Cryopreservation of the Mammalian Kidney

35, 114–131 (1997) CY972026

CRYOBIOLOGY ARTICLE NO.

Cryopreservation of the Mammalian Kidney II. Demonstration of Immediate ex Vivo Function after Introduction and Removal of 7.5 M Cryoprotectant

Gregory M. Fahy1 and Suja E. Ali Tissue Cryopreservation Section, Transfusion and Cryopreservation Research Program, Naval Medical Research Institute, Building 29, 8901 Wisconsin Avenue, Bethesda, Maryland 20889 The objective of the present study was to determine whether rabbit kidneys could be perfused with a 7.5 M vitrification solution (VS4, which vitrifies under applied pressure) without loss of function. To answer this question, kidneys were perfused with VS4 using a computer-based machine to gradually raise and lower concentration and then attached to the aorta and vena cava of a perfusor rabbit using an apparatus that permitted renal blood flow and renal function to be measured. About half (6/13) of the kidneys so evaluated resumed substantial immediate function after a transient period of severely reduced blood flow. Loss of function did not occur if cryoprotectant concentration was limited to 3.8 M. The loss of function produced by VS4 could be partially reproduced by artificially limiting blood reflow in control kidneys to simulate the transiently low flows caused by VS4 exposure. These results provide the first evidence that both the parenchyma and the vascular system of a sensitive mammalian organ can survive exposure to a vitrifiable concentration of cryoprotectant. q 1997 Academic Press

We have proposed that the cryopreservation of solid organs might be possible through vitrification (6, 13, 17). Rabbit renal cortical slices retain viability after being treated with vitrifiable media (5, 12, 17), but the ability of the vascular system to survive similar exposure by perfusion must also be demonstrated. We recently described computer-operated perfusion equipment for introducing and removing vitrification solutions (9, 10) and applicable methodology for perfusion and transplantation of control rabbit kidneys (22). We now report methods for the introduction and removal of vitrifiable concentrations of cryoprotectant that permit immediate renal function during normothermic blood reperfusion ex vivo. Preliminary accounts of this work were presented elsewhere (7, 11).

Received August 29, 1996; accepted April 9, 1997. 1 To whom correspondence should be addressed at N. M. R. I.; Dr. Fahy is Chief Scientist, Organ, Inc., 1510 W. Montana St., Chicago IL 60014.

MATERIALS AND METHODS

Animals and Anesthesia All procedures described here were consistent with guidelines established by the Animal Care and Use Committee of the American Red Cross Jerome Holland Laboratory for the Biomedical Sciences, the principles set forth in the ‘‘Guide for the Care and Use of Laboratory Animals’’ (Institute of Laboratory Animal Resources, National Research Council, DHHS, Publication No. (NIH) 86-23 (1985)), and USDA guidelines. New Zealand White rabbits of either sex weighing 2–4 kg were anesthetized by intramuscular injection of 1.14 ml of Ketamine HCl (Parke-Davis, 100 mg/ml) plus 0.46 ml of Xylazine (Haver-Lockhart, 20 mg/ml). Endotracheal intubations were performed (3), and anesthesia was maintained by connecting a stream of 50% oxygen/50% nitrous oxide (300 ml/min of each gas) to an endotracheal tube via a ‘‘T’’ connector. Inhalational anesthesia was supplemented as necessary by slow intravenous (iv) infusion of 1% methohexital

114 0011-2240/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

CRYO 2026

/

a90a$$$161

08-22-97 06:40:29

cryoas

AP: CRYO

115

FUNCTION OF VITRIFIABLE KIDNEYS TABLE 1 Solutions Componenta

DR

WF

CF

RPS-1

RPS1M

VS4

VS1

25% CPA

NaCl Sodium lactate KCl MgCl2 CaCl2 NaHCO3 Dextrose Reduced glutathione Adenine HCl K2HPO4 Na2HPO4 Mannitol HES (Mr 450 kDa) (g/l) Dextran 40 (g/l) Dibenzyline (mg/l) Regitine (mg/l) Heparin (units/l) Papaverine (mg/l) CPZ (mg/l) DMSO (g/l) Formamide (g/l) Acetamide (g/l) 1,2-Propanediol (g/l) PEG 8,000 (g/l) pH mOsm/kg H2O

51 14 2 — 1.4 6 139 — — — — — — — — — — — — — — — — — 7.4 290

77 — — — — — 10 5 — 30 — 80 30 50 5 0.1 1000 30 3 — — — — — 7.6 355

— — 28.3 — — — 30 5 1 7.2 7.7 150 30 — 5 — — — 1 — — — — — 7.6 310

10 — 28.3 2 1 — 180 5 1 7.2 — — — — — — — — — — — — — — 7.4 290

10 — 28.3 2 1 — 10 5 1 7.2 — 170 — — — — — — — — — — — — 7.4 290

10 — 28.3 0.40 0.05 — 180 5 1 7.2 — — — — — — — — — 215.7 124.3 — 150.0 — 7.6 —

10 — 28.3 0.4 0.05 — 180 5 1 7.2 — — — — — — — — — 205 — 155 100 60 8.0 —

10 — 28.3 0.40 0.05 — 180 5 1 7.2 — — — — — — — — — 110.1 63.4 — 76.5 — 7.6 —

a Component amounts given in mM units unless stated otherwise. Note. DR, dextrose–Ringer’s; WF, warm flush; CF, cold flush; HES, hydroxyethyl starch; CPZ, chlorpromazine (used as a 1 mg/ml solution in 0.9% NaCl); DMSO, dimethyl sulfoxide. Dibenzyline stock, 1 mg/ml, in 0.9% NaCl. Regitine was supplied in 100 ml of lactated Ringer’s solution; RPS-2, RPS-1 with 10 mM NaHCO3 replacing 10 mM NaCl.

sodium (Brevital; Eli Lilly & Co., Indianapolis, IN) via the marginal ear vein. Body (rectal or esophageal) temperature was monitored and maintained with a heating pad. Donor Kidney Preparation A midventral laparotomy was performed and the gut reflected to the right and exteriorized into a plastic bag containing gauze moistened with normal saline. IntraMedic polyethylene tubing (PE 50) was used to cannulate the left ureter. The donor was hydrated with 40 ml of dextrose–lactated Ringer’s solution (DR, Table 1) and 20 ml of 5% mannitol. One milliliter furosemide (1.67 mg/ml, in

