Acta Physiol Scand 1990, 138, 13-23
Mechanism of erythrocyte trapping in ischaemic acute renal failure A. B A Y A T I , R. CHRISTOFFERSON*, 0. KALLSKOG and M. WOLGAST Departments of Physiology and Medical Biophysics, and * H u m a n Anatomy, University of Uppsala, Sweden A., CHRISTOFFERSON, R., KALLSKOG, 0. & WOLGAST, M. 1990. Mechanism of BAYATI, erythrocyte trapping in ischaemic acute renal failure. Acta Physiol Scand 138, 13-23, Received 8 June 1989, accepted 28 August 1989. ISSN 00014772. Departments of Physiology and Medical Biophysics, and Human Anatomy, University of Uppsala, Sweden. Forty-five minutes of warm ischaemia and 20 min of recirculation in the rat kidney was found to result in ( I ) a massive transient extravasation of plasma upon recirculation and (2) an increase in plasma-lymph transport of proteins during the first hours after onset of circulation. This was accompanied by trapping of erythrocytes, as determined with 'lCr-labelled erythrocytes, in the capillaries, mainly in the inner stripe of the outer medulla. At scanning electron microscopy of vibratome sections, the trapping appeared as aggregates of polygonally shaped erythrocytes. It is concluded that 45 min of ischaemia and 20 min of recirculation results in an increase in the permeability of the renal capillaries. This increase leads to extravasation of capillary plasma with consequent local haemoconcentration, causing an increase in vascular resistance and in capillary hydrostatic pressure. This elevated pressure will, in turn, lead to perpetuating extravasation of plasma, further haemoconcentration and so on, eventually resulting in dense packing of polygonal erythrocytes, obstructing the blood flow. It is believed that oxygen-derived free radicals generated in the early recirculation phase contribute to the increase in macromolecular permeability, since the scavenger bovine superoxide dismutase and allopurinol, a xanthine oxidase inhibitor, were found to prevent this unfavourable chain of events.
Key words : acute renal failure, allopurinol, congestion, erythrocyte trapping, kidney, permeability, superoxide dismutase (SOD).
I t has been proposed that trapping of erythrocytes in the kidney - primarily in the inner stripe of the outer medulla - is of great importance in the pathogenesis of post-ischaemic acute renal failure (NorlCn et al. 1978a, Karlberg et al. 1982a, Wolgast et al. 1982, Mason 1988). T h e sustained medullary ischaemia subsequent to this trapping (Karlberg et al. 1982b, 1983) may thus underlie the reduced capacity for urine concentration and the disturbed potassium secretion, both of which are typical features in post-ischaemic acute renal failure. Correspondence : A. Bayati, Department of Physiology and Medical Biophysics, University of Uppsala, Box 572, s-751 23 Uppsala, Sweden.
T h e trapping has been found to occur in the early recirculation phase u p to 20 min after the onset of circulation (Karlberg et al. 1982a). T h e factors responsible for this unfortunate event remain, however, unexplained. Since the trapped erythrocytes occupied almost the entire vascular space of the renal medulla (Karlberg et al. 1982a), extravasation of plasma through a damaged capillary membrane (Hansson et al. 1983, Karlberg et al. 1983, Bayati et al. 1987, Ojteg et al. 1987) might be the first step in this process. In order to test this hypothesis, the following were investigated on termination of induced ischaemia : (a) T h e plasma-lymph transport of macroI3
11
,4. Bayati et al.
molecules during the first 2 h after the onset of recirculation. (b) 'The trapping of erythrocytes in terms of the distribution volume of "'Cr-labelled erythrocytes in the different kidnej- regions. (c) The extravasation of plasma during the first minute after onset of circulation, as studied from the haematocrit of renal venous blood. (d) T h e morphology of the kidney, in particular the inner stripe of the outer medulla, by scanning electron microscopy of vibratome sections. Since oxygen-derived free radicals may be responsible for the increase in macromolecular permeability (Del Maestro et al. 1981a, b), the effects of bovine copper/zinc superoxide dismutase @-SOD) as a scavenger of oxygenderived free radicals and the effects of allopurinol, an inhibitor of xanthine oxidase, on these parameters were also studied.
macromolecular permeability, as reflected by the plasma-lymph transport of the plasma proteins, was studied in 16 rats weighing 301 F 10 g. For this purpose, a hilar lymph vessel was dissected free and cannulated with a 100-zoo p m polyethylene catheter prefilled with heparinized saline. Lymph was sampled at 30-min intervals both before the ischaemia and up to 120 min after recirculation. Rats in which the lymph flow before clamping was less than I yl min-' were discarded. The volumes of the samples were measured from their length in 75-pI constant-bore capillaries (Drummond Scientific Co., USA). Plasma samples were withdrawn in the middle of each p m i n period. The protein concentrations of plasma and lymph were measured by the method of Lowrey et al. (1951).The protein transport was calculated by multiplying the lymph flow by the lymph protein concentration. The experiments were performed both in untreated rats and after intravenous administration of up to 40 mg of b-SOD (Grunenthal GmbH, FRG), given 30 s before the onset of recirculation (Bayati et al. 1987).
M A 4 T E R I A I , S AND M E T H O D S The experiments were performed on 89 male Sprague-Dawley rats (Alab, Stockholm, Sweden) weighing 281 F 4 g. They had free access to standard rat chow (Rj, EWOS, Sweden) and water prior to the experiments. The animals wrere anaesthetized with Inactin (Byk Gulden, Konstanz, FRG) given intraperitonealll- in a dose o f 120 mg kg ', After tracheostomy they were placed on a servo-controlled heating pad to stabilize the body temperature at 3j . j "C. The left femoral artery was cannulated for monitoring of the sj-stemic blood pressure and withdrawal o f blood samples. The left femoral vein was cannulated for administration of test substances and for continuous infusion of Ringer-bicarbonate, which was given at a rate of 5 ml kg h- I . All the animals received 40 IC: of heparin intravenously 30 min before clamping of the renal artery. The left kidney was exposed through a flank incision, dissected free from its perirenal tissue in order to cut off collateral blood supply, and suspended in a Lucite cup. Ischaemia was evoked by clamping the renal artery for 45 min, during which time the kidney was placed in its natural position with the abdomen closed. In order to secure a kidney temperature of 3 7 . 5 "C, cotton wool was placed over the incision. After termination of the ischaemia, the abdomen was reopened, the kidney was resuspended in the L,ucite cup, and the arterial clamp was removed. In all the animals the right kidney served as a control. The animals were divided into four main groups: ( I ) .Wacromolecular permeability. The transcapillary
( 2 ) Erythrocyte distribution volume. The distribution volume of "Cr-labelled erythrocytes was studied in 33 rats weighing 292 & 36 g. For the purpose of labelling, 10 nil of blood from a donor rat was dissolved in I ml of ACD solution (Travenol Laboratories, Thetford, Norfolk, England) and centrifuged at 1000 r.p.m. for 10min. After removaI of the supernatant, roo yCi of 51Cr(sodium chromate, Amersham International, Amersham, UK) was added and the sample was incubated at room temperature for 20 min. The erythrocytes were then washed four times and resuspended in an equal volume of saline. One millilitre of the suspension was injected intravenously 10min before clamping of the renal artery. After 20 rnin of recirculation the two kidneys were excised. For microdissection the two poles of the kidney were first removed. A triangular core was then obtained by making two sagittal incisions in the striatal direction. Thereafter the core was subdivided into: (a) cortex, (b) outer stripe of the outer medulla, (c) inner stripe of the outer medulla and (d) inner medulla. The "Cr activity of the tissue specimens and that of the erythrocytes in the circulating blood withdrawn at the same time were measured by gamma-spectrophotometry. The erythrocyte distribution volume was calculated as the ratio of the activity in the kidney specimens to that in the erythrocytes in systemic blood and is represented as per cent of the zonal volume. The material was subdivided into three groups : (a) A group of 10 rats in which the left kidney was subjected to 45 rnin of ischaemia and zomin of recirculation. (b) -4group of 18 rats subjected to the same trauma
