The Time of Origin of the Parabiotic Anastomosis' MARVIN SODICOFF AND ROBERT T. BINHAMMER DepaTtment of Anatomy, University of Cincinnati, College of Medicine, Cincinnati, Ohio ABSTRACT The course of development of parabiotic anastomosis was followed with respect to time in Holtzman ( Sprague-Dawley) parabionts. Comparison of nonirradiated pairs was made to pairs in which one of the partners had received 800r. The development of the anastomosis was studied by determining the per cent transfer of CR51 labeled erythrocytes (which represented a cellular element) and radio-iodinated serum albumin (which represented a molecular element) from one partner to the other. Erythrocytes were found in the non-injected animal in small numbers at 22 hours after pairing; at 47 hours the rate of transfer became much more rapid. RISA was detected in the non-injected partner as early as five hours after pairing and accumulated steadily thereafter. Irradiation had no effect on the course of development of the parabiotic anastomosis as evidenced by similarity of accumulation rates of Cr51 labeled erythrocytes and RISA when compared to the non-irradiated pairs. There is no exact information in the literature concerning the time of development of the parabiotic vascular anastomosis. The present study was designed to provide more precise information regarding the early phase in the development of the parabiotic union. That an actual vascular anastomosis does develop was suggested by the work of Hill ( ' 3 2 ) using a colorimetric determination of the plasma of parabionts after injection of brilliant vital red into the vascular system of one partner. Van Dyke et al. ('48) demonstrated vascular continuity between partners using Fe59tagged erythrocytes. Several investigators have indicated that the anastomosis becomes functional between two and four days after union Anderson ('54) demonstrated the presence of capillary connections as early as 3 to 4 days after pairing. Van Dyke et al. ('48) reported almost no exchange of FeS9labeled erythrocytes until 2 to 3 days after pairing. The present study' determines more precisely the time of origin of vascular anastomosis with respect to cellular elements, as represented by erythrocytes, and non-cellular elements, as represented by the protein serum albumin. It was also of interest to determine whether or not irradiation of one animal of a pair would alter the development of the anastomosis. MATERIALS AND METHODS The experimental animals were female rats of the Sprague-Dawley strain (Holtz- man) weighing 160-180 gm. Animals were maintained at a temperature of 24°C and fed a diet of Purina laboratory chow pellets and tap water ad libitum. Rats of similar weight were united surgically under pentobarbital anesthesia using the technique of Bunster and Meyer ('33). In those experiments in which one of the partners was irradiated, the animals were exposed two hours prior to pairing. Three or four unanesthetized rats were placed in a ventilated plastic box 15 X 15 X 15 cm for irradiation. Animals received a single 800 r dose directed to their dorsums which in this laboratory represents an LD 60/30 days. Radiation was administered with a 250 KVP machine using 250 KV, 15 ma, 0.5 mm Cu and 1.0 mm Al, HVL 1.7 mm Cu, STD 70 cm, at a dose rate of about 67 r/min. The machine was calibrated before each exposure using a 100 r Victoreen ionization chamber in a paraffin phantom. Cross-circulation was studied by following NaCP04 tagged red blood cells in some pairs and radioiodinated serum albumin (RISA) in different pairs. Erythrocytes from donor rats were tagged in vitro after separation by centrifugation. Approximately 8 wc NaCrS1O4/mlwhole blood was added to the erythrocytes and incubation was carried out at room temperature for 30-45 minutes using gentle agitation. At the end of the incubation period 1 ml of 1 This inyestigation was supported by a Public Health Service Fellowship (GPM 19,030) from Natlonal Institute of General Medical Sciences and by grant CA-03390from the National Canc Cancer Institute. 625 MARVIN SODICOFF A N D ROBERT T. BINHAMMER BINHAMMER 626 50 - 0 Eu 40- L e F! L ..- 2. 