AID

CRYO 2026

/

a90a$$$161

saline), 3 ml chlorpromazine (CPZ; 1 mg/ml, in saline), and 2 ml heparin (1000 units/ml, in saline) were given iv. CPZ was useful for blocking vasospasm and permitting a more effective and consistent flush despite its hypotensive effects. Procaine (1%) and papaverine (0.003%) were also initially injected into the fat surrounding the left renal artery and vein and later used to irrigate the free vessels as needed to prevent or reverse vasoconstriction. Some hilar fat was often left in place to avoid damaging the renal vessels at the hilum, and most perirenal fat was left undisturbed until the moment the kidney was removed. The renal artery was ligated, incised, and

08-22-97 06:40:29

cryoas

AP: CRYO

116

FAHY AND ALI

FIG. 1. Modified ex vivo perfusion system. ET, endotracheal tube; IV, intravenous fluid line; SR, saline rinse syringe; BF, syringe barrel used for timed blood collection; SC, three-way stopcock; DI, dobutrex infusion syringe; SP, syringe pump; CE, cannula extensions (black segments are shunt tips, white segment is silastic tubing); LA, length adjustor (horizontal piece); V, venous line (to vena cava); A, arterial line (to aorta); BC, blood sample collection syringe; TD, Trimed diaphragm; SG, sphygmomanometer gauge. The tubing segment shown with a manual clamp attached was used to purge air from the line prior to use. (Inset) Cannula securing technique: V, vessel, S, silastic tubing.

cannulated in situ. The cannula was a beveled Extracorporeal T-120 tapered Scribner shunt tip (22) (equivalent to tips currently available from Akcess, North Brunswick, NJ). Immediately after cannulation, the kidney was flushed with a mildly hypertonic room temperature solution (warm flush solution, WF, Table 1) from a hanging bottle at a pressure of approximately 40 mm Hg. The average time from artery ligation to flushing was 1 min. The tapered shunt tip was secured in the artery as shown in the inset to Fig. 1. While continuing arterial perfusion, the renal vein was cannulated with a nonbeveled, nontapered size T110 shunt tip. When blood no longer appeared in the venous effluent, the flushing solution was changed by means of a stopcock, the new cold (47C) flush solution (CF, Table 1) passing through a coil of tubing covered with ice before reaching the kidney. During this in vivo cold flush, the kidney was cleared of perirenal

AID

CRYO 2026

/

a90a$$$161

fat and connective tissue. Once the kidney was cold to the touch, the flushing solution was changed again via another stopcock to 0–47C RPS-1 or RPS-2 (Table 1). The kidney was then transferred with its arterial, venous, and ureter cannulae to a beaker surrounded by ice and containing ice-cold RPS-1 or RPS-2. The kidney was allowed to continue perfusing for another 5–10 min in the beaker. The kidney was weighed immediately after perfusion, then placed back in cold RPS-1 or RPS-2. Preparation of Perfusor Rabbits Kidneys were evaluated using a modification of the ex vivo blood reperfusion technique of Carter et al. (3). Anesthesia, endotracheal intubation, laparotomy, and intestinal handling were performed on the perfusor rabbit as described above, but the renal vessels were ligated bilaterally, and 2 ml of 1% acetylsalicylic acid (aspirin, ASA) and 2000 units of heparin were infused intravenously.

08-22-97 06:40:29

cryoas

AP: CRYO

117

FUNCTION OF VITRIFIABLE KIDNEYS

When necessary, blood was collected from a third rabbit and administered to the perfusor rabbit via the venous line in order to raise the perfusor rabbit’s blood pressure (3). Transfusion reactions were not observed in this highly inbred population. Dopamine tended to produce visible renal vasoconstriction when used to support blood pressure, so 5 mg/dl dobutamine HCl (Dobutrex, Eli Lilly & Co.), delivered through the venous return line (Fig. 1), was used when necessary. Ex Vivo Perfusion Figure 1 depicts the ex vivo tubing circuit. The aorta was cannulated using a T-110 shunt tip connected to an arterial perfusion line (A) composed of thin-walled silastic tubing (equivalent to T5715-11 tubing; Baxter Scientific Products, McGaw Park, IL), and the vena cava was cannulated in similar fashion (V). Channels through the wall of the heavy silastic tubing segment LA formed nonslip contacts with the arterial and venous lines near the kidney, permitting essential adjustments of the arterial and venous cannulae positions to compensate for variations in vessel length or tubing length (3). Arterial blood pressure was monitored with a Tycos sphygmomanometer gauge (SG) protected by a Trimed diaphragm (TD) (equivalent to a ‘‘Transducer Hepatitis Protector,’’ Kendall McGaw Laboratories, Sabana Grande, Puerto Rico). All tubing was primed with 0.9% (w/v) NaCl just prior to use to minimize air bubble nucleation on the inner tubing walls. The Carter et al. apparatus was modified to include a sidearm in the venous line for infusion of pressor and to include arterial sidearms for blood collection and air bubble evacuation. The venous pressure measurement standpipe was replaced by a syringe of heparinized saline (SR) to clear the venous sidearm of blood between flow measurements. It was difficult to connect the renal shunt tip cannulae directly to the blood lines without introducing air, so we provided silastic tubing cannula extensions