Erythrocyte trapping in acute renal failure
15
but treated with up to 20 mg of b-SOD given The triangular core and the two kidney poles were intravenously 30 s before recirculation. immersion-fixed for 24 h at room temperature, using (c) A group of five rats subjected to the same 2.5 yo glutaraldehyde (Kebo Lab AB, Stockholm, trauma but treated with 40 mg of allopurinol (Sigma Sweden) in a phosphate buffer solution (pH 7.4). The Chemical Co., MO, USA), given intravenously 4 h osmolalities of the phosphate buffer solution used as before clamping of the kidney. carrier and the fixative solution were 280 and 530 (3) Extravasation of plasma. The extravasation of mosmol I-' respectively. plasma during the first minute after onset of circulation After the above prefixation with glutaraldehyde, the as determined from the haematocrit in renal venous specimens were washed four times with the carrier. blood was studied in 20 rats weighing 258 4 g. The samples were then suspended in the same carrier For the catheterization of the renal vein a sharpened and cut into I O ~ Z O Opm thick slices with an Oxford steel cannula with an outer diameter of 0.9 mm was vibratome set at its maximal frequency and lowest inserted through the vessel wall. After 15 rnin the steel speed. However, in order to facilitate the sectioning, a cannula was gently removed, whereupon the puncture rigid frame surrounding the specimen was obtained hole would be sealed with fibrin. A silicon catheter by immersing it for 5 min in I % osmium tetroxide (outer diameter 0.9 mm, inner diameter 0.5 mm) (Expectron Man. AB, Stockholm, Sweden), dissolved could then easily be inserted through the preformed in the same carrier. The sections were then washed again and post-fixed hole. This narrow catheter was attached to a wider silicon tubing which was used to harvest the blood at room temperature in 1 % osmium tetroxide samples; this had an outer diameter of 3 mm and an overnight. Thereafter surplus fixative was removed inner diameter of I mm. The tubing was connected to and the specimens were dehydrated in a graded series a suction pump set at a rate of 0.6 ml min-'. In order of acetone. After 12 h in absolute acetone, the sections to separate the blood samples from each other, small were dried in a critical point drying apparatus air bubbles were sucked into the tubing every 2 s, (Polaron, Agar, UK), mounted .on Cambridge aluusing a revolver principle described previously (Lars- minium stubs, coated with a 400 A gold layer in a fine son et al. 1984). coat sputter (Jeol JFC I 100)and observed in a Philips The sampling was started 10 s before re- 525 scanning electron microscope (SEM) at an circulation and continued for up to 50 s after. acceleration voltage of 20 kV. During this time 30 samples, each with a volume To rule out the possibility of a preparational of zopl, were harvested. They were transferred artifact, parallel studies on the right control kidney to 2 0 4 heparinized microcapillaries and, for were performed in all series. The finding of normal measurement of the haematocrit, centrifuged at erythrocytes (Fig. I b) and podocytes, both of which 9000 r.p.m. are sensitive to preparation mishaps, provides evidence In this series parallel studies of the intrarenal against such an artifactual cause of the polygonally erythrocyte volume were also made. The material was shaped erythrocytes found in these experiments. subdivided into four groups : In order to quantify the erythrocyte trapping, 30 (a) A 'control' group of five rats, in which the left photographs from each kidney (original magnification kidney was subjected to only 5 rnin of ischaemia, i.e. 100-1100) were analysed with respect to the pera minimal trauma. centage area occupied by capillaries, using a MOP (b) A group of five rats in which the left kidney was system (Quantitative Bilauswertung, Kontorn Messsubjected to 45 rnin of ischaemia and 20 rnin of gerate GmbH, FRG). recirculation. The material was subdivided into the following (c) A group of five rats subjected to the same groups : ischaemia and recirculation but treated with 20 mg of (a) A group of five rats in which the left kidney was b-SOD given intravenously 30 s before recirculation. subjected to 45 min of ischaemia but with no (d) A group of five rats subjected to the same recirculation. ischaemia and recirculation but treated with 40 mg of (b) A group of five rats subjected to both 45 min of allopurinol in a 67 mM NaOH/saline solution (pH 9) ischaemia and 20 min recirculation. given intravenously 4 h before the clamping of the (c) A group of five rats subjected to 45 min of kidney. ischaemia and 20 min of recirculation, but which (4) Morphological studies. The morphology of the recieved a single intravenous dose of 20 mg of b-SOD left experimental and right control kidney was 30 s before recirculation. investigated in 20 rats weighing 273 f7 g. (d) A group of five rats subjected to the same In order to prevent collapse of the tubules resulting ischaemia and recirculation, but treated with a single from the reabsorption of fluid which occurs soon after intravenous dose of 40 mg of allopurinol 4 h before nephrectomy, z ml of a 12% sucrose-Ringer solution clamping of the kidney. was injected intravenously 10 rnin prior to the removal All the results are presented as mean _+ SEM. The of the kidneys. comparison between the different groups was done
16
A . Bayati et al.
Fig. I . Scanning electron micrographs of the inner stripe of the outer medulla of a normal kidney (a x 6jo; b x 1490) and of kidneys subjected to both ischaemia and recirculation (c x 650; d x 1490).The capillaries of the damaged kidneys (c and d) are wide open and are occupied by
deformed erythrocytes.
null hvpothesis.
the mass transport of proteins was the same as that in the preischaemic period. The same tendency could also be traced in the group of animals treated with 20 mg of b-SOD.
RESULTS
Erythrocyte distribution volume
using Student’s t-test. T h e results of this comparison are presented as the highest possibility in favour of the
As shown in Fig. z(b), 45 min of ischaemia and 20 min of recirculation resulted in trapping of In kidneys subjected to 45 min of ischaemia and erythrocytes, mainly in the inner stripe of the 20 min of recirculation, the lymph protein outer medulla, comprising IS”/:,of the inner concentration increased from a control value of stripe volume. T h e corresponding figure for the 1 . 7 & 0 . 1 3 to2.210.23 mg ml-’(Fig. 2a). At the inner zone was 676, while in the cortex no same time the lymph flow increased from significant increase was observed. In contrast, kidneys subjected to only 5 min of I .6 0.13 to 2.9 fo.64 pl min-’ and the mass transport from 2. j i0.25 to 7.3 & 0.85 mg min-’ ; ischaemia showed essentially no increase in the thus there was a threefold increase in the net distribution volume of 51Cr-labelled erythrocytes (Fig. 3 b). transcapillary transport. After treatment with 40 mg of b-SOD, the Treatment with 40 mg of allopurinol reduced increase in the lymph flow was outweighed by a the trapping considerably (Fig. 3 b), and 20 mg reduction in the protein concentration such that of b-SOD had the same effect (Fig. 2b). .Ifarrornolecular transport
Erythrocyte trapping
1
acute renal failure
h
E '
E
i
E"
T
Cortex
8
I7
Outer stripe Inner stripe
Inner zone
Fig. 2. The upper panel (a) shows the average plasma-lymph data of plasma proteins in the period up to 1.5 h after the onset of recirculation (mean +SEM) in different groups: normal kidneys ischaemic control kidneys (a),ischaemic kidneys treated with 20 mg of b-SOD (a)and ischaemic kidneys treated with 40 mg b-SOD (a).The lower panel (b) depicts the percentage regional renal erythrocyte volume after 20 min of recirculation as investigated by the use of X r labelled erythrocytes in the same groups as in (a). *represents a probability less than 0.05.
(a),
Extravasation of plasma
The haematocrit of renal venous blood before and shortly after the onset of circulation 'is depicted in Fig. 3 (a). I n kidneys subjected to 45 min of ischaemia the haematocrit increased from an initial value of 46 1.6 to a peak value of 61 ko.7 after 15 s of recirculation, whereafter it
returned to an almost normal value within the following minute. I n contrast. there was no alteration in the haematocrit 'in the kidneys subjected to only 5 min of ischaemia. As seen in Fig. 3 (a), treatment with 20 mg of b-SOD or 40 mg of allopurinol effectively prevented the post-circulation rise in haematocrit resulting from 45 min of ischaemia.
