioxide in gas phase. Comparison of methods for measuring S:M ratios in cortical slices (four per vial) and biopsy-size fragments (one per vial) of rat kidney. The drop in counts method (S:Md) results in values comparable to the tissue extraction method (S:M,). S:M, values have been corrected for recoveries less than 100 per cent by adding the counts lost to the measured tissue counts. Single cortical biopsy fragments consistently give lower S:M ratios than do slices from the same kidneys. Incubations were conducted in Cross and Taggart phosphate (C&T) or Krebs-Ringer bicarbonate (KRB) buffer with IO mM acetate and 1.5 mg/lOO ml ‘H-PAH at 25°C. Data are expressed as mean f standard deviation. *Seven tissues prepared for autoradiography unavailable for direct measurement of tissue 3H-PAH concentration; No. = 6.
METHODS Kidney 3H-PAH uptake studies were performed on 27 biopsy specimens obtained from 25 patients with chronic renal disease. In two patients, renal cortical tissue was obtained by open surgical biopsy, and in two additional patients with end-stage renal disease by nephrectomy. The remaining 23 biopsy specimens were obtained by the percutaneous needle technic. Fifteen of these specimens were obtained during infusion of 50 to 200 ml 50 per cent sodium iothalamate (Conray@, Mallinckrodt, Inc., Montreal, Canada) under image-intensification fluoroscopy. Renal biopsy specimens were obtained only for routine clinical indications after informed consent had been obtained from each patient. The inner fragment of needle biopsy specimens (2 to 10 mg) was used for in vitro incubation. The outer portion of the specimens-that part most likely to contain glomeruli-was preserved for diagnostic microscopy. The glomerular filtration rate was determined by 24-hour creatinine clearances using a Technicon Autoanalyzer@, except in four cases (Cases 11, 12A, 128, 21) in which the glomerular filtration rate was determined by standard inulin and/or iothalamate-‘251 clearances, as previously described [2]. Twenty-four hour urinary protein excretion was determined by the sulfosalicylic acid method. Tissue Incubation. Renal cortical slice incubation technics were modified so that tissue to medium 3H-PAH concentration ratios could be measured without destruction of the biopsy fragment. Instead of using four outer cortical slices as previously reported from this laboratory [3], only a single fragment, approximately 5 mm in length and 1 mm in diameter, was available for incubation from needle biopsy specimens. A 50 by 15 mm test tube containing a 2 mm diameter glass mixing bead, 0.5 ml of buffer and 3H-PAH was usedfor each biopsy fragment incubation. After completion of the incubation, the tissue was weighed on buffer-soaked filter paper in a Roller-Smith balance (50 mg scale), snap-frozen and prepared for section freeze-dry autoradiography. Initially, the medium selected by Cross and Taggart [4] to maximize PAH uptake was used for incubations. Investigations conducted in this laboratory on rat kidney cortex slices indicated that the Cross and Taggarl phosphate buffer had
244
February 1977
The American Journal of Medicine
several distinct disadvantages [5]. However, Krebs-Ringer bicarbonate buffer (KRB) sustained optimal PAH uptake when 10 mM acetate was added (Table I). Consequently, the acetate KRB medium replaced Cross and Taggart buffer in later studies. Either 1.5 or 6 mg/lOO ml 3H-PAH (p-aminohippuric-acid [glycyl-2-3H] specific activity 91 to 233 mCi/mmol, New England Nuclear, Boston, Mass.) was added to the medium, and incubations were conducted for 60 or 180 minutes at 25OC. No central anoxic region free of 3H-PAH uptake [6] was encountered under these incubation conditions in rat or human cortical fragments. The longer incubation and lower medium PAH concentration were selected as a standard technic in order to approximate the plasma PAH concentrations used in clearance measurements in vivo and to maximize PAH uptake ratios. Because only a single small fragment