AID

CRYO 2026

/

a90a$$$161

(CE) which could be plugged into the main blood lines much more easily. The donor kidney was positioned on a gauze-covered platform near the ends of the ex vivo tubing assembly, and the CEs were connected to the clamped perfusor lines. The venous clamp was removed, the arterial line was digitally occluded, the arterial clamp was released, and the digital occlusion was slowly released to allow gradual blood reflow to the kidney and hence a gentle rather than abrupt increase in perfusion pressure. Blood flow was measured by diverting the venous return from the kidney into a 5-ml syringe barrel (BF in Fig. 1) and measuring the time required to collect 2 ml of blood using a stopwatch. After an initial 15-min equilibration period, urine was collected from 15 to 30, 30 to 60, 60 to 90, and 90 to 120 min. Blood leakage into the urine could usually be minimized by advancing the ureter cannula tip closer to the hilum (presumably thus avoiding damaged blood vessels on the ureter lining) and gently religating the cannula near its tip. Hilar bleeding was remedied by the liberal application of fibrin powder (Sigma Chemical Co., St. Louis, MO), and unnecessary stripping of hilar fat was avoided to minimize this problem. Renal surface drying was prevented by draping the kidney with a stretched square of Parafilm. The ex vivo perfusion lines were recycled three to five times. They were cleaned using hot tap water and sonication, rinsed with distilled water, and stored in a refrigerator. Siliconizing the apparatus to avoid clotting (3) was not necessary. Cryoprotectant Perfusion Equipment and Data Handling The perfusion apparatus has been described in detail elsewhere (9, 10). The device uses a computer coupled to pumps, solenoids, and sensors through a standard interface. Linear concentration changes are carried out using computer-assisted gradient formers, and step changes in concentration are effected using

08-22-97 06:40:29

cryoas

AP: CRYO

118

FAHY AND ALI

solenoids to switch from one reservoir to another. The arteriovenous (A-V) concentration difference across the kidney was determined by continuously perfusing the reference cell of a differential refractometer with arterial perfusate and the sample cell with venous effluent. The differential signal was corrected for a baseline (no A-V concentration difference) shift that was dependent upon the absolute concentration of cryoprotectant (CPA). The corrected signal was calibrated using fixed concentration differences. The kidney was assumed to be in osmotic equilibrium with the perfusate (though not necessarily in equilibrium with respect to cell volume) when the difference signal between arterial and venous samples approximated zero. This assumption was subsequently verified by direct determinations of tissue CPA concentration and by DSC determinations of tissue vitrifiability and glass transition temperature (unpublished observations). Perfusion data were assigned post hoc to standardized 0.6- to 1.5-min time blocks to permit experiments to be averaged and the averaged data to be displayed. All experiments are represented at each plotted data point. Data were tested for statistical significance using Student’s unpaired t test or the Mann– Whitney rank sum test, depending upon the outcome of assumption testing using SigmaStat, a scientific statistical software package (Jandel Scientific, San Rafael, CA). Solutions The vitrification solution was VS4 (12, 14) (Table 1). In VS4, the acetamide of VS1 (26) is replaced by formamide to speed cryoprotectant permeation (15) and avoid carcinogenicity (27), and 1,2-propanediol is increased to compensate for the poor vitrifiability of formamide (8). High-molecular mass species (PEG) are avoided to control viscosity and thus enhance perfusability. VS4 contains a total of 7.5 molar (490 g/liter) CPA, has a melting point of about 0337C, and vitrifies at 1000 atmospheres of hydrostatic pressure (12, 24).

AID

CRYO 2026

/

a90a$$$161

The vehicle for VS4 was RPS-1 (Table 1). To avoid precipitation in the presence of VS4, the concentration of Ca2/ in RPS-1 was lowered from 1 to 0.05 mM as VS4 was introduced and raised back to 1 mM as VS4 was removed; similarly, the Mg2/ concentration was lowered to 0.4 mM as VS4 was introduced. In two successful and three unsuccessful VS4 perfusions, RPS-1-M (Table 1) was used in the final reservoir. The first six control kidneys were perfused with RPS-2 rather than with RPS-1, but the results were similar and were therefore pooled. Addition and Removal of VS4 CPA was introduced and removed in seven phases according to reasoning described earlier (5, 17). Phase 1 is an initial equilibration phase; phase 2 is a gradual introduction of about half of the final target concentration; phase 3 is a pause in introduction designed to permit osmotic equilibration prior to phase 4, which is a step change to and a holding period at the final target concentration of 7.5 M; phase 5 is a step change downward from the peak concentration and an equilibration period at the reduced concentration; phase 6 is a gradual washout of the remaining cryoprotectant; and phase 7 is a final equilibration period prior to the normothermic ex vivo blood reperfusion test. Step changes in concentration were based on tissue slice experiments showing such steps to be essential for controlling CPA toxicity (5). To offset some of the osmotic stress associated with CPA washout, mannitol was incorporated into the solutions used in phases 5, 6, and 7. In all but one experiment, 300 mM mannitol was present during phase 5. In two successful and two unsuccessful experiments, this concentration was maintained during phase 6, then stepped down to 100 mM at phase 7, and then reduced gradually to 0 mM over 30 min. In the remaining nine experiments (four successful, five unsuccessful), the mannitol concentration was linearly reduced

08-22-97 06:40:29

cryoas

AP: CRYO

119

FUNCTION OF VITRIFIABLE KIDNEYS

to 100 mM during phase 6 as CPA levels fell, then stepped down to 50 mM after all CPA had been removed, and then linearly reduced to 0 mM over 30 min. Functional Measurements Renal functional indices were calculated in the standard way. Renal blood flow (RBF) is reported as the directly measured venous outflow, since the difference between arterial and venous blood flow rates (arterial flow Å venous flow / urine flow) was typically only 1–5% of the blood flow rate. Glomerular filtration rate (GFR) is expressed as ml/hr as per Carter et al. (3). Filtration fraction was estimated as GFR/RBF. Tissues were fixed in 10% formalin or in Karnovsky’s fixative for routine histological evaluation. Experimental Design There were five experimental groups. After perfusion in vitro, kidneys in all groups were evaluated by ex vivo blood perfusion. Control (CON) group. This group consisted of kidneys perfused with RPS-1 or RPS-2 for approximately 4.5 h. In this group, the calcium and magnesium concentrations were lowered to 50 mM and to 1 mM, respectively, at t Å 15 min, and then restored to normal 45 min before the end of the perfusion, simulating ion changes during the introduction and removal of cryoprotectant in other perfusions. Low concentration (25% CPA) group. This group consisted of kidneys perfused with 25% (w/v) CPA (Table 1) (equivalent to 51% of full strength VS4, or ‘‘51% VS4’’). The total time of exposure to CPA was the same as in the VS4 groups. Functioning high concentration (49% CPA) group (VS4 group). A total of 13 kidneys were perfused with full-strength VS4 without disqualifying technical failures. After these experiments were completed, it was noted that the results fell rather cleanly into two categories, kidneys that functioned and kidneys that did not. The ‘‘VS4 group’’ consisted of the