18
A. Bayati et al.
62
50
44
I
10
0
I
I
20
30
Time
I
I
40
’
50
(sec)
10
0 Cortex
Outer stripe
Innerstripe
lmer zone
Fig. 3. T h e upper panel (a) illustrates the venous haematocrit changes in the early recirculation phase after 5 min (*) and 45 min of ischaemia in untreated kidneys (a),kidneys treated with 20 mg of b-SOD and kidneys treated with 40 mg of allopurinol (V). Note the very steep increase in the haematocrit in the untreated kidneys after 15 s of recirculation. This increase was almost normalized after administration of b-SOD or of allopurinol. In the lower panel (b) the percentage erythrocyte volumes in different regions of the same kidneys are shown. Right control kidneys (0) and left kidneys after 5 min of ischaemia (m), 45 rnin of ischaemia with no treatment and 45 min of ischaemia treated with 40 mg of allopurinol (m).
(m),
(a)
Erythrocyte trapping in acute renal failure 40
19
gonally shaped erythrocytes. Apart from this occasional finding, the morphology was similar to that of the right control kidney. The morphometric studies on kidneys subjected to 45 rnin of ischaemia and 20 min of recirculation showed that in the inner stripe of the outer medulla as much as 34kz.oyo of the total area was occupied by capillaries. This is in sharp contrast to the kidneys subjected to 45 min of ischaemia but with no recirculation, where only 4 f0.3 yo was occupied by capillaries. Treatment with 20 mg of b-SOD or 40 mg of allopurinol diminished this area to almost normal values (Fig. 4).
F
Fig. 4. The percentage area of the inner stripe occupied by capillaries in right control and left damaged kidneys in different groups. Right control kidney left kidneys after 45 rnin of ischaemia with no recirculation (a)and kidneys subjected to 45 min of ischaemia and 20 rnin of recirculation with no treatment ( 0 )treated with 20 mg of b-SOD (a)and treated with 40 mg of allopurinol (a).
(a),
DISCUSSION
In our view the key event underlying the erythrocyte trapping after 45 min of ischaemia and 20 min of recirculation is a dramatic increase in the macromolecular permeability of the renal capillaries. There is a possibility, however, that the increased mass transport of proteins into the lymphatic system might be Morphological studies caused by an increase in the transcapillary After 45 min of warm ischaemia and 20 min of hydrostatic pressure difference, since the fluid recirculation, the peritubular capillaries of the filtration also increases the transport of large renal medulla were distended and contained molecules by convection. aggregates of polygonal erythrocytes (Fig. I c). As discussed by Reed (1988), however, the The tubules appeared to be compressed by the enhanced fluid filtration would lead to a dilution surrounding capillaries (Fig. ~ d and ) in most of the lymph proteins, which would be incases exhibited no lumen. The vasa recta showed compatible with the present finding that ischano such aggregates, but contained normal-shaped emia and recirculation resulted in an elevated erythrocytes. lymph protein concentration. Moreover in a These findings were in sharp contrast to those recent study at this laboratory it was found that in the right control kidney (Fig. ra). Here, the the transcapillary hydrostatic pressure difference peritubular capillary lumen displayed its normal did not change after ischaemic damage (Hellberg narrow dimension and only occasionally con- et al. 1988). It would thus seem likely that the tained erythrocytes (Fig. I b). When erythrocytes increased transport of macromolecules is due to were present, they were of normal size and a true increase in permeability (Durbin 1960), an shape. Nearly all tubules were open and had interpretation which is in accordance with that in previous studies both in the kidney (Hansson their normal rounded appearance. After 45 rnin of ischaemia but without recircu- et al. 1983, Karlberg et al. 1983, Bayati et al. lation, the overall morphological picture of the 1987, Ojteg et al. 1987) and in other organs (Del inner stripe of the outer medulla was much 1;ke Maestro et al. 1981a, b, McCord & Roy 1982, that in the control kidneys, although the tub. rr Parks et al. 1984). Since SOD (a scavenger of oxygen free epithelial cells seemed swollen. No polygo~ial .,dicals) and allopurinol (an inhibitor of xanthine erythrocytes were observed. tidase) were able to restore the plasma to In kidneys subjected to ischaemia and recirculation, but treated with either b-SOD or lymph transport of proteins, the injury to the allopurinol, only some of the peritubular capil- capillary barrier seems to be caused by oxygenlaries were obstructed by aggregates of poly- derived free radicals (DeI Maestro et a[. 1981a,
20
A . Bayati et al.
Fig. 5. The upper diagram illustrates that, in a capillary with a comparatively short transit time, the extravasation of plasma leads to congestion of erythrocytes in the distal end of the capillary soon after the onset of recirculation. In the lower diagram, which represents a capillary with a long transit time, all plasma has escaped from the capillary. The erythrocytes are thereby forced, in the ideal case, to assume a hexagonal shape.
b, Paller et al. 1984, Parks et af. 1984, Oriel et al. 1985). The course of events underlying the production of these free radicals may be as follows: During the ischaemia and the subsequent anoxia, the purine metabolism leads to accumulation of hypoxanthine (Hansson et al. 1983). In addition, the inactive P form of xanthine oxidase is converted into its active 0 form (Stripe & Della Corte 1969). After the onset of circulation and return of oxygen, the latter form of xanthine oxidase will convert the accumulated hypoxanthine into xanthine and further to uric acid. In each of these steps an oxygen free radical (0;)is generated. Since this radical is known to break down polysaccharides, such as glucosaminoglycans (Fink & Lengfelder 1986), of the capillary barrier (Valsala & Avasthi 1987), 0; is a likely candidate for the damage to the capillary barrier. The effect of these oxygen radicals would of course be on all the capillaries of the kidney, although with different intensity. The fact that the trapping is mainly seen in the outer medulla of the kidney could be explained (see below) by additional structural and rheological characteristics of this special part of the kidney. Regarding the mechanism of the erythrocyte trapping, the essential event, in our opinion, is not a compression of the erythrocytes but a complete escape of plasma. This may be explained as follows (Fig. 5 ) :
Firstly, major transport of proteins into the perivascular space will lead to a reduction of the colloid osmotic pressure difference. The consequent large outward filtration of fluid will result in local haemoconcentration (Fig. 3 a) and thereby in an increase in vascular resistance. The fact that the blood flow will be reduced will also, in itself, increase the vascular resistance (Vann & Fitz-Gerald 1982). T h e concomitant rise in capillary hydrostatic pressure will then lead to further extravasation of plasma, further haemoconcentration, a further increase in hydrostatic pressure, and so on. It is also evident that for one and the same macromolecular permeability, a specific volume of blood will lose more of its plasma the longer the transit time (Fig. 5). In capillaries with a short transit time, as in the cortex (Wolgast 1973), the above process will only lead to ‘congestion’ of the erythrocytes in the distal part of the capillary. T h e capillaries will thereby remain patent to the blood flow. It may also be expected that as soon as the macromolecular permeability is restored, the intracapillary haematocrit and the blood flow will also be restored. This idea is supported by the present finding of a transient increase in the haematocrit in renal venous blood soon after the onset of the recirculation (Fig. 3 a). T h e finding that b-SOD and allopurinol both reduced early extravasation of plasma emphasizes the role of oxygen-derived
Erythrocyte trapping in acute renal failure
Fig. 6. The polygonally shaped erythrocytes in a peritubular capillary in the inner stripe of the outer medulla of a kidney subjected to 45 min of ischaemia and 20 min of recirculation ( x 5270). Note the very distended capillary containing a large number of deformed erythrocytes.
Fig. 7. Scanning electron micrograph of the inner stripe of the outer medulla, showing the compression of the tubules by the distended capillaries ( x 4080).