of kidney tissue was available from needle biopsy specimens, traditional slice methods for determining PAH uptake had to be modified to preserve the same tissue for autoradiography. In renal biopsy fragments, uptake of 3H-PAH was calculated by assuming that the decrease in medium tritium concentration was caused solely by tissue accumulation of 3H-PAH. The uptake ratio determined by this drop in counts method (designated S:Md) was calculated only when the medium PAH concentration was 1.5 mg/lOO ml. Tritium was measured in a Packard liquid scintillation spectrophotometer (Model No. 3244), as previously described [3]. The validity of determining S:M ratios without destruction of tissue during analysis and the reliability of biopsy-size cortical fragments for measurements of 3H-PAH uptake were determined in rat kidneys. The S:Md calculated by the drop in counts method (Table I) closely approximattid the slice to medium concentration ratio obtained by the usual tissue extraction method (S:M& However, renal cortical fragments (approximately 10 mg) consistently gave lower S:Md ratios than four slices (approximately 125 mg) from the same rat kidneys. Whether the reduced uptake of small single cortical fragments was caused by physical (e.g., mixing, surface damage) or metabolic (e.g., endogenous inhibitor production) factors is not apparent from these studies. Table I demonstrates that the biopsy fragment incubation technic is ade-
Volume 62
CONCENTRATIVE
quately reproducible in rat despite the reduction in 3H-PAH uptake by small tissue samples. Autoradiography. The method of section freeze-dry autoradiography devised in this laboratory for cellular localization of diffusible molecules has been described elsewhere [7]. This method minimizes diffusion of water-soluble molecules by preventing contact with solvents prior to developing the autoradiograph. Briefly, following in vitro incubation in medium containing 3H-PAH, the tissue is snap-frozen at liquid nitrogen temperature. Sectioning and freeze-drying are accomplished in a closed cryostat operating at about -50%. The dried sections are brought to room temperature in vacua and placed against microscope slides precoated with liquid nuclear track photographic emulsion (Kodak NTB-2, Eastman Kodak Co., Rochester, New York; or Ilford L-2, Polyscience, Inc., Warrington, Penn.). Exposures were usually from seven to 14 days, depending on the cellular 3H-PAH concentration determined by intermittent developing of sample autoradiographs. Following development of the autoradiographs and gold latensification, the adherent tissues were stained by the periodic acid-Schiff and hematoxylin method. Autoradiographic and Histologic Grading. The number of tubules showing 3H-PAH accumulation in a low power field (4X objective lens) was graded by inspection of the autoradiographs (Table II). If virtually all proximal tubules in a field showed concentrative transport, the autoradiograph was graded “4+.” If only a rare cluster of proximal convolutions showed 3H-PAH uptake, the biopsy specimen was graded “1-t.” Autoradiographic grain density was not used in this grading, since this is determined by a number of variables in addition to the tritium concentration in the tubule, e.g., exposure time, tissue thickness, film thickness, gold latensification and latent image fading. The degree of tubular-interstitial disease in the tissue adherent to the autoradiograph was similarly graded from 4+ (normal) to 0 (severe) (Table II). Interstitial and tubular involvement was assessed in terms of flattening of proximal tubular epithelium, loss of brush border, tubular basement membrane thickening, inflammatory cell infiltration, interstitial fibrosis and loss of tubules. Cortex showing occasional areas of tubular-interstitial involvement was graded 3-l-. When no proximal tubules in a field appeared normal and extensive interstitial fibrosis was present, the tissue was graded “0.” Fixed tissue prepared for light microscopy was used to confirm the interstitial and tubular findings seen in the freezedried sections. The grading of interstitial involvement was made without regard to glomerular or vascular histopathology and does not imply the primacy of the tubular-interstitial process. Nine of the 27 biopsy specimens were obtained from patients with apparent nonglomerular disease, including five with nephrosclerosis associated with hypertension, two from a patient with lead nephropathy (Case 1, [2]), and one each with primary hyperparathyroidism and obstructive pyelonephritis. RESULTS