AID

CRYO 2026

/

a90a$$$161

FIG. 2. Pressure, flow, and temperature profiles of control kidneys (Con, solid lines; CNIS, dotted lines) in vitro. Means { SEM.

kidneys that recovered substantial immediate function. Nonfunctioning high concentration group (NF group). These kidneys were perfused with VS4 by methods similar to those used in the VS4 group, yet failed to develop significant function. They were analyzed separately from the VS4 group to gain insight into the causes of success vs failure after exposure to VS4 and to permit this group to be differentiated from the VS4 group. Control ischemic (CNIS) group. These kidneys were not perfused with CPA but were subjected to reduced blood flow during ex vivo blood perfusion by applying a partially occluding clamp to the arterial perfusion line to partially simulate the temporary deficit in blood flow that was observed to occur in the functioning VS4 subgroup. RESULTS

Control Baseline Figure 2 shows the in vitro perfusion dynamics of kidneys perfused without cryopro-

08-22-97 06:40:29

cryoas

AP: CRYO

120

FAHY AND ALI

et al. for fresh kidneys (3) and those reported by Hawkins et al. for kidneys formerly perfused in vitro with RPS-2 (19). 25% CPA Perfusion Protocol and Outcome Figure 4 shows the protocol used for the 25% (w/v) CPA group. After the 25% CPA (3.8 M) plateau, the perfusate was switched to 25% CPA plus 300 mM mannitol (near 140 min). The time spent at these steps was equal to the time spent during analogous stages of the introduction and removal of VS4 (see below). The A-V concentration differential (middle) appeared to respond appropriately and sensitively to changing conditions. The bottom panel depicts the mean arterial (T.a) and venous (T.v) temperatures, which were similar and remained between 0 and 27C. The ex vivo functional indices for this group showed no lasting injury (see below), indicatFIG. 3. Physiological performance of the control (Con) group. FF, filtration fraction, in dimensionless units. ClPr , protein clearance, in ml/h or as a percentage of GFR. % Reab., fractional reabsorption. Means { SEM.

tectant (CON and CNIS groups). The flow rates obtained were consistent with previous observations (22), particularly considering the generally lower temperatures in the present experiments and use of the flushed, undrained kidney vs the fully drained kidney for the reference kidney weight. Changing Ca2/ and Mg2/ concentrations during the perfusion had no visible effect on flow rate. The CON and CNIS groups perfused similarly and experienced similar conditions. The functional indices for the control kidneys are given in Fig. 3 and appear satisfactory. The GFR was in agreement with the results of Carter et al. (3) for fresh kidneys ex vivo. Fractional volume, sodium, glucose, and potassium reabsorption was satisfactory, and protein exclusion from the urine was virtually total (ú99%). Our ex vivo blood flows were intermediate between those reported by Carter

AID

CRYO 2026

/

a90a$$$161

FIG. 4. Protocol and perfusion dynamics of kidneys perfused with ‘‘half-strength’’ VS4 (25% (w/v) cryoprotectant). P/10 Å 0.1 1 perfusion pressure (in mmHg); M, ˆ , programmed molarity of measured molarity of CPA; M CPA; F, flow (in [ml/min]/g); A-V (mM), the arteriovenous CPA concentration difference; T.a, arterial temperature; T.v, venous temperature. Vertical lines by each parameter abbreviation identify the line type used to represent the parameter. The upward jump in apparent concentration (top) and in the A-V difference at about 140 min reflects the introduction of mannitol, whose effect on refractive index was not fully compensated for in the calibration. n Å 3, means { SD.

08-22-97 06:40:29

cryoas

AP: CRYO

121

FUNCTION OF VITRIFIABLE KIDNEYS

FIG. 5. In vitro perfusion data for the first ‘‘successful’’ perfusion of an organ with a vitrifiable concentration of CPA. Parameter abbreviations as in the legend to Fig. 4. The seven phases of this experiment are indicated by numerals by each phase. The initial equilibration time (phase 1) was 15 min; the CPA addition rate (during phase 2), 47.2 mM/min; first plateau (phase 3) duration, 18.46 min; duration of VS4 perfusion (phase 4), 15.93 min; second plateau duration (phase 5), 17.11 min; and CPA washout rate (during phase 6), 034.4 mM/min. After CPA washout (phase 7), the mannitol used as an osmotic buffering agent was gradually washed out over 30 min (data not depicted). The figure is a replotted facsimile of the computer screen at the end of the experiment (49DFP4) on November 21, 1986. In this experiment, mannitol was present at 300 mM throughout phase 5 and 6 and was dropped to 100 mM upon beginning phase 7.

ing that perfusion with approximately 4 M VS4 solutes is not particularly harmful. VS4 Perfusion Protocol and Overall Outcome Figure 5 shows the first well-controlled and well-documented attempt to introduce and remove VS4. The numbers 1–7 in the top panel represent the seven phases of the CPA treatment protocol described above. The middle and bottom panels are as in Fig. 4. This was the first protocol that yielded a ‘‘surviving’’ kidney after exposure to a vitrifiable solution. During VS4 perfusion (phase

AID

CRYO 2026

/

a90a$$$161

4), the A-V concentration gradient came to 50 mM or less, which is equivalent to about 0.3% (w/v), or 3 times less than the resolution within which the concentration needed to vitrify (Cv) was originally determined (8). A photograph of this kidney after about 60 min of blood reflow (Fig. 6) shows visually normal cortical blood flow, no hemorrhaging or other cortical abnormalities, and virtually bloodfree urine. In all, 13 VS4 perfusions were accomplished without technical failures. It was noticed that kidneys perfused with VS4 appeared to fall into two distinct categories: kidneys that produced appreciable amounts of urine early in the evaluation period and kidneys that did not. Further analysis confirmed that kidneys producing little urine were impaired in other respects as well (see below) in comparison to the remaining VS4-perfused kidneys. In view of this apparent bimodal response to VS4 exposure, it appeared necessary to analyze the functioning and the nonfunctioning kidneys as separate groups. Of the 13 perfusions, 6 yielded functioning kidneys, and 7 resulted in nonfunctioning kidneys. We were unable to account for success or failure on the basis of variations in the VS4 perfusion protocol (Table 2) or other perfusion conditions (Fig. 7). On the other hand, some nonfunctioning kidneys did not attain rapid and/or complete blood washout during procurement, and this could have affected outcome. One failed kidney had a phase 5 mannitol concentration of 200 mM vs the standard 300 mM. Renal Blood Flow, Vascular Resistance, and Renin Release Blood reflow histories and renovascular resistance results for all groups are documented in Fig. 8. Blood flow (Fig. 8, top) became profoundly depressed in all VS4-perfused kidneys shortly after revascularization, but recovered to near normal. The low flow (‘‘vascular crisis’’) period was considerably longer in the nonfunctioning VS4 (NF) group than in the