21
22
.4. Ba-luti et al.
free radicals in the ischaemia-recirculation damage. In the renal medulla, where the transit time is -5-10 times longer than in the cortex (Wolgast r q 3 ) , the above chain of events may result in a complete escape of plasma and hence in dense packing of erythrocytes in the distal portion of the capillary (Figs. 5 and 6). The subsequent deformation of the erythrocytes would thus be caused not by elevated hydrostatic pressure but by the complete escape of plasma. Since the trapped erythrocytes obstruct the capillary, the extravasation and trapping will proceed in the retrograde direction until the whole capillary has been filled with aggregates of deformed erythrocytes. Howc\ er, if the macromolecular permeability in a part of a capillary loop remains normal, the above complete extravasation of plasma will be prevented by the subsequent rise in the intravascular colloid osmotic pressure; that is, part of a capillar!. loop may escape trapping during the 20 min of recirculation. The present study thus shows that in some areas (primarily in the cortex) there is only transient ‘congestion ’ of normally shaped erythrocytes. In capillaries with a long transit time, as in the renal medulla (Wolgast 1973), the membrane damage may lead to true ‘trapping’ of polygonally shaped erythrocytes, obstructing the blood flow. When the blood flow is blocked, the hydrostatic pressure in the capillaries proximal to the blockade will rise progressively. Since the damaged membrane also seems to weaken the supporting basement membrane, the elevated hydrostatic pressure proximal to the blockade will cause dilatation of the capillary, so that a large number of erythrocytes will accumulate within it (Fig. 6). It must be mentioned that the dilatation of the capillary and the retrograde formation of the aggregates of polygonally shaped erythrocytes are parallel phenomena. This will mean that although the erythrocyte volume, measured as the distribution of 5’Cr-labelled erythrocytes, is as large as the capillary volume in the inner stripe of the outer medulla (Karlberg et al. 1982a) there will be some capillaries that will escape the damage and maintain their blood flow. This is in accordance with a previous finding that after ischaemic damage the med-
ullary blood flow of the kidney decreased to about 10% of the normal. Since this distension of the capillaries also seems to compress the adjacent tubules (Fig. 7), it might contribute to the tubular obstruction which is typical for ischaemic acute renal failure (Arendhorst et al. 1975, NorlCn et al. 1978b, Karlberg et al. 1982b). It is concluded that aggregates of polygonally shaped erythrocytes in the inner stripe of the outer medulla, in particular, are caused by total extravasation of plasma proteins, due to an increase in the macromolecular permeability of the capillaries. This, in turn, may be evoked by the direct or indirect action of oxygen-derived free radicals generated soon after the onset of recirculation. These aggregates would obstruct the blood flow, sustaining medullary ischaemia. This work was supported by the Swedish Medical Research Council (grant B87-oqX-02867-16.
REFERENCES .-\RENDHORST, W.J.,FINN,W.F. & GOTTSCHALK, C.W.
1975. Pathogenesis of acute renal failure following temporary renal ischemia in the rat. Circ Res 37, j j8-568. B.A;ATI,A,, HELLBERG, O., ODLIND, B. & W’OLGAST, 34. 1987. Prevention of acute renal failure by superoxide dismutase and sucrose. Acta Phjxiol Scand ~ 3 0 367-372. , DELMAESTRO, R.F.,BJORK, J. & ARFORS,K.E. 198I a. Increase in microvascular permeability induced by enzymatically generated free radicals. 11. Role of superoxide anion radicals, hydrogen peroxidase and hydroxyl radicals. Microvas Res 22, 255-270. DELMAESTRO, R.F., BJORK, J. & ARFORS,K.E. 1981 b. Increase in microvascular permeability induced by enzymatically generated free radicals. I. In vivo study. Microcasc Res 22, 239-254. DURBIN, R.P. 1960. Osmotic flow of water across permeable cellulose membranes. 3 Gen Ph-ysd 44, 31j-326. FINK,M.R. & LENGFELDER, E. 1986. Hyaluronic acid degrading reactions under hyperthermic conditions and effects of thiols. In: G. Rotilio (ed.) Sufwoside (irrtl Si(ptrv.wit! Dtsniiitrrse in Chrwiisrr)!,PI’. 60--63. Else\ ier, New Yo&. f1ANSSON, R., JONSSON, 0.& LUNDSTAM, S. 1983. Effects of free radical scavengers on renal circulation after ischemia in the rabbit: Clin ,Sci 65, 605-610. HELI.BERG. 0. 8z OITEG. G. 1088. ,~ 0.. KXL.LSKOG. ~, Permeability of renal capillaries after ischaemic
Erythrocyte trapping in acute renal failure damage in the rat. Influence of neutrophils. Acta Physiol Scand 132, 4 9 4 3.9. KARLBERG, L., KALLSKOG, O., NYGREN,K. & WOLGAST, M. 1982a. Erythrocyte and albumin distribution on the kidney following warm ischemia. Scand 3 Urol Nephrol 16, 173-177. KARLBERG, L., KALLSKOG,O., NORL~N,B.J. & WOLGAST, M. 1982b. Nephron function in postischemic acute renal failure. Scand 3 Urol Nephrol 16, 167-172. KARLBERG, L., NORL~N, B.J., KALLSKOG, O., OJTEG, G. & WOLGAST,M. 1983. Impaired medullary circulation in post-ischaemic acute renal failure. Acta Physiol Scand 118,11-17. LARSSON, M., SJ~QUIST, M. & WOLGAST, M. 1984. Renal interstitial volume of the rat kidney. Acta Physiol Scand 120, 297-304. LOWREY,O.H., ROSBROUGH, N.J., FARR,A.L. & RANDALL, R.J. 1951. Protein measurements with Folin phenol reagent. 3 Biol Chem 193, 265-275. MASON,J. 1988. The cause and the consequences of the perfusion defect that follows renal ischaemia. In: A.M. Davidson (ed.) Nephrology vol. 11, pp. 955363. Cambridge University Press, Cambridge. MCCORD,J.M. & ROY, R.S. 1982. The pathophysiology of superoxide : roles in inflammation and ischemia. Can 3 Physiol Pharmacol 60, 1 3 4 6 '352. NORLEN,B.J., ENGBERG, A., KALLSKOG, 0. & WOLGAST, M. 1978a. Intrarenal hemodynamics in the transplanted rat kidney. Kidney Int.!4, 13. NORL~N, B.J., ENGBER, A., KALLSKOG, 0.& WOLGAST, M. 1978b. Nephron function of the transplanted rat kidney. Kidney Int 14, 10-20.
23
OJTEG,G., BAYATI, A,, KALLSKOG, 0.& WOLGAST, M. I 987. Renal capillary permeability and intravascular red cell aggregation after ischaemia. I. Effects of xanthine oxidase activity. Acta Physiol Scand 129, 295-304. ORIEL,K., SMEDIRA, N.G. & RICOTTA,J.J. 1985. Protection of the kidney after temporary ischemia : free radical scavengers. 3 Vasc Surg 2, 49-53. PALLER,J.S., HOIDAL,J.R. & FERRIS, T.F. 1984. Oxygen free radicals in ischemic acute renal failure in the rat. 3 Clin Invest 74, I 1 5 6 1 1 6 4 . PARKS, D.A., SHAH,A.K. & GRANGER, D.N. 1984. Oxygen radicals : effects on intestinal vascular permeability. Am 3 Physiol247, G167-G170. REED,R.K. 1988. Transcapillary albumin extravasation in rat skin and skeletal muscle: effect of increased venous pressure. Acta Physiol Scand 134, 375-382. STRIPE, F. &DELLACORTE,E. 1969.The regulation of rat liver xanthine oxidase. 3 Biol Chem 244, 385 5-3863. VALSALA, K. & AVASTHI,P.S. 1987. The anionic sites at the luminal surface of peritubular capillaries in rats. Kidney Int 31, 52-58. VANN, P.G. & FITZ-GERALD,J.M. 1982. Flow mechanics of red cell trains in very narrow capillaries. Microvasc Res 24, 2 9 6 3 13. WOLGAST, M. 1973. Renal medullary red cell and plasma flow as studied with labelled indicators and internal detection. Acta Physiol Scand 88, 215-225. WOLGAST, M., KARLBERG, L., KALLSKOG, O., NORL~N, B.J. & NYGREN, K. 1982. Hemodynamic alterations in ischemic acute renal failure. Nephron 31,301-303.