3H-PAH Uptake Ratios in Kidney Biopsy Specimens from Human Subjects. Diagnoses and clinical characteristics of all patients are presented in Table Ill. The 3H-PAH uptake ratios determined by the drop in Counts
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Histologic
Grading Criteria
TABLE
II
Grade
“H-PAH Uptake Tubular-interstitial Disease Frequency ___.___-._-.__~.____-.___~_.._~_
4+
3+
2+
All proximal tubules show 3H-PAH uptake (except at edges of biopsy) Small groups of proximal tubules show no 3H-PAH uptake Less than half of the tubules show 3HPAH uptake
I+
Rare proximal tubules show 3H-PAH uptake
0
No cellular accumulation of ‘H-PAH seen
No abnormality
Occasional areas of tubular damage and interstitial fibrosis Less than half of the renal cortical tissue on slide shows tubular damage and interstitial fibrosis; most proximal tubules appear normal More than half of renal cortical tissue on slide shows tubular damage, interstitial fibrosis and inflammatory cell infiltration Extensive tubular damage, interstitial fibrosis and inflammatory cell infiltration are present; many tubules are dilated or obsolete
method showed little relationship to autoradiographic patterns of tritium distribution. In four biopsy specimens in which two to four biopsy fragments were studied separately, considerable variability in S:Md was evident in each biopsy fragment obtained from the same kidney. In one patient (Case 11) in whom tissue was obtained by the open surgical technic, four separate biopsy-size fragments gave 3H-PAH S:Md values of 0.4, 0.8, 1.6 and 2.9. These repetitive measurements of uptake showed such variability as to render the measurement of S:Md ratios in human biopsy fragments unreliable. When carefully selected cortical slices are used from open biopsy material, more reproducible results may be obtained [ 81. Autoradiography. Section freeze-dry autoradiographs from human renal biopsy specimens demonstrated selective proximal tubule accumulation of tritium following in vitro incubation in media containing 3H-PAH. Considerable variation was evident in segmental uptake of 3H-PAH even in apparently normal human proximal tubules (Figure 1). Glomeruli, interstitium and distal tubules showed no concentrative transport of 3H-PAH. Some proximal tubule lumens in each tissue showed grain densities which were greater than that in surrounding epithelial cells (Figures 2D and 3). When the frequency of proximal tubule uptake of 3H-PAH is graded 0 to 4-I, no correlation between uptake frequency and specific diagnosis, blood pressure or amount of proteinuria was evident (Table Ill). On the other hand, definite correlations were found to exist between frequency of 3H-PAH uptake, glomerular filtration rate and the degree of interstitial involvement
February 1977
The American Journal of Medicine
Volume 62
245
Histologic
Age
Extracapillary Pyelonephritis
Nephrosclerosis Nephrosclerosis End stage
End stage Nephrosclerosis
20 21
22 23 15B
24 25
0t
.
68 93 75 & (29) (86)
160/110 160/100 170/100 150/100 170190 140190 2401140
.
9 5
0.6 0.4
40 39
. . .
3.8 0.0
0.9 0.8
ot
ot
25 15
30 25
32
170/110 loo/70
0.2
35
32
52 48 36
120170
0.1
0.0 2.4 3.6
55
65
67
92
100
101
58
43 24 44
6.5
0.0
0.0
2.9
1 .1
0.0
109
117
0.2
0.1
144 125
150 146
154
0.0 3.4
10.5 12.9
1.5
-
0 0 0
1+ 1+
0 0 0 0 0
1+ It
2+ 2t
2t
2t
0 0 2t
2+
3+
3+
2+
1+ 0
0 Ii
2+
2+
1+ 0 1+
1+
2+
2+
2+
0.6
2.7
0.0 0.4
0.8 1 .o
1.9
. . . . .
. ’
. .
. . . .
.
2.9, 1 .6, 0.8,0.4 .