08-22-97 06:40:29

cryoas

AP: CRYO

122

FAHY AND ALI

FIG. 6. The kidney whose perfusion data are given in Fig. 5. This was the first rabbit kidney to ‘‘survive’’ after being perfused with VS4 in vitro and then with rabbit blood ex vivo. Renal color is essentially normal and uniform over the entire renal surface, and the urine is free of blood, is significant in volume, and is of a normal color. The kidney rests on a gauze pad supported by an inverted histology tub. The parafilm cover normally resting on the kidney to prevent surface dehydration has been removed for the photograph. This photograph was taken by Dr. Robert J. Williams after approximately 60 min of blood reflow.

TABLE 2 Protocol Comparison for Functioning and Nonfunctioning Kidneys Functional group: Mean { SEM (range)

Parameter dC/dt (loading) [CPA] at phase 3 Time at phase 3 Time at VS4 [CPA] at phase 5 Time at phase 5 dC/dt (washout)

48.8 3.67 12.8 20.2 4.31 15.0 043.3

{ { { { { { {

0.51 mM/min (47.2–50.7) 0.13 M (3.0–3.87) 1.9 min (9.2–18.5) 1.8 min (14.6–25.1) 0.18 M (3.8–4.6) 2.4 min (8.9–21.7) 4.0 mM/min (033.4 to 053.8)

Non-functional: Mean { SEM (range) 50.8 3.85 8.26 22.9 4.35 11.3 047.7

{ { { { { { {

1.00 mM/min (48.4–56.5) 0.02 M (3.8–3.93) 1.3 min (5.2–12.7) 1.0 min (19.1–26.4) 0.16 M (3.8–4.6) 3.0 min (2.63–23.10) 3.6 mM/min (033.6 to 055.0)

Note. dC/dt, rate of change of cryoprotectant concentration; [CPA], total concentration of cryoprotectant.

AID

CRYO 2026

/

a90a$$2026

08-22-97 06:40:29

cryoas

AP: CRYO

123

FUNCTION OF VITRIFIABLE KIDNEYS

functioning VS4 group, and the vascular resistance was significantly higher in NF vs VS4 kidneys as well (Fig. 8, bottom), but even the VS4 group did not recover fully normal resistance (asterisks at 60–120 min, Fig. 8, bottom), even though VS4 flow was not statistically different from control flow at any time after 30 min. Kidneys perfused with 25% (w/ v) VS4 solutes showed a similar but far less profound early perfusion defect that reversed itself much faster and then changed into hyperemia (P õ 0.05 at 120 min). Hyperemia in the CNIS group did not attain statistical significance at any given time point. Blood renin levels were measured to determine whether elevated resistance and reduced blood flow could be related to renin release (Fig. 9A). Although renin was elevated in the NF group at 22.5 min in comparison to controls (P Å 0.022), the absolute renin levels present in this group at that time were not significantly different from control renin levels at 75–105 min, and renin levels were not elevated at all in the VS4 group despite severe flow reduction. Although low blood flow at 22.5 min might have minimized release of re-

FIG. 7. Perfusion parameters and the extent of cryoprotectant osmotic equilibration in kidneys perfused with VS4. Solid lines, kidneys that functioned after perfusion with VS4; dotted lines, kidneys that did not function. Means { SEM. Arterial and venous temperatures did not differ from the mean temperature ([arterial / venous]/2) by more than 0.757C at most.

AID

CRYO 2026

/

a90a$$$161

FIG. 8. (Top) Blood reflow in kidneys previously perfused with vehicle alone [Con; control ischemic (CNIS)], 25% (w/v) CPA (25%), or 49% (w/v) CPA [VS4; nonfunctioning VS4 (NF)]. All data (including the control data itself) are normalized to the control group for clarity of representation. Means { SEM. The asterisk on the NF curve at 30 min indicates a significant difference between the VS4 and NF groups (P Å 0.022, Mann–Whitney rank sum test); the asterisk on the VS4 curve indicates a significant difference (P Å 0.043) between the VS4 group and the controls (t test). Asterisk on 25% curve indicates P õ 0.05 compared to controls. (Bottom) Vascular resistance (mm Hg/[(ml/min)/g]). Asterisks indicate significant (P õ 0.05) differences between the VS4 group and the Con group (* by VS4 points) and between the NF group and the VS4 group (* by NF points). Means { SEM.

nin to the general circulation, no rise in renin levels was seen when blood flow to the NF kidneys recovered. Finally, an attempt to improve renal blood flow using capotan (an ACE inhibitor) failed (data not shown). Interestingly, blood pressure was elevated substantially in all experimental groups, but particularly in the NF group (Fig. 9B). Gener-

08-22-97 06:40:29

cryoas

AP: CRYO

124

FAHY AND ALI

Glomerular Function

FIG. 9. Plasma renin activity (A) and mean arterial blood pressure (B). Means { SEM. Abbreviations as in the legend to Fig. 8.

ally, blood pressure could not be related to circulating renin levels. Comparison of Fig. 8 and Fig. 9B shows that the flow deficits of the VS4 and NF groups were intrinsic to the kidney and were not the result of reduced blood pressure. Hyperemia in the 25% CPA group was due to both lower renal resistance and elevated mean arterial pressure.