Mechanism of erythrocyte trapping in ischaemic acute renal failure A. B A Y A T I , R. CHRISTOFFERSON*, 0. KALLSKOG and M. WOLGAST Departments of Physiology and Medical Biophysics, and * H u m a n Anatomy, University of Uppsala, Sweden A., CHRISTOFFERSON, R., KALLSKOG, 0. & WOLGAST, M. 1990. Mechanism of BAYATI, erythrocyte trapping in ischaemic acute renal failure. Acta Physiol Scand 138, 13-23, Received 8 June 1989, accepted 28 August 1989. ISSN 00014772. Departments of Physiology and Medical Biophysics, and Human Anatomy, University of Uppsala, Sweden. Forty-five minutes of warm ischaemia and 20 min of recirculation in the rat kidney was found to result in ( I ) a massive transient extravasation of plasma upon recirculation and (2) an increase in plasma-lymph transport of proteins during the first hours after onset of circulation. This was accompanied by trapping of erythrocytes, as determined with 'lCr-labelled erythrocytes, in the capillaries, mainly in the inner stripe of the outer medulla. At scanning electron microscopy of vibratome sections, the trapping appeared as aggregates of polygonally shaped erythrocytes. It is concluded that 45 min of ischaemia and 20 min of recirculation results in an increase in the permeability of the renal capillaries. This increase leads to extravasation of capillary plasma with consequent local haemoconcentration, causing an increase in vascular resistance and in capillary hydrostatic pressure. This elevated pressure will, in turn, lead to perpetuating extravasation of plasma, further haemoconcentration and so on, eventually resulting in dense packing of polygonal erythrocytes, obstructing the blood flow. It is believed that oxygen-derived free radicals generated in the early recirculation phase contribute to the increase in macromolecular permeability, since the scavenger bovine superoxide dismutase and allopurinol, a xanthine oxidase inhibitor, were found to prevent this unfavourable chain of events.
Key words : acute renal failure, allopurinol, congestion, erythrocyte trapping, kidney, permeability, superoxide dismutase (SOD).
I t has been proposed that trapping of erythrocytes in the kidney - primarily in the inner stripe of the outer medulla - is of great importance in the pathogenesis of post-ischaemic acute renal failure (NorlCn et al. 1978a, Karlberg et al. 1982a, Wolgast et al. 1982, Mason 1988). T h e sustained medullary ischaemia subsequent to this trapping (Karlberg et al. 1982b, 1983) may thus underlie the reduced capacity for urine concentration and the disturbed potassium secretion, both of which are typical features in post-ischaemic acute renal failure. Correspondence : A. Bayati, Department of Physiology and Medical Biophysics, University of Uppsala, Box 572, s-751 23 Uppsala, Sweden.
T h e trapping has been found to occur in the early recirculation phase u p to 20 min after the onset of circulation (Karlberg et al. 1982a). T h e factors responsible for this unfortunate event remain, however, unexplained. Since the trapped erythrocytes occupied almost the entire vascular space of the renal medulla (Karlberg et al. 1982a), extravasation of plasma through a damaged capillary membrane (Hansson et al. 1983, Karlberg et al. 1983, Bayati et al. 1987, Ojteg et al. 1987) might be the first step in this process. In order to test this hypothesis, the following were investigated on termination of induced ischaemia : (a) T h e plasma-lymph transport of macroI3
11
,4. Bayati et al.
molecules during the first 2 h after the onset of recirculation. (b) 'The trapping of erythrocytes in terms of the distribution volume of "'Cr-labelled erythrocytes in the different kidnej- regions. (c) The extravasation of plasma during the first minute after onset of circulation, as studied from the haematocrit of renal venous blood. (d) T h e morphology of the kidney, in particular the inner stripe of the outer medulla, by scanning electron microscopy of vibratome sections. Since oxygen-derived free radicals may be responsible for the increase in macromolecular permeability (Del Maestro et al. 1981a, b), the effects of bovine copper/zinc superoxide dismutase @-SOD) as a scavenger of oxygenderived free radicals and the effects of allopurinol, an inhibitor of xanthine oxidase, on these parameters were also studied.
macromolecular permeability, as reflected by the plasma-lymph transport of the plasma proteins, was studied in 16 rats weighing 301 F 10 g. For this purpose, a hilar lymph vessel was dissected free and cannulated with a 100-zoo p m polyethylene catheter prefilled with heparinized saline. Lymph was sampled at 30-min intervals both before the ischaemia and up to 120 min after recirculation. Rats in which the lymph flow before clamping was less than I yl min-' were discarded. The volumes of the samples were measured from their length in 75-pI constant-bore capillaries (Drummond Scientific Co., USA). Plasma samples were withdrawn in the middle of each p m i n period. The protein concentrations of plasma and lymph were measured by the method of Lowrey et al. (1951).The protein transport was calculated by multiplying the lymph flow by the lymph protein concentration. The experiments were performed both in untreated rats and after intravenous administration of up to 40 mg of b-SOD (Grunenthal GmbH, FRG), given 30 s before the onset of recirculation (Bayati et al. 1987).
M A 4 T E R I A I , S AND M E T H O D S The experiments were performed on 89 male Sprague-Dawley rats (Alab, Stockholm, Sweden) weighing 281 F 4 g. They had free access to standard rat chow (Rj, EWOS, Sweden) and water prior to the experiments. The animals wrere anaesthetized with Inactin (Byk Gulden, Konstanz, FRG) given intraperitonealll- in a dose o f 120 mg kg ', After tracheostomy they were placed on a servo-controlled heating pad to stabilize the body temperature at 3j . j "C. The left femoral artery was cannulated for monitoring of the sj-stemic blood pressure and withdrawal o f blood samples. The left femoral vein was cannulated for administration of test substances and for continuous infusion of Ringer-bicarbonate, which was given at a rate of 5 ml kg h- I . All the animals received 40 IC: of heparin intravenously 30 min before clamping of the renal artery. The left kidney was exposed through a flank incision, dissected free from its perirenal tissue in order to cut off collateral blood supply, and suspended in a Lucite cup. Ischaemia was evoked by clamping the renal artery for 45 min, during which time the kidney was placed in its natural position with the abdomen closed. In order to secure a kidney temperature of 3 7 . 5 "C, cotton wool was placed over the incision. After termination of the ischaemia, the abdomen was reopened, the kidney was resuspended in the L,ucite cup, and the arterial clamp was removed. In all the animals the right kidney served as a control. The animals were divided into four main groups: ( I ) .Wacromolecular permeability. The transcapillary
( 2 ) Erythrocyte distribution volume. The distribution volume of "Cr-labelled erythrocytes was studied in 33 rats weighing 292 & 36 g. For the purpose of labelling, 10 nil of blood from a donor rat was dissolved in I ml of ACD solution (Travenol Laboratories, Thetford, Norfolk, England) and centrifuged at 1000 r.p.m. for 10min. After removaI of the supernatant, roo yCi of 51Cr(sodium chromate, Amersham International, Amersham, UK) was added and the sample was incubated at room temperature for 20 min. The erythrocytes were then washed four times and resuspended in an equal volume of saline. One millilitre of the suspension was injected intravenously 10min before clamping of the renal artery. After 20 rnin of recirculation the two kidneys were excised. For microdissection the two poles of the kidney were first removed. A triangular core was then obtained by making two sagittal incisions in the striatal direction. Thereafter the core was subdivided into: (a) cortex, (b) outer stripe of the outer medulla, (c) inner stripe of the outer medulla and (d) inner medulla. The "Cr activity of the tissue specimens and that of the erythrocytes in the circulating blood withdrawn at the same time were measured by gamma-spectrophotometry. The erythrocyte distribution volume was calculated as the ratio of the activity in the kidney specimens to that in the erythrocytes in systemic blood and is represented as per cent of the zonal volume. The material was subdivided into three groups : (a) A group of 10 rats in which the left kidney was subjected to 45 rnin of ischaemia and zomin of recirculation. (b) -4group of 18 rats subjected to the same trauma