. .
f GFR assumed to approximate
0 in patients
undergoing
dialysis.
__--
.
= 0.72
at 25°C).
.
.
0.0 . . . Chronic Chronic Chronic
hemodialysis hemodialysis hemodialysis
..
Heroin user Postobstruction, infection, EPAH = 0.31
. I . .
Lupus erythematosus
Poststreptococcal
Lead colic, EPAH Hyperparathyroid . . .
Postchelation therapy, EPA,_, = 0.82 Lupus erythematosus
K,) Gold nephropathy, rheumatoid arthritis Gold nephropathy, rheumatoid arthritis E PAH = 0.31
Gold nephropathy in remission, rheumatoid arthritis Eight years poststreptococcal acute GN Low serum complement
Multiple allergies Heroin user
.
.
hematuria
. .
Unexplained
..
Intermittent proteinuria Steroid resistant
Comment
. . .
.
.
2.9
‘ii
. . .
. .
. . .
.
11.3
4+
3+
. .
1.7
0
4+
.
. . . .
. . .
4.4
180 min
4+
.
.
3.1,2.9,0,8 . *.
4.4,‘d.i
60 min
Incubation Time
2+
4+
3+ 3+
3+ 3+
4+
Autoradiograph
2+
4+
4+ 2+
4+ 1+
4+
GlomerularTubularFiltration InterRate stitial (mllmin) Disease
160190
120180 200/l 10 200/100
-
Proteinuria (gldav)
NOTE: Leaders indicate study not done. S:Md was not measured when 6 mg/lOO ml 3H-PAH was used. GN = glomerulonephritis. +S:Md = the slice to medium ‘H-PAH concentration ratio determined by the drop in counts method (I .5 mg/lOO ml ‘H-PAH incubated
17 36
46 38 74
19 56
26 27
Nephrosclerosis Membranoproliferative G N
18 19
GN
11
Membranoproliferative GN
52
17
16
nephropathy Acute proliferative GN
28 54 73
12A 14 15A
57
nephropathy Lead nephropathy Nephrocalcinosis Membranous
1751115
39
Membranous
47
13
29
120180
Lead nephropathy
28
128
46
165/100
Nephrosclerosis
11
18
90160
Minimal change
10 53
36
Minimal change
9
13 13
130170
28
8
13
120180
130/80
16
7
Healed proliferative GN Focal GN
13
6
120170
38
Membranous
4 5
nephropathy
40 32
nephropathy Normal kidney Membranoproliferative GN
21 15 15 17
105/70 11 O/70
7 28
Minimal change Membranous
12
Blood Urea Nitrogen hog/ 100 ml)
Data in Patients
130180 120170
120180
-
Blood Pressure (mm Hg)
(yr)
Uptake
and ‘H-PAH
13
2 3
Diagnosis
Clinical,
Minimal change
III
1
Case No.
TABLE
CONCENTRATIVE
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Figure 1. Autoradiograph showing 3H-PAH uptake by all proximal tubules, grade 4i-, except damaged segments at edge of biopsy specimen (Case 6). Electron microscopy demonstrated thickening of giomerular basement membranes. Tubules and interstitial regions were morphologically within normal limits (grade 4+). Periodic acid-Schiff and hematoxylin stain; original magnification X 76, reduced by 50 per cent.