The recovery of GFR (Fig. 11A) correlated with the magnitude of blood reflow within the first 30 min. The GFR of 25% CPA kidneys, which suffered little restriction in blood flow, recovered fully within 45 min. The filtration rates of the CNIS kidneys were identical to those of the VS4 group at all times, indicating that even brief flow restriction of severe magnitude was sufficient to simulate the effect of VS4 on filtration rate. The NF group, whose blood flow was reduced the most, recovered very little filtration. Figure 11B shows filtration fraction (FF) as a percent of control FF. The results parallel the GFR data, but with lower resolution. The full recovery of FF but only partial recovery of GFR in the VS4 group suggests that GFR at 105 min was limited by blood flow rather than by FF in this group. As FF improved, so too did apparent protein retention by the glomeruli in the VS4 group (Fig. 11C), suggesting an improvement in both quantity and quality of the filtrate over time. The lack of effect of 25% CPA or of transient hypoxia (CNIS group) implies that the transient effect of VS4 on apparent protein retention is a direct chemical or osmotic effect of VS4 rather than an effect of hypoxia. In the

Urine Flow Rates Although urine flow rates indeed separated the VS4 group from the NF group (Fig. 10), it was not possible in general to judge renal function on the basis of urine output alone. For example, the undamaged 25% CPA group and the highly damaged NF group produced equal urine outputs at all times, and the NF urine output was synonymous with that of controls by 75–105 min of reflow (Fig. 10).

AID

CRYO 2026

/

a90a$$$161

FIG. 10. Urine flow in the experimental (solid lines) vs the control (dotted line) groups. Abbreviations as in the legends to Figs. 8 and 9. Means { SEM.

08-22-97 06:40:29

cryoas

AP: CRYO

125

FUNCTION OF VITRIFIABLE KIDNEYS

FIG. 11. Glomerular function in all experimental groups. (A) GFR. (B) Filtration fraction. (C) Apparent protein retention (100%-[ClPr/GFR] 1 100%). (D) Absolute protein clearance in ml/h. Data in A, B, and C normalized to controls. Group nomenclature as in the legends to Figs. 8–10. Means { SEM.

NF group, glomerular semipermeability was apparently lost. Despite the dramatic difference between protein retention in the VS4 and NF groups, the absolute protein excretion in the VS4 group resembled that of the NF group for the first 75 min of reflow (Fig. 11D), reflecting the lower urine flow rates in the NF group and a protein concentration in the NF filtrate that approached or equalled the protein concentration in plasma. Although protein retention approached 100% in the VS4 group by 105 min, absolute protein excretion was still very much above control limits, but appeared to be improving. The fact that GFR and FF were identical for the CNIS and VS4 groups whereas protein retention and protein excretion rates for these groups were quite different indicates

AID

CRYO 2026

/

a90a$$$161

that filtration rate and filtrate quality vary independently in this model. Tubular Function Figure 12 presents the fractional reabsorptions of sodium (Fig. 12A), potassium (Fig. 12B), and glucose (Fig. 12C) as well as water clearance information (urine osmolality, Fig. 12D). Twenty-five percent CPA actually raised sodium reabsorption by a small but consistent amount over controls (Fig. 12A), but sodium reabsorption was partially reduced in the VS4 group and was not significantly different from zero in the NF group. Unlike GFR, reduced Na/ reabsorption in the VS4 group could not be accounted for by the hypoxia of this group (compare to the CNIS results).

08-22-97 06:40:29

cryoas

AP: CRYO

126

FAHY AND ALI

FIG. 12. Tubular function in all experimental groups. Fractional sodium (A) and glucose (C) reabsorptions were normalized to control data. Abbreviations as in prior figures. Means { SEM. The increase in urine osmolality in the Con group was significant (P õ .02 for a linear regression). The urine osmolality at 75– 105 min in the 25% CPA group was significantly higher than the baseline (22.5 min) control urine osmolality (P õ 0.05).

As with Na/ transport, kidneys in the Con, CNIS, and 25% CPA groups reabsorbed significant and similar fractions of filtered potassium (Fig. 12B, top). VS4treated kidneys (Fig. 12B, bottom) excreted more than the entire filtered load of potassium, indicating either potassium secretion, passive loss of potassium from renal parenchyma to the urine, or both. The VS4 data on glucose reabsorption (Fig. 12C) were superimposable on the results for the CNIS group, suggesting that glucose transport in the VS4 group was affected primarily by the hypoperfusion experienced by this group. The reabsorption of glucose among the NF kidneys was not statistically different from

AID

CRYO 2026

/

a90a$$$161

zero and, if anything, the NF group tended to actually secrete glucose rather than reabsorb it. Twenty-five percent CPA permitted full recovery of glucose reabsorption. Con and 25% CPA kidneys were able to concentrate urine (Fig. 12D). Kidneys in the VS4 and NF groups were not, and this was accounted for by exposure to hypoxia (CNIS data). Circulating Metabolite and Enzyme Concentrations Figure 13 describes the circulating levels of most metabolites and enzymes that might be affected by altered renal clearances or renal injury. SGOT and SGPT were significantly

08-22-97 06:40:29

cryoas

AP: CRYO

127

FUNCTION OF VITRIFIABLE KIDNEYS

FIG. 13. Plasma metabolite levels. Like symbols refer to like groups in different panels. Con group shown as open circles in most panels for easy reference. At 105 min, serum glutamic-oxaloacetic transaminase (SGOT) was elevated over controls in both the NF (P õ 0.003, Mann-Whitney rank sum test) and the VS4 (P Å 0.034, t test) groups (P õ 0.02 for VS4 vs NF groups as well), and serum glutamic–pyruvic transaminase (SGPT) was also elevated in these groups (P Å 0.003 for NF vs Con, Mann–Whitney rank sum test, and P Å 0.001 for VS4 vs Con, by t test).

elevated at 105 min in both the NF and the VS4 groups, whereas the more kidney-specific GGT and the more nonspecific LDH showed no meaningful trends. Groups experiencing significant ischemia (CNIS, VS4, and NF) tended to have lower circulating total protein

AID

CRYO 2026

/

a90a$$$161

and calcium concentrations (open points), and recipients of kidneys perfused with 25% (w/ v) CPA tended to have higher circulating levels of phosphate and potassium. Urine pH was alkaline (usually between 7.7 and 8.1) in all groups (data not shown). Only