Erythrocyte trapping in acute renal failure
15
but treated with up to 20 mg of b-SOD given The triangular core and the two kidney poles were intravenously 30 s before recirculation. immersion-fixed for 24 h at room temperature, using (c) A group of five rats subjected to the same 2.5 yo glutaraldehyde (Kebo Lab AB, Stockholm, trauma but treated with 40 mg of allopurinol (Sigma Sweden) in a phosphate buffer solution (pH 7.4). The Chemical Co., MO, USA), given intravenously 4 h osmolalities of the phosphate buffer solution used as before clamping of the kidney. carrier and the fixative solution were 280 and 530 (3) Extravasation of plasma. The extravasation of mosmol I-' respectively. plasma during the first minute after onset of circulation After the above prefixation with glutaraldehyde, the as determined from the haematocrit in renal venous specimens were washed four times with the carrier. blood was studied in 20 rats weighing 258 4 g. The samples were then suspended in the same carrier For the catheterization of the renal vein a sharpened and cut into I O ~ Z O Opm thick slices with an Oxford steel cannula with an outer diameter of 0.9 mm was vibratome set at its maximal frequency and lowest inserted through the vessel wall. After 15 rnin the steel speed. However, in order to facilitate the sectioning, a cannula was gently removed, whereupon the puncture rigid frame surrounding the specimen was obtained hole would be sealed with fibrin. A silicon catheter by immersing it for 5 min in I % osmium tetroxide (outer diameter 0.9 mm, inner diameter 0.5 mm) (Expectron Man. AB, Stockholm, Sweden), dissolved could then easily be inserted through the preformed in the same carrier. The sections were then washed again and post-fixed hole. This narrow catheter was attached to a wider silicon tubing which was used to harvest the blood at room temperature in 1 % osmium tetroxide samples; this had an outer diameter of 3 mm and an overnight. Thereafter surplus fixative was removed inner diameter of I mm. The tubing was connected to and the specimens were dehydrated in a graded series a suction pump set at a rate of 0.6 ml min-'. In order of acetone. After 12 h in absolute acetone, the sections to separate the blood samples from each other, small were dried in a critical point drying apparatus air bubbles were sucked into the tubing every 2 s, (Polaron, Agar, UK), mounted .on Cambridge aluusing a revolver principle described previously (Lars- minium stubs, coated with a 400 A gold layer in a fine son et al. 1984). coat sputter (Jeol JFC I 100)and observed in a Philips The sampling was started 10 s before re- 525 scanning electron microscope (SEM) at an circulation and continued for up to 50 s after. acceleration voltage of 20 kV. During this time 30 samples, each with a volume To rule out the possibility of a preparational of zopl, were harvested. They were transferred artifact, parallel studies on the right control kidney to 2 0 4 heparinized microcapillaries and, for were performed in all series. The finding of normal measurement of the haematocrit, centrifuged at erythrocytes (Fig. I b) and podocytes, both of which 9000 r.p.m. are sensitive to preparation mishaps, provides evidence In this series parallel studies of the intrarenal against such an artifactual cause of the polygonally erythrocyte volume were also made. The material was shaped erythrocytes found in these experiments. subdivided into four groups : In order to quantify the erythrocyte trapping, 30 (a) A 'control' group of five rats, in which the left photographs from each kidney (original magnification kidney was subjected to only 5 rnin of ischaemia, i.e. 100-1100) were analysed with respect to the pera minimal trauma. centage area occupied by capillaries, using a MOP (b) A group of five rats in which the left kidney was system (Quantitative Bilauswertung, Kontorn Messsubjected to 45 rnin of ischaemia and 20 rnin of gerate GmbH, FRG). recirculation. The material was subdivided into the following (c) A group of five rats subjected to the same groups : ischaemia and recirculation but treated with 20 mg of (a) A group of five rats in which the left kidney was b-SOD given intravenously 30 s before recirculation. subjected to 45 min of ischaemia but with no (d) A group of five rats subjected to the same recirculation. ischaemia and recirculation but treated with 40 mg of (b) A group of five rats subjected to both 45 min of allopurinol in a 67 mM NaOH/saline solution (pH 9) ischaemia and 20 min recirculation. given intravenously 4 h before the clamping of the (c) A group of five rats subjected to 45 min of kidney. ischaemia and 20 min of recirculation, but which (4) Morphological studies. The morphology of the recieved a single intravenous dose of 20 mg of b-SOD left experimental and right control kidney was 30 s before recirculation. investigated in 20 rats weighing 273 f7 g. (d) A group of five rats subjected to the same In order to prevent collapse of the tubules resulting ischaemia and recirculation, but treated with a single from the reabsorption of fluid which occurs soon after intravenous dose of 40 mg of allopurinol 4 h before nephrectomy, z ml of a 12% sucrose-Ringer solution clamping of the kidney. was injected intravenously 10 rnin prior to the removal All the results are presented as mean _+ SEM. The of the kidneys. comparison between the different groups was done
16
A . Bayati et al.
Fig. I . Scanning electron micrographs of the inner stripe of the outer medulla of a normal kidney (a x 6jo; b x 1490) and of kidneys subjected to both ischaemia and recirculation (c x 650; d x 1490).The capillaries of the damaged kidneys (c and d) are wide open and are occupied by
deformed erythrocytes.
null hvpothesis.
the mass transport of proteins was the same as that in the preischaemic period. The same tendency could also be traced in the group of animals treated with 20 mg of b-SOD.
RESULTS
Erythrocyte distribution volume
using Student’s t-test. T h e results of this comparison are presented as the highest possibility in favour of the
As shown in Fig. z(b), 45 min of ischaemia and 20 min of recirculation resulted in trapping of In kidneys subjected to 45 min of ischaemia and erythrocytes, mainly in the inner stripe of the 20 min of recirculation, the lymph protein outer medulla, comprising IS”/:,of the inner concentration increased from a control value of stripe volume. T h e corresponding figure for the 1 . 7 & 0 . 1 3 to2.210.23 mg ml-’(Fig. 2a). At the inner zone was 676, while in the cortex no same time the lymph flow increased from significant increase was observed. In contrast, kidneys subjected to only 5 min of I .6 0.13 to 2.9 fo.64 pl min-’ and the mass transport from 2. j i0.25 to 7.3 & 0.85 mg min-’ ; ischaemia showed essentially no increase in the thus there was a threefold increase in the net distribution volume of 51Cr-labelled erythrocytes (Fig. 3 b). transcapillary transport. After treatment with 40 mg of b-SOD, the Treatment with 40 mg of allopurinol reduced increase in the lymph flow was outweighed by a the trapping considerably (Fig. 3 b), and 20 mg reduction in the protein concentration such that of b-SOD had the same effect (Fig. 2b). .Ifarrornolecular transport
Erythrocyte trapping
1
acute renal failure
h
E '
E
i
E"
T
Cortex
8
I7
Outer stripe Inner stripe
Inner zone
Fig. 2. The upper panel (a) shows the average plasma-lymph data of plasma proteins in the period up to 1.5 h after the onset of recirculation (mean +SEM) in different groups: normal kidneys ischaemic control kidneys (a),ischaemic kidneys treated with 20 mg of b-SOD (a)and ischaemic kidneys treated with 40 mg b-SOD (a).The lower panel (b) depicts the percentage regional renal erythrocyte volume after 20 min of recirculation as investigated by the use of X r labelled erythrocytes in the same groups as in (a). *represents a probability less than 0.05.
(a),
Extravasation of plasma
The haematocrit of renal venous blood before and shortly after the onset of circulation 'is depicted in Fig. 3 (a). I n kidneys subjected to 45 min of ischaemia the haematocrit increased from an initial value of 46 1.6 to a peak value of 61 ko.7 after 15 s of recirculation, whereafter it
returned to an almost normal value within the following minute. I n contrast. there was no alteration in the haematocrit 'in the kidneys subjected to only 5 min of ischaemia. As seen in Fig. 3 (a), treatment with 20 mg of b-SOD or 40 mg of allopurinol effectively prevented the post-circulation rise in haematocrit resulting from 45 min of ischaemia.