(Figure 4). Fewer proximal tubules per low power field accumulated 3H-PAH in the presence of reduced glomerular filtration rate and increased interstitial disease. Proximal tubule uptake could not be predicted from tubular morphology in the freeze-dried sections since adjacent tubules with comparable structural changes could be associated with intense to absent autoradiograph grain density (Figures 1, 26, 2D, 2F and 3). COMMENTS
The most striking observation in this study is that the frequency of proximal tubular uptake of 3H-PAH correlated directly with the glomerular filtration rate and inversely with tubular-interstitial disease (Figure 4). A correlation between glomerular filtration rate and the degree of tubular-interstitial involvement irrespective of glomerular histopathology has been recognized [9]. Regardless of the etiology of the renal disease, the lower the glomerular filtration rate and the greater the tubular-interstitial involvement, the fewer the proximal tubules that demonstrated cellular accumulation of 3H-PAH in the autoradiographs. In the presence of renal failure, the persistence of 3H-PAH uptake in only isolated groups of proximal tubules is consistent with the view that a few relatively normal nephrons support residual renal function rather than a more uniformly damaged population of nephrons. In the presence of reduced glomerular filtration rate and interstitial changes, broad areas of each kidney cortex section showed no uptake of PAH. It has previously been shown that serum from uremic subjects inhibits PAH uptake in kidney cortex slices [lo]. Since some tubules showed persistence of 3H-PAH uptake even in azotemic patients (Figure 2), failure of concentrative transport was apparently not due to competitive inhibitors of PAH found in the serum of uremic patients. Inhibitors of PAH uptake in such serum would be expected to produce uniform effects on all nephrons; such inhibition of uptake should have produced a ho-
mogeneous autoradiographic pattern. On the contrary, in tissue obtained from azotemic patients, proximal segments demonstrating 3H-PAH uptake tended to be clustered in isolated areas (Figures 2C and 2E). Moreover, patients undergoing hemodialysis with end-stage renal disease did not show improved tubular uptake of 3H-PAH due to dialysis (Table Ill). Endogenous organic molecules within the biopsy fragments which act as competitive inhibitors of PAH transport are apparently greatly diluted by the 1,OOO-fold excess volume of the incubation medium. The failure of proximal tubules to accumulate 3H-PAH similarly cannot be attributed to a diffusion barrier produced by interstitial fibrosis or tubular basement membrane thickening. Concentrative transport was often demonstrable deep within a biopsy fragment despite broad areas of no uptake in adjacent but more superficial regions. If the widened interstitium had prevented 3H-PAH from reaching some proximal tubule segments, this effect would have been most apparent in the regions of the tissue that were more remote from the incubation medium. The long incubation times and relatively high 3H-PAH concentrations used in this study insured full penetration of the tissue by 3H-PAH. Moreover, proximal tubule segments with comparable basement membrane thickening and surrounding interstitial fibrosis showed variations in uptake that could not be explained by morphologic changes alone (Figures 2 and 3). Cellular damage due to the biopsy needle appeared to be limited to the surface of the specimen so that interference with concentrative transport due to gross cellular trauma was limited to the outer edges of the tissue specimen (Figure 1). Linear fractures apparent in the interior of the tissue were produced during frozen tissue sectioning. Such sectioning artifacts occur after the in vitro incubation and, thus, could not alter the concentrative transport process. The loss of PAH uptake by proximal tubules evident in the autoradiographs thus appears to be caused by cellular damage accompanying renal failure.
February 1977
The American Journal of Medicine
Volume 62
247
CONCENTRATlVE TRANSPORT IN HUMAN RENAL BlOPSY SPECIMENS-WEDEEN
Figure 2. Autoradiographs demonstrating intermediate grades of 3H-PAH uptake. A and B, Case I 7: 3H-PAH uptake grade 3+, tubular-interstitial disease grade 2+. C and D, Case 17: 3H-PAH uptake grade 2+, tubular-interstitial disease grade 2+. E and F, Case 2 I: 3H-PAH uptake grade I+. tubular-interstitial disease grade 0 (see Table II).
248
February 1977
The American Journal of Medicine
Volume 62
CONCENTRATIVE TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Figure 3. Autoradiographfrom hypertensive patient (Case 71) demonstrating cellular accumulation of 3H-PAH in proximal tubules (3+). Higher 3H-PA H concentrations are present in some lumens of the proximal tubule than in cells. Moderateinterstitial fibrosis, mononuclearcell infiltration and tubular basement membrane thickening is present (grade 2+). Periodic acid-Schiff and hematoxylin stain; original magnification X 795, reduced by 50 per cent.