30- L c =g 20- ... c c r s O 10- - 0 0 10 20 1 30 I I I 40 60 50 Hours after pairing I I I i 70 80 90 100 Fig. 1 Erythrocyte and pIasma transfer at various sampling intervals after pairing. Vertical lines on the curves denote standard error. Label was injected at time of pairing. ascorbic acid (50 mg/ml) was added to the incubation mixture. Cells were washed twice with normal saline and resuspended in saline; 1.0 ml of this preparation was injected into the tail vein of one member of each pair immediately after pairing. In those pairs in which one animal had been irradiated the non-irradiated partner was injected. In those pairs in which RISA was used, 7-8 vc diluted in 0.5 ml saline was injected into the tail vein immediately after pairing. It was assumed that immediately after injection 100% of the activity resided in the injected animal. Development of the cross-circulation was determined by sampling the blood of each animal of a pair at several intervals after pairing and determining the per cent of available label that had transfused into the non-injected animal. This was done by determining the radioactivity present in equal volumes of blood from each animal corrected for hematocrit differences. No correction for blood volume was applied since the paired animals were about of equal weight and the blood volumes were therefore considered to be similar. The per cent of activity transferred was therefore represented by the CPM in the blood sample of the non-injected animal divided by the total CPM of both samples. Approximatelv 0.3 ml of blood was drawn into heparhized syringes from cardiac punctures. Aliquots were then taken for micro- 10 30 50 70 Hours after pairing 90 Fig. 2 Erythrocyte and plasma transfer at various sampling intervals after pairing plotted on a probability scale. 627 PARABIOTIC ANASTOMOSIS Hours after pairing Fig. 3 Comparison of erythrocyte and plasma transfer in the irradiated (one partner received 800 r) and non-irradiated pairs. hematocrits and 0.1 ml or 0.2 ml samples of the blood were used for determinations of radioactivity present. The activity of each sample was determined in a deep well chamber using a Nuclear-Chicago Scaler Model 186. Cross-circulation was followed with CrS1tagged erythrocytes in 15 nonirradiated pairs and in 14 pairs in which one member was irradiated. Transfer of RISA was followed in 15 non-irradiated pairs and in nine pairs in which on? partner had received radiation. RESULTS The accumulation of tagged erythrocytes and serum albumin in the non-injected partner of non-irradiated pairs is shown in figure 1. CrS' was f m t detected at 22 hours after pairing in very low amounts (0.04 t 0.01% ). Cr"' label accumulated slowly, so that at 47 hours after pairing only 2.1 t 0.46% of the label had been transferred. From this point on the label transferred to the non-injected animal at a faster rate, as indicated by the steeper slope, 8.4 t 1 . 5 % , 27.5 2 2.9%, 34.0 -+- 4.0%, and 47.4 2 1.5% at 54, 70, 77 and 95 hours, respectively. At the earliest sampling interval after pairing, five hours, Ii3' was found in small quantities in the non-injected animal (0.27 2 0.06% ). The label thereafter accumulated steadily, 3.7 2 0.4%, 8.3 t- 0.8%, 15.9 0.9%, 17.9 2 1.3%, 30.6 2.1%, and 44.2 2 1.5% at 22, 31, 46, 53, 70 and 94 hours, respectively. When the data were * plotted on a normal probability scale (fig. 2 ) , the points in the case of erythrocytes as well as RISA appeared to have a linear distribution. The differences in accumulation of labeled erythrocytes and labeled serum albumin were statistically significant at 22, 32, 47 and 54 hours ( P < 0.001). In those pairs where one partner was irradiated, the pattern of the transfer of the label was not significantly different from that exhibited by those which received no irradiation (fig. 3). DISCUSSION The shape of the curve showing the accumulation of the Cr5' labeled erythrocytes (fig. 1 ) into the non-injected animal was clearly sigmoid indicating that development of the vascular anastomosis followed a normal growth pattern. Although the shape of the curve for RISA accumulation was not as clearly sigmoid (fig. l ) , the fact that both curves showed a linear distribution when plotted on a probability scale indicated that transfer of RISA as well as tagged erythrocytes described a similar pattern of vascular inosculation. Transfer of the CrS1label began later after pairing than did RISA, which was found in the non-injected partner as early as five hours after surgical union. Appearance of RISA in the non-injected animal at such a short time after pairing suggested that its early transfer, at least, may not have been by va.scular anastomosis but 628 MARVIN SODICOFF AND ROBERT T. B I N H A M M E R rather by diffusion. Schiff and Plagge (‘55) have suggested that relatively small molecules might diffuse from one animal to the other rather than cross via blood vessels. Jacobsohn (’48) also maintained that parabiotic exchange was by non-vascular pathways and suggested diffusion into the lymphatics. Albumin, the smallest of the plasma proteins, escaped through capillaries in large quantities (Wasserman and Mayerson, ’ 5 2 ) , and was returned to the blood via lymphatic vessels. Injury to capillaries increased their permeability and allowed increased protein leakage. Incision of the skin and lateral body wall at the time of surgical parabiotic union certainly severed lymphatics, capillaries and larger blood vessels and allowed blood and lymph to escape into the tissue spaces. Since it has been shown that cut lymphatics may remain open for 24-48 