18
A. Bayati et al.
62
50
44
I
10
0
I
I
20
30
Time
I
I
40
’
50
(sec)
10
0 Cortex
Outer stripe
Innerstripe
lmer zone
Fig. 3. T h e upper panel (a) illustrates the venous haematocrit changes in the early recirculation phase after 5 min (*) and 45 min of ischaemia in untreated kidneys (a),kidneys treated with 20 mg of b-SOD and kidneys treated with 40 mg of allopurinol (V). Note the very steep increase in the haematocrit in the untreated kidneys after 15 s of recirculation. This increase was almost normalized after administration of b-SOD or of allopurinol. In the lower panel (b) the percentage erythrocyte volumes in different regions of the same kidneys are shown. Right control kidneys (0) and left kidneys after 5 min of ischaemia (m), 45 rnin of ischaemia with no treatment and 45 min of ischaemia treated with 40 mg of allopurinol (m).
(m),
(a)
Erythrocyte trapping in acute renal failure 40
19
gonally shaped erythrocytes. Apart from this occasional finding, the morphology was similar to that of the right control kidney. The morphometric studies on kidneys subjected to 45 rnin of ischaemia and 20 min of recirculation showed that in the inner stripe of the outer medulla as much as 34kz.oyo of the total area was occupied by capillaries. This is in sharp contrast to the kidneys subjected to 45 min of ischaemia but with no recirculation, where only 4 f0.3 yo was occupied by capillaries. Treatment with 20 mg of b-SOD or 40 mg of allopurinol diminished this area to almost normal values (Fig. 4).
F
Fig. 4. The percentage area of the inner stripe occupied by capillaries in right control and left damaged kidneys in different groups. Right control kidney left kidneys after 45 rnin of ischaemia with no recirculation (a)and kidneys subjected to 45 min of ischaemia and 20 rnin of recirculation with no treatment ( 0 )treated with 20 mg of b-SOD (a)and treated with 40 mg of allopurinol (a).
(a),
DISCUSSION
In our view the key event underlying the erythrocyte trapping after 45 min of ischaemia and 20 min of recirculation is a dramatic increase in the macromolecular permeability of the renal capillaries. There is a possibility, however, that the increased mass transport of proteins into the lymphatic system might be Morphological studies caused by an increase in the transcapillary After 45 min of warm ischaemia and 20 min of hydrostatic pressure difference, since the fluid recirculation, the peritubular capillaries of the filtration also increases the transport of large renal medulla were distended and contained molecules by convection. aggregates of polygonal erythrocytes (Fig. I c). As discussed by Reed (1988), however, the The tubules appeared to be compressed by the enhanced fluid filtration would lead to a dilution surrounding capillaries (Fig. ~ d and ) in most of the lymph proteins, which would be incases exhibited no lumen. The vasa recta showed compatible with the present finding that ischano such aggregates, but contained normal-shaped emia and recirculation resulted in an elevated erythrocytes. lymph protein concentration. Moreover in a These findings were in sharp contrast to those recent study at this laboratory it was found that in the right control kidney (Fig. ra). Here, the the transcapillary hydrostatic pressure difference peritubular capillary lumen displayed its normal did not change after ischaemic damage (Hellberg narrow dimension and only occasionally con- et al. 1988). It would thus seem likely that the tained erythrocytes (Fig. I b). When erythrocytes increased transport of macromolecules is due to were present, they were of normal size and a true increase in permeability (Durbin 1960), an shape. Nearly all tubules were open and had interpretation which is in accordance with that in previous studies both in the kidney (Hansson their normal rounded appearance. After 45 rnin of ischaemia but without recircu- et al. 1983, Karlberg et al. 1983, Bayati et al. lation, the overall morphological picture of the 1987, Ojteg et al. 1987) and in other organs (Del inner stripe of the outer medulla was much 1;ke Maestro et al. 1981a, b, McCord & Roy 1982, that in the control kidneys, although the tub. rr Parks et al. 1984). Since SOD (a scavenger of oxygen free epithelial cells seemed swollen. No polygo~ial .,dicals) and allopurinol (an inhibitor of xanthine erythrocytes were observed. tidase) were able to restore the plasma to In kidneys subjected to ischaemia and recirculation, but treated with either b-SOD or lymph transport of proteins, the injury to the allopurinol, only some of the peritubular capil- capillary barrier seems to be caused by oxygenlaries were obstructed by aggregates of poly- derived free radicals (DeI Maestro et a[. 1981a,
20
A . Bayati et al.
Fig. 5. The upper diagram illustrates that, in a capillary with a comparatively short transit time, the extravasation of plasma leads to congestion of erythrocytes in the distal end of the capillary soon after the onset of recirculation. In the lower diagram, which represents a capillary with a long transit time, all plasma has escaped from the capillary. The erythrocytes are thereby forced, in the ideal case, to assume a hexagonal shape.
b, Paller et al. 1984, Parks et af. 1984, Oriel et al. 1985). The course of events underlying the production of these free radicals may be as follows: During the ischaemia and the subsequent anoxia, the purine metabolism leads to accumulation of hypoxanthine (Hansson et al. 1983). In addition, the inactive P form of xanthine oxidase is converted into its active 0 form (Stripe & Della Corte 1969). After the onset of circulation and return of oxygen, the latter form of xanthine oxidase will convert the accumulated hypoxanthine into xanthine and further to uric acid. In each of these steps an oxygen free radical (0;)is generated. Since this radical is known to break down polysaccharides, such as glucosaminoglycans (Fink & Lengfelder 1986), of the capillary barrier (Valsala & Avasthi 1987), 0; is a likely candidate for the damage to the capillary barrier. The effect of these oxygen radicals would of course be on all the capillaries of the kidney, although with different intensity. The fact that the trapping is mainly seen in the outer medulla of the kidney could be explained (see below) by additional structural and rheological characteristics of this special part of the kidney. Regarding the mechanism of the erythrocyte trapping, the essential event, in our opinion, is not a compression of the erythrocytes but a complete escape of plasma. This may be explained as follows (Fig. 5 ) :
Firstly, major transport of proteins into the perivascular space will lead to a reduction of the colloid osmotic pressure difference. The consequent large outward filtration of fluid will result in local haemoconcentration (Fig. 3 a) and thereby in an increase in vascular resistance. The fact that the blood flow will be reduced will also, in itself, increase the vascular resistance (Vann & Fitz-Gerald 1982). T h e concomitant rise in capillary hydrostatic pressure will then lead to further extravasation of plasma, further haemoconcentration, a further increase in hydrostatic pressure, and so on. It is also evident that for one and the same macromolecular permeability, a specific volume of blood will lose more of its plasma the longer the transit time (Fig. 5). In capillaries with a short transit time, as in the cortex (Wolgast 1973), the above process will only lead to ‘congestion’ of the erythrocytes in the distal part of the capillary. T h e capillaries will thereby remain patent to the blood flow. It may also be expected that as soon as the macromolecular permeability is restored, the intracapillary haematocrit and the blood flow will also be restored. This idea is supported by the present finding of a transient increase in the haematocrit in renal venous blood soon after the onset of the recirculation (Fig. 3 a). T h e finding that b-SOD and allopurinol both reduced early extravasation of plasma emphasizes the role of oxygen-derived
Erythrocyte trapping in acute renal failure
Fig. 6. The polygonally shaped erythrocytes in a peritubular capillary in the inner stripe of the outer medulla of a kidney subjected to 45 min of ischaemia and 20 min of recirculation ( x 5270). Note the very distended capillary containing a large number of deformed erythrocytes.
Fig. 7. Scanning electron micrograph of the inner stripe of the outer medulla, showing the compression of the tubules by the distended capillaries ( x 4080).