The correlation of the frequency of 3H-PAH uptake with the glomerular filtration rate and tubular-interstitial disease contrasts with the independence of 3H-PAH uptake from proteinuria. Although tubular reabsorption and catabolism of proteins is an important proximal tubular function [ Ill, an increased tubular protein load in vivo does not appear to interfere with accumulation of 3H-PAH in vitro. As renal disease progresses, fewer tubules participate in concentrative transport of PAH, whatever the underlying etiology. Since the nontransporting nephron segments are not necrotic, they must continue to receive some blood supply. PAH contained in the blood reaching nephrons which have lost the ability to accumulate 3H-PAH is presumably not extracted. The failure to secrete PAH undoubtedly contributes to the decrease in PAH extraction associated with renal failure [ 121.
PAH extraction ratios (E~AH)obtained at the time of renal biopsy in four of our cases (Table III) failed, however, to show any consistent correlation with glomerular filtration rate, grade of tubular interstitial disease or autoradiographic uptake pattern. These few measurements of EPAHdo not suggest any simple correlation with 3H-PAH uptake in vitro. This is, perhaps, not surprising since EPAH appears to be determined by a number of factors other than cellular transport processes and glomerular filtration rate [ 131. The finding that cellular accumulation of 3H-PAH in vitro correlated with changes in PAH extraction &d maximal transport rate (TmpAH) in vivo in a case of occupational lead nephropathy, reported in detail elsewhere [ 21, remains unique and is presumably related to specific proximal tubule effects of lead. The relationship between cellular accumulation of PAH and transepithelial transport remains to be defined.
-
PAH-3
H Uptake
C-4Intrrstitial
Frequency
Disease
160 .
140
Figure 4. Relationship of 3H-PAH uptake frequency and interstitial disease to glomerular filtration rate. Uptake frequency (circles) and tubular-interstitial disease (triangles) were graded0 to 4+ ( Table IIf). The glomerular filtration rate was considered to be 0 in three chronic hemodialysis patients. Both 3H-PAH uptake frequency and tubular-interstitial disease correlate with the glomerular filtration rate. The calculated linear regression for uptake frequency is plotted as a solid line and that for tubular-interstitial disease as a broken line. Both regressions had correlation coefficients of 0.8 and T values of 6.
120
-;
100
.
-i a IL
60 60
(3 40
0
0
I+
AUTORADlOGRAPHlC
February 1977
2+
3+
AND HISTOLOGIC
The American Journal of Medicine
4t GRADE
Volume 62
249
CONCENTRATIVE TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
Precise measurements of 3H-PAH concentrations in proximal tubule cells would require a more quantitative approach to autoradiography than has been feasible with freeze-dried sections. Grain counting has not proved useful in these preparations. In addition to the well known technical difficulties inherent in quantitative autoradiography [ 141, the histologic preparations used in this study present formidable problems for grain counting. The usual variables (tissue-film contact, latent image fading, silver halide saturation, depth of focus and background correction) are compounded by problems introduced by the freeze-dried sections. Freeze-drying sections minimize diffusion of water-soluble labeled compounds at the expense of introducing variability of tissue section thickness, cutting artifacts and possible physical displacement of 3H-PAH during sectioning and freeze-drying [7]. Quantitative autoradiography of tubular cell uptake is further handicapped by the obvious segmental variation in 3H-PAH uptake within even the normal proximal tubule (Figure 1). Some of the factors which interfere with grain counting can be overcome by using embedded freeze-dried tissue blocks [ 151 but only at the risk of permitting redistribution of labeled compounds due to solubility and water permeability of the embedding medium. Assessment of 3H-PAH uptake in human renal biopsy specimens by determining S:M ratios has proved equally futile. Minor modification of traditional slice incubation technics permitted reproducible measurement of tissue uptake in biopsy-size fragments of rat renal cortex. Comparable fragments from human renal biopsy specimens did not, however, show reproducible S:M ratios. The inability to obtain reliable 3H-PAH uptake data on human biopsy specimens in vitro appears to be caused by the sampling problems inherent in the needle biopsy technic. Portions of this minute biopsy material often contain capsule, medulla, perirenal fat or excretory tract epithelium. The diseased kidney, in addition, has regions of scarring adjacent to regions of relative preservation of morphologic integrity. The range of variation in S:M uptake data from diseased kidneys proved too great to permit reliable uptake measurements in single human biopsy fragments. The sampling errors inherent in needle biopsy material interfere with autoradiographic evaluation far less than with the determination of S:M ratios. Noncortical tissue can readily be excluded from analysis on morphologic grounds in the autoradiographs, thereby greatly reducing observational variability. Biopsy material containing no renal cortex can easily be excluded. On the other hand, a tissue fragment containing only a minute portion of cortex and therefore providing invalid S:M ratios may, nevertheless, contain enough proximal tubules to evaluate concentrative transport by autoradiography. Errors due to sampling would be expected to contribute