hours (McMaster and Hudock, ’34), protein could have continued to enter the interstitial fluid for this period of time. It was this fluid, at the areas of surgical union, which was probably absorbed by the uninjured lymphatics of the other animal, thus resulting in transfer of the labeled protein at earlier time intervals than tagged erythrocytes. Eichwald et al. (’59) using injections of Evans blue immediately after surgery found that the dye passed to the non-injected animal one and two days after surgery. However, on these days the dye was limited to the region of the suture line and did not gain entrance to the vascular system. He found it wasn’t until the third day after pairing that the dye could be transmitted systemically. In contrast, the present work with tagged albumin indicated that it entered the vascular system within hours after surgery. Erythrocyte transfer proceeded very slowly until approximately 48 hours after union when a more rapid phase was initiated. These results suggested that during the first 47 hours after pairing only a few and/or small vascular channels were present, and after 47 hours the vessels of the capillary anastomosis had increased sufficiently in number and size to permit a rapid transfer. RISA did not give such a clear illustration of the development of the anastomosis, for it was not possible to detect any point in the transfer of RISA in which there was a sudden marked increase in transfer rate. The difference in transfer of erythrocytes and RISA was attributed to the physical characteristics of the labeled particle. The erythrocyte requires a vascular channel of greater size than does the albumin molecule. On the basis of the work of Clark (’18), who described the formation of new capillaries as originating as solid endothelial sprouts in which a lumen forms and then widens enough to permit circulation of blood cells, one would have anticipated earlier transfer of the small particle through the new anastomoses. The rather abrupt rise in rate of transfer of erythrocytes after 48 hours can thus be interpreted as indicating the time in the development of the capillary anastomosis when lumens of the newly developed capillaries have attained sufficient size to permit passage of erythrocytes. RISA showed no similar abrupt increase in rate probably due to its having a virtually nonexistent “threshold size.” In the later phase of the development of the anastomosis any diffusion of RISA via tissue fluid to lymphatics of the non-injected partner would probably be insignificant as compared to the vascular capillary transfer. This is supported by the fact that equal cross-circulation times for erythrocytes and albumin were found in pairs with well established capillary anastbmosis (Binhammer and Hull, ’62). Irradiation of one animal of each pair did not significantly change the development of the parabiotic anastomosis. The accumulation of Cr51 labeled erythrocytes and RISA in the non-injected (non-irradiated) partner followed much the same pattern as in those pairs where no irradiation was administered. This might support the work of Pohle et al. (’49) who found that skin irradiated with less than 1,000 r prior to its incision healed at a normal rate. ACKNOWLEDGMENT The authors gratefully acknowledge the fine technical assistance of Mrs. Marvin Sodicoff and Mrs. Paul Joyce. LITERATURE CITED Andresen, R. H. 1954 Cross circulation and tissue reaction in parabiosis. Arch. Path., 58: 455474. PARABIOTIC ANASTOMOSIS Binhammer, R. T., and J. K. Hull 1962 Cross circulation in normal and intoxicated parabionts. Proc. SOC.Exper. Biol. and Med., 2 2 : 134-139. Bunster, E., and R. K. Meyer 1933 A n improved method of parabiosis. Anat. Rec., 57: 339-343. Clark, E. R. 1918 Studies on the growth of blood vessels in the tail of the frog larva. Am. J. Anat., 23: 37-88. Eichwald, E. J., E. C. Lustgraaf and M. Strainer 1959 Genetic factors in parabiosis. J. Nat. Cancer Inst., 23: 1193-1213. Hill, R. T. 1932 Blood exchange and hormonic reactions in parabiotic rats. J. Exp. Zool., 63: 203-234. Jacobsohn, D. 1948 On the mode of action of ovarian hormones on growth and development 629 of the mammary gland. Acta Physiol. Scandinav., 17: Supp. 57. McMaster, P. D., and S. Hudock 1934 The participation of skin lymphatics in repair of the lesion due to incision and burns. J. Exper. Med., 84: 473494. Pohle, E. A., G. Ritchie and W. W. Moir 1949 Studies of the effect of roentgen rays on healing of wounds. Radiology, 52: 707-713. Schiff, G. J., and J. C. Plagge 1955 Use of fluorescein in testing the union of parabiotic rats. Proc. SOC. Exper. Biol. and Med., 88: 559-561. Van Dyke, D. C., R. L. Huff and H. M. Evans 1948 The efficiency of the vascular union i n parabiosis. Stanford M. Bull., 6: 271-275. Wasserman, K., and H. S. Mayerson 1952 Dynamics of lymph and plasma protein exchange. Cardiologia, 21: 296-307.