21
22
.4. Ba-luti et al.
free radicals in the ischaemia-recirculation damage. In the renal medulla, where the transit time is -5-10 times longer than in the cortex (Wolgast r q 3 ) , the above chain of events may result in a complete escape of plasma and hence in dense packing of erythrocytes in the distal portion of the capillary (Figs. 5 and 6). The subsequent deformation of the erythrocytes would thus be caused not by elevated hydrostatic pressure but by the complete escape of plasma. Since the trapped erythrocytes obstruct the capillary, the extravasation and trapping will proceed in the retrograde direction until the whole capillary has been filled with aggregates of deformed erythrocytes. Howc\ er, if the macromolecular permeability in a part of a capillary loop remains normal, the above complete extravasation of plasma will be prevented by the subsequent rise in the intravascular colloid osmotic pressure; that is, part of a capillar!. loop may escape trapping during the 20 min of recirculation. The present study thus shows that in some areas (primarily in the cortex) there is only transient ‘congestion ’ of normally shaped erythrocytes. In capillaries with a long transit time, as in the renal medulla (Wolgast 1973), the membrane damage may lead to true ‘trapping’ of polygonally shaped erythrocytes, obstructing the blood flow. When the blood flow is blocked, the hydrostatic pressure in the capillaries proximal to the blockade will rise progressively. Since the damaged membrane also seems to weaken the supporting basement membrane, the elevated hydrostatic pressure proximal to the blockade will cause dilatation of the capillary, so that a large number of erythrocytes will accumulate within it (Fig. 6). It must be mentioned that the dilatation of the capillary and the retrograde formation of the aggregates of polygonally shaped erythrocytes are parallel phenomena. This will mean that although the erythrocyte volume, measured as the distribution of 5’Cr-labelled erythrocytes, is as large as the capillary volume in the inner stripe of the outer medulla (Karlberg et al. 1982a) there will be some capillaries that will escape the damage and maintain their blood flow. This is in accordance with a previous finding that after ischaemic damage the med-
ullary blood flow of the kidney decreased to about 10% of the normal. Since this distension of the capillaries also seems to compress the adjacent tubules (Fig. 7), it might contribute to the tubular obstruction which is typical for ischaemic acute renal failure (Arendhorst et al. 1975, NorlCn et al. 1978b, Karlberg et al. 1982b). It is concluded that aggregates of polygonally shaped erythrocytes in the inner stripe of the outer medulla, in particular, are caused by total extravasation of plasma proteins, due to an increase in the macromolecular permeability of the capillaries. This, in turn, may be evoked by the direct or indirect action of oxygen-derived free radicals generated soon after the onset of recirculation. These aggregates would obstruct the blood flow, sustaining medullary ischaemia. This work was supported by the Swedish Medical Research Council (grant B87-oqX-02867-16.
REFERENCES .-\RENDHORST, W.J.,FINN,W.F. & GOTTSCHALK, C.W.
1975. Pathogenesis of acute renal failure following temporary renal ischemia in the rat. Circ Res 37, j j8-568. B.A;ATI,A,, HELLBERG, O., ODLIND, B. & W’OLGAST, 34. 1987. Prevention of acute renal failure by superoxide dismutase and sucrose. Acta Phjxiol Scand ~ 3 0 367-372. , DELMAESTRO, R.F.,BJORK, J. & ARFORS,K.E. 198I a. Increase in microvascular permeability induced by enzymatically generated free radicals. 11. Role of superoxide anion radicals, hydrogen peroxidase and hydroxyl radicals. Microvas Res 22, 255-270. DELMAESTRO, R.F., BJORK, J. & ARFORS,K.E. 1981 b. Increase in microvascular permeability induced by enzymatically generated free radicals. I. In vivo study. Microcasc Res 22, 239-254. DURBIN, R.P. 1960. Osmotic flow of water across permeable cellulose membranes. 3 Gen Ph-ysd 44, 31j-326. FINK,M.R. & LENGFELDER, E. 1986. Hyaluronic acid degrading reactions under hyperthermic conditions and effects of thiols. In: G. Rotilio (ed.) Sufwoside (irrtl Si(ptrv.wit! Dtsniiitrrse in Chrwiisrr)!,PI’. 60--63. Else\ ier, New Yo&. f1ANSSON, R., JONSSON, 0.& LUNDSTAM, S. 1983. Effects of free radical scavengers on renal circulation after ischemia in the rabbit: Clin ,Sci 65, 605-610. HELI.BERG. 0. 8z OITEG. G. 1088. ,~ 0.. KXL.LSKOG. ~, Permeability of renal capillaries after ischaemic
Erythrocyte trapping in acute renal failure damage in the rat. Influence of neutrophils. Acta Physiol Scand 132, 4 9 4 3.9. KARLBERG, L., KALLSKOG, O., NYGREN,K. & WOLGAST, M. 1982a. Erythrocyte and albumin distribution on the kidney following warm ischemia. Scand 3 Urol Nephrol 16, 173-177. KARLBERG, L., KALLSKOG,O., NORL~N,B.J. & WOLGAST, M. 1982b. Nephron function in postischemic acute renal failure. Scand 3 Urol Nephrol 16, 167-172. KARLBERG, L., NORL~N, B.J., KALLSKOG, O., OJTEG, G. & WOLGAST,M. 1983. Impaired medullary circulation in post-ischaemic acute renal failure. Acta Physiol Scand 118,11-17. LARSSON, M., SJ~QUIST, M. & WOLGAST, M. 1984. Renal interstitial volume of the rat kidney. Acta Physiol Scand 120, 297-304. LOWREY,O.H., ROSBROUGH, N.J., FARR,A.L. & RANDALL, R.J. 1951. Protein measurements with Folin phenol reagent. 3 Biol Chem 193, 265-275. MASON,J. 1988. The cause and the consequences of the perfusion defect that follows renal ischaemia. In: A.M. Davidson (ed.) Nephrology vol. 11, pp. 955363. Cambridge University Press, Cambridge. MCCORD,J.M. & ROY, R.S. 1982. The pathophysiology of superoxide : roles in inflammation and ischemia. Can 3 Physiol Pharmacol 60, 1 3 4 6 '352. NORLEN,B.J., ENGBERG, A., KALLSKOG, 0. & WOLGAST, M. 1978a. Intrarenal hemodynamics in the transplanted rat kidney. Kidney Int.!4, 13. NORL~N, B.J., ENGBER, A., KALLSKOG, 0.& WOLGAST, M. 1978b. Nephron function of the transplanted rat kidney. Kidney Int 14, 10-20.
23
OJTEG,G., BAYATI, A,, KALLSKOG, 0.& WOLGAST, M. I 987. Renal capillary permeability and intravascular red cell aggregation after ischaemia. I. Effects of xanthine oxidase activity. Acta Physiol Scand 129, 295-304. ORIEL,K., SMEDIRA, N.G. & RICOTTA,J.J. 1985. Protection of the kidney after temporary ischemia : free radical scavengers. 3 Vasc Surg 2, 49-53. PALLER,J.S., HOIDAL,J.R. & FERRIS, T.F. 1984. Oxygen free radicals in ischemic acute renal failure in the rat. 3 Clin Invest 74, I 1 5 6 1 1 6 4 . PARKS, D.A., SHAH,A.K. & GRANGER, D.N. 1984. Oxygen radicals : effects on intestinal vascular permeability. Am 3 Physiol247, G167-G170. REED,R.K. 1988. Transcapillary albumin extravasation in rat skin and skeletal muscle: effect of increased venous pressure. Acta Physiol Scand 134, 375-382. STRIPE, F. &DELLACORTE,E. 1969.The regulation of rat liver xanthine oxidase. 3 Biol Chem 244, 385 5-3863. VALSALA, K. & AVASTHI,P.S. 1987. The anionic sites at the luminal surface of peritubular capillaries in rats. Kidney Int 31, 52-58. VANN, P.G. & FITZ-GERALD,J.M. 1982. Flow mechanics of red cell trains in very narrow capillaries. Microvasc Res 24, 2 9 6 3 13. WOLGAST, M. 1973. Renal medullary red cell and plasma flow as studied with labelled indicators and internal detection. Acta Physiol Scand 88, 215-225. WOLGAST, M., KARLBERG, L., KALLSKOG, O., NORL~N, B.J. & NYGREN, K. 1982. Hemodynamic alterations in ischemic acute renal failure. Nephron 31,301-303.