250
February 1977
The American Journal ol Medicine
to the considerable variation in individual autoradiograph grades when compared to whole kidney glomerular filtration rate (Figure 4). The 23 percutaneous renal biopsy specimens discussed here were selected from a consecutive series of 30 in vitro biopsy incubations because renal cortical tissue was present for autoradiographic evaluation. Of the seven biopsy specimens excluded, two were lost because of early technical problems in specimen handling, four contained only renal medulla, and one contained only fibrous tissue. In addition to demonstrating a correlation between the frequency of proximal tubular uptake of 3H-PAH with glomerular filtration rate and tubular interstitial disease, these autoradiographic studies provide insight into the function of normal human proximal tubules. They demonstrate that the human proximal tubule accumulates 3H-PAH intracellularly in vitro. The finding that 3H-PAH concentrations in the lumens are occasionally higher than that in adjacent cells, as judged by inspection of autoradiograph grain density, is similar to previous observations in the rat [ 161. Luminal accumulation of 3H-PAH was present in both percutaneous needle biopsy fragments and renal cortical slices obtained by the open surgical biopsy technic. Assuming that the renal cortex in man in vitro behaves like that of the rat, the luminal channel of proximal tubules is not in open communication with the incubation medium. In the rat, neither labeled inulin nor labeled sucrose is detectable in the lumens of the proximal tubule following in vitro slice incubation [ 161. 3H-PAH thus reaches the lumen only by transport across the epithelial cell. Since the autoradiographic grain density is greater in the lumen than in the ,cell in this freeze-dried tissue (Figures 2D and 3) 3H-PAH transport appears to be uphill across the luminal as well as the peritubular membrane. The failure to demonstrate comparable uphill transport of PAH at the luminal surface in isolated perfused rabbit tubules [ 171 may relate to species differences or to the method of isolating and perfusing individual nephron segments. Moreover, apparent uphill luminal secretion in vitro is by no means universal in proximal tubules in the autoradiographs and may therefore be missed in individual perfused tubular segments. The high luminal grain density arising from 3H-PAH does not appear to be a technical artifact. Differences in tissue absorption of the beta emissions from tritium cannot explain the autoradiographic grain distribution pattern. Freeze-dried unembedded 3 p thick cells have too little density to account for the manifold reduction in autoradiograph grains underlying cytoplasm as compared to lumens (Figure 3). Even when the lumens of the proximal tubule appear to be closed or filled with proteinaceous material, increased luminal grain density is evident in apposition to the brush borders. The autoradiographic pattern, suggesting uphill transport of
Volume 62
CONCENTRATIVE
3H-PAH from cell to lumen, has consistently been evident in renal autoradiographs from man, rat and dog in this laboratory. Moreover, high luminal radioactivity has never been seen in renal cortical slices incubated with a variety of other tritium or carbon-14 labeled compounds including inulin, sucrose, iodoacetamide, dinitrophenol, ouabain, cY-aminoisobutyric acid, cycloleutine, lysine or methionine. These initial autoradiographic investigations of human tubular function have been confined to PAH, a nonmetabolized molecule, which is avidly secreted and accumulated by the proximal tubule cell. Similar morphologic-functional correlations could be undertaken with other labeled diffusible compounds which undergo
TRANSPORT IN HUMAN RENAL BIOPSY SPECIMENS-WEDEEN
concentrative transport, such as nonmetabolized amino acids, sugars and antibiotics. If specific patterns of tubular cell dysfunction can be identified by section freeze-dry autoradiography, a clearer understanding of the role of interstitial and tubular changes in renal disease may emerge. ACKNOWLEDGMENT These studies were performed with the technical assistance of Ms. Claffertene Cheeks and Mr. Bhagirath Vyas. Clinical microscopy was performed by Drs. Michael Lyons, Norman Ende and Gabriel Mulcahy. Dr. Jacob Churg generously provided guidance in the grading of tubular-interstitial disease.