08-22-97 06:40:29

cryoas

AP: CRYO

128

FAHY AND ALI

kidneys in the CON group were able to lower urine pH significantly (from 8.00 { 0.09 at 75 min to 7.45 { 0.15 at 105 min). Figure 14 presents representative histological views of control (Fig. 14A), CNIS (Fig. 14B), VS4 (Fig. 14C), and NF (Fig. 14D) kidneys. Control kidney histology was normal. CNIS kidneys were noticably more damaged and resembled the ‘‘viable’’ VS4 kidneys. Changes in the ‘‘viable’’ VS4 group (not all represented in Fig. 14) included tubular cell swelling and vacuolation, apical membrane blebbing and distortion, tubular obstruction, the appearance of hyaline material in tubular lumena, and faintly stainable material in Bowman’s capsule despite reasonable preservation of glomerular structure. Tubular cell nuclei appeared reasonably well preserved. The injury observed in the NF group appeared to represent a significant intensification of the more mild changes seen in the VS4 and CNIS groups rather than a qualitatively different pattern of injury. DISCUSSION

The current results document the first evidence that a sensitive mammalian organ can tolerate perfusion with a vitrifiable concentration of cryoprotectant without losing function. Slice experiments had previously indicated that renal tissue could endure exposure to vitrification solutions (16, 17), and muscle systems not dependent on blood reflow for functional evaluation had previously been found to withstand vitrifiable concentrations of dimethyl sulfoxide (4, 18), but until the present experiments it was not known whether the vascular system could also maintain its integrity following such exposure. Previous investigations involving rabbit and dog kidneys were limited to concentrations no higher than 4 M (20) to 4.2 M (21) CPA, respectively,

which are insufficient for vitrification. In the experiments described here, substantial immediate function in the form of tubular (sodium and glucose reabsorption), glomerular (GFR, FF, and protein exclusion from the urine) and vascular (RBF) measures was documented. Not all kidneys were able to regain function, and those that did regain function did not regain full function. On the other hand, the impairments in the functioning kidneys were comparable to those induced by a mild ischemic stress that is known to be consistent with subsequent renal recovery and life support (2). Furthermore, human kidney recipients commonly require three or more dialysis treatments prior to resumption of adequate renal function (25), whereas kidneys in the functioning VS4 group showed appreciable immediate function. Although the current results show that the vascular system can withstand exposure to VS4 in an intact organ sufficiently well to permit some organs to function, impairment of vascular function did occur and appeared to play a direct mechanistic role in the overall injury observed. The sequence of events suggested by Figures 8, 9, 11, and 12 is that VS4 exposure results in a transient state of reduced blood flow upon restitution of circulation and that this reduced blood flow either results in outright organ failure or, when it is milder, accounts for the observed impairment of filtration rate and of glucose and water reabsorption. Increased glomerular permeability and reduced sodium and potassium reabsorption after VS4 exposure appear to be direct osmotic or chemical effects of VS4 rather than secondary effects of hypoxia. The basis of blood flow impairment in VS4-perfused kidneys is unknown. Increased protein excretion could result from either elevated glomerular permeability

FIG. 14. Histology after VS4 perfusion and subsequent ex vivo blood reperfusion for 2 h. (A) Control kidney, (B) CNIS (control ischemic) kidney, (C) Kidney from the functioning VS4 group, (D) kidney from the nonfunctioning VS4 group.

AID

CRYO 2026

/

a90a$$$161

08-22-97 06:40:29

cryoas

AP: CRYO

129

FUNCTION OF VITRIFIABLE KIDNEYS

AID

CRYO 2026

/

a90a$$2026

08-22-97 06:40:29

cryoas

AP: CRYO

130

FAHY AND ALI

to plasma protein or loss of protein from the renal parenchyma. Similarly, net potassium ‘‘secretion’’ in the VS4 and NF groups could reflect passive losses from injured renal cells. However, blood levels of potassium were not elevated in these groups, and blood protein content was reduced. Net glucose ‘‘secretion’’ by some NF kidneys could result from CPAstimulated gluconeogenesis (1). Surprisingly, SGOT and SGPT were the most meaningful circulating markers of renal injury observed. The mild hypocalcemia and hyperphosphatemia seen in some groups may not be abnormal (22). The natural history of the CPA-perfused kidney in vitro shows some interesting features. As concentration rises, renal flow rate initially increases, presumably due to osmotic vasodilation. With higher concentrations and higher viscosities, flow rate decreases and then recovers when concentration is lowered and viscosity falls. These features have been seen in all CPA perfusions carried out since this initial study and are very reproducible (Fig. 7). These earliest pilot results represent only a first step toward the goal of introducing and removing vitrifiable media without loss of viability. Further optimization and the results of permanent orthotopic transplantation of VS4perfused kidneys are described in the next paper of the current series (23). ACKNOWLEDGMENTS Naval Medical Research and Development Command, Work Unit No. 61153 N. MRO 4120.001-1462 NMR& D. The opinions and assertions contained herein are those of the authors and are not to be construed as official nor as representing those of the Department of Defense or of the Navy. We thank Daniel I. Levy, Brian Ackerly, Marianne Martin, and Martin Kamya for expert technical assistance. We thank Dr. Robert J. Williams for taking the photograph that appears as Fig. 5 and Dr. Bijan S. Khirabadi for preparing the micrographs that appear as Fig. 14. This research was supported in part by grants from the National Institute of General Medical Science (GM 17959 and BSRG 2 S07 RR05737). We gratefully acknowledge the critical support provided by Organ, Inc., LRT, Inc., the G. Harold and Leila Y. Mathers Charitable Foundation, the Office of Naval Research, the American