REFERENCES 1.
2. 3.
4.
5.
6.
7.
8.
Andres GA, McCluskey RT: Tubular and interstitial renal disease due to immunologic mechanisms. Kidney Int 7: 271, 1975. Wedeen RP, Maesaka JK, Weiner B, et al.: Occupational lead nephropathy. Am J Med 59: 630, 1975. Wedeen RP, Thier SO: The intrarenal distribution of nonmetabolized amino acids in vivo. Am J Physiol 220: 507, 1971. Cross RJ, Taggart JV: Renal tubular transport: accumulation of p-aminohippurate by rabbit kidney slices. Am J Physiol 161: 181, 1950. Wedeen RP, Weiner B: The effect of acetate on p-aminohippuric acid uptake in kidney slices. Fed Proc 33: 1103, 1974. Wedeen RP, Weiner B: The distribution of p-aminohippuric acid in rat kidney slices. II. Depth of uptake. Kidney Int 3: 214, 1973. Bordier B, Ornstein L, Wedeen RP: The intrarenal distribution of para-aminohippuric acid determined by a modified technique of section freeze-dry autoradiography. J Cell Biol 46: 518, 1970. Thier SO: Amino acid transport i:r the renal tubule cell. Recent Advances in Renal Physiology and Pharmacology (Wesson LG. Fanelli GM Jr, eds), Baltimore, University Park Press, 1974, p 39.
February
9.
10.
11. 12. 13.
14.
15.
16.
17.
1977
Schainuck LI, Striker GE, Cutler RE, et al.: Structural-functional correlations in renal disease. Il. The correlations. Human Pathol 1: 631, 1970. Preuss HG, Massry SG, Maher JF, et al.: Effects of uremic sera on renal tubular p-aminohippurate transport. Nephron 3: 265, 1966. Maack T: Renal handling of low molecular weight proteins. Am J Med 58: 57, 1975. Wesson LG, ed: Physiology of the Human Kidney, New York, Grune & Stratton, 1969, p 165. Haberle D: Influence of glomerular filtration rate on the rate of para-aminohippurate secretion by the rat kidney. Kidney Int 7: 385, 1975. Perry RP: Quantitative autoradiography. Methods in Cell Physiology, vol 1 (Prescott DM, ed), New York, Academic Press, 1964, p 305. Kinter WB, Wong M: Peritubular membrane transport of PAH into proximal cells in necturus kidney slices. Am J Physiol 227: 50, 1974. Wedeen RP, Weiner B: The distribution of p-aminohippuric acid in rat kidney slices. I. Cellular localization. Kidney Int 3: 205, 1973. Tune BM, Burg MB, Patlak CS: Characteristics of p-aminohippurate transport in proximal renal tubules. Am J Physiol 217: 1057, 1969.
The American
Journal
of Medicine
Volume 62
251