AID

CRYO 2026

/

a90a$$$161

Red Cross, and the Life Extension Foundation, all of which was essential for completion and publication of this work. This research was undertaken under the auspices of the American Red Cross, initially at the Blood Research Laboratory in Bethesda, MD, and continuing at the Jerome Holland Laboratory’s Transplantation Laboratory in Rockville, MD. We thank the American Red Cross for its critical role in housing and supporting this work. Animal Disclaimer: The experiments reported herein were conducted according to the principles set forth in the ‘‘Guide for the Care and Use of Laboratory Animals,’’ Institute of Laboratory Animal Resources, National Research Council, DHHS, Publication No. (NIH) 86-23 (1985). REFERENCES 1. Baxter, S. J., and Lathe, G. H. Biochemical effects on kidney of exposure to high concentrations of dimethyl sulphoxide. Biochem. Pharmacol. 30, 1079–1091 (1971). 2. Booster, M. H., van der Vusse, G. J., Wijnen, R. M., Yin, M., Stubenitsky, B. M., and Kootstra, G. University of Wisconsin solution is superior to histidine tryptophan ketoglutarate for preservation of ischemically damaged kidneys. Transplantation 58, 979–984 (1994). 3. Carter, J. N., Green, R. D., Halasz, N. A., and Collins, G. M. Ex vivo perfusion. A renal preservation model. J. Surg. Res. 31, 55 (1981). 4. Elford, B. C., and Walter, C. A. Effects of electrolyte composition and pH on the structure and function of smooth muscle cooled to 0797C in unfrozen media. Cryobiology 9, 82–100 (1972). 5. Fahy, G. M. Cryoprotectant toxicity: Biochemical or osmotic? Cryo-Letters 5, 79–90 (1984). 6. Fahy, G. M. Vitrification: A new approach to organ cryopreservation. In ‘‘Transplantation: Approaches to Graft Rejection’’ (H. T. Meryman, Ed.), pp. 305–335. A. R. Liss, New York, 1986. 7. Fahy, G. M. Vitrification of multicellular systems and whole organs. Cryobiology 24, 580–581 (1987). 8. Fahy, G. M. Vitrification. In ‘‘Low Temperature Biotechnology: Emerging Applications and Engineering Contributions’’ (J. J. McGrath and K. R. Diller, Eds.), pp. 113–146. ASME, New York, 1988. 9. Fahy, G. M. Organ perfusion equipment for the introduction and removal of cryoprotectants. Biomed. Instr. Technol. 28, 87–100 (1994). 10. Fahy, G. M. ‘‘Computer Controlled Cryoprotectant Perfusion Apparatus,’’ United States Patent 5,472,876 (1995). 11. Fahy, G. M., Ali, S. E., and Levy, D. I. Physiology of rabbit kidneys after perfusion with VS4. Cryobiology 31, 573 (1994). 12. Fahy, G. M., da Mouta, C., Tsonev, L., Khirabadi, B. S., Mehl, P., and Meryman, H. T. Cellular in-

08-22-97 06:40:29

cryoas

AP: CRYO

131

FUNCTION OF VITRIFIABLE KIDNEYS

13.

14.

15.

16.

17.

18.

19.

jury associated with organ cryopreservation: Chemical toxicity and cooling injury. In ‘‘Cell Biology of Trauma’’ (J. J. Lemasters and C. Oliver, Eds.). CRC Press, Boca Raton, 1994. Fahy, G. M., and Hirsh, A. Prospects for organ preservation by vitrification. In ‘‘Organ Preservation, Basic and Applied Aspects’’ (D. E. Pegg, I. A. Jacobsen, and N. A. Halasz, Eds.), pp. 399–404. MTP Press, 1982. Fahy, G. M., Khirabadi, B. S., and Mehl, P. Equipment, solutions, perfusion techniques, and medications permitting survival of kidneys perfused with vitrifiable media. Cryobiology 28, 512 (1991). Fahy, G. M., Levy, D. I., and Ali, S. E. Some emerging principles underlying the physical properties, biological actions, and utility of vitrification solutions. Cryobiology 24, 196–213 (1987). Fahy, G. M., Lilley, T. H., Linsdell, H., Douglas, M. St. J., and Meryman, H. T. Cryoprotectant toxicity and cryoprotectant toxicity reduction: In search of molecular mechanisms. Cryobiology 27, 247–268 (1990). Fahy, G. M., MacFarlane, D. R., Angell, C. A., and Meryman, H. T. Vitrification as an approach to cryopreservation. Cryobiology 21, 407–426 (1984). Farrant, J. Mechanism of cell damage during freezing and thawing and its prevention. Nature 205, 1284– 1287 (1965). Hawkins, H. E., Clark, P., Lippert, E. C., and Karow, A. M., Jr. Functional preservation of the mammalian kidney. VI. Viability assessment of rabbit kidneys perfused at 257C. J. Surg. Res. 38, 281–288 (1985).

AID

CRYO 2026

/

a90a$$$161

20. Jacobsen, I. A., Pegg, D. E., Wusteman, M. C., and Robinson, S. M. Transplantation of rabbit kidneys perfused with glycerol solutions at 107C. Cryobiology 15, 18–26 (1978). 21. Karow, A. M., Jr., McDonald, M., Dendle, T., and Rao, R. Functional preservation of the mammalian kidney. VII. Autologous transplantation of dog kidneys after treatment with dimethyl sulfoxide (2.8 and 4.2 M). Transplantation 41, 669–674 (1986). 22. Khirabadi, B. S., and Fahy, G. M. Cryopreservation of the mammalian kidney. I. Transplantation of rabbit kidneys perfused with EC and RPS-2 at 2– 47C. Cryobiology 30, 10–25 (1994). 23. Khirabadi, B. S., Fahy, G. M., and Ali, S. E. Cryopreservation of the mammalian kidney. III. Demonstration of life support function after introduction and removal of 7.5 molar cryoprotectant. In preparation. 24. Mehl, P. Nucleation and crystal growth in a vitrification solution tested for organ cryopreservation by vitrification. Cryobiology 30, 509–518 (1993). 25. Merion, R. M., Oh, H. K., Port, F. K., Toledo-Pereyra, L. H., and Turcotte, J. G. A prospective controlled trial of cold-storage versus machine-perfusion preservation in cadaveric renal transplantation. Transplantation 50, 230–233 (1990). 26. Rall, W. F., and Fahy, G. M. Ice-free cryopreservation of mouse embryos at 01967C by vitrification. Nature 313, 573–575 (1985). 27. Weisburger, J. H., Yamamoto, R. S., Glass, R. M., and Frankel, H. H. Prevention by arginine glutamate of the carcinogenicity of acetamide in rats. Toxicol. Pharmacol. 14, 163–175 (1969).

08-22-97 06:40:29

cryoas

AP: CRYO