The Properdin System and Immunity. Iv. the Hemolysis of Erythrocytes from Patients with Paroxysmal Nocturnal Hemoglobinuria

Home , Properdin

THE PROPERDIN SYSTEM AND IMMUNITY. IV. THE HEMOLYSIS OF ERYTHROCYTES FROM PATIENTS WITH PAROXYSMAL NOCTURNAL HEMOGLOBINURIA

Carl F. Hinz Jr., … , William S. Jordan Jr., Louis Pillemer

J Clin Invest. 1956;35(5):453-457. https://doi.org/10.1172/JCI103296.

Research Article

Find the latest version: https://jci.me/103296/pdf THE PROPERDIN SYSTEM AND IMMUNITY. IV. THE HEMOLYSIS OF ERYTHROCYTES FROM PATIENTS WITH PAROXYSMAL NOCTURNAL HEMOGLOBINURIA 1 By CARL F. HINZ, JR.,2 WILLIAM S. JORDAN, JR., AND LOUIS PILLEMER (Froin the Departm7tentt of M11edicine, School of Medicine, and the Intstitutte of Pathology, Western Reserve University, Clevelantd, 0.) (Submitted for publication November 14, 1955; accepted January 9, 1956)

Paroxysmal nocturnal hemoglobinuria (PNH) main for 30 minutes at room temperature (200 to 25°C) is a rare chronic hemolytic anemia characterized and then stored overnight at 40C. Serum was separated by an acquired defect of the from the clot by centrifugation and stored at 4°C. On erythrocyte which storage at 4°C serum loses its ability to hemolyze PNH renders it susceptible to hemolysis by normal hu- erythrocytes in several days; therefore all sera were used man serum. Previous reports have indicated that within three days of collection. Serum, however, can be magnesium and factors resembling the components stored at - 70°C for at least six months without loss of of complement are required for the hemolysis in hemolytic activity. Serum was diluted with isotonic sodium chloride containing 0.0005 M magnesium chloride. vitro of PNH erythrocytes by normal serum Heated serum was prepared by heating in a 56°C water (1, 2). bath for 30 minutes. Properdin is inactivated in such Recently a naturally occurring serum protein, serum (3). properdin, has been described which requires mag- Hemi,olytic system: Except wvhere indicated all tests nesium and complement for its activity in vitro were conducted in a buffered system at pH 6.8 to 6.9. Since minor variations in pH have been noted to re- (3). Thus, properdin, magnesium, and comple- sult in marked changes in degree of hemolysis, a well ment have been termed the properdin system, buffered system is essential. The buffer was made by dis- which participates in the lysis of certain bacteria solving 3.4 gm. of imidazole in 90 ml. of 0.2 M hydro- and the inhibition of certain viruses (4, 5). chloric acid, and adding distilled water to 100 ml. This The resemblance between the factors required was titrated to pH 6.7 with approximately 0.2 volumes of normal hydrochloric acid (7). All sera were acidified for PNH hemolysis and the constituents of the and buffered after heating or other pre-treatment. The properdin system prompted this study on the role addition of 0.1 volume of 0.145 MI hydrochloric acid and of the properdin system in the hemolysis of PNH 0.1 volume of the buffer to one volume of serum, fol- erythrocytes. lowed by 0.1 volume of the 50 per cent erythrocyte sus- pension resulted in a stable final pH of 6.8 to 6.9. The METHODS mixture was incubated for 15 minutes at 37°C in a water bath. After incubation and centrifugation, hemolysis The test system in all instances was the hemolysis of was scored by inspection on a scale of 0 to +++, or the PNH erythrocytes obtained from three patients with free hemoglobin was determined quantitatively on an typical PNH whose case histories will be reported else- Evelyn photoelectric colorimeter to obtain per cent he- where (6). The red blood cells from all three patients molysis. Changes in the number of transfused cells pres- had similar characteristics. ent in patient's cell population caused the variations in Erythrocytes from oxalated or defibrinated blood were hemolysis observed in different experiments. washed three times in isotonic sodium chloride and re- Special serumiii Preparations: Serum samples from suspended to a concentration of 50 per cent in isotonic which components of complement, C'1, C'2, C'3, or C'4 sodium chloride containing 0.0005 M magnesium chloride. had been removed, referred to respectively as RI, R2, R3, They were stored at 4°C and discarded after 5 days. and R4, were prepared as previously described (8). Sera: Fresh, clotted venous blood was allowed to re- Properdin was removed from serum by exposing it to zymosan at 17° C (3). Serum rendered deficient of ' Presented in part to the Central Society for Clinical properdin by this means is called RP. It contains all Research, Chicago, Illinois, October 29, 1954. Supported measurable factors of the clotting system and all com- in part by research grant No. H1263C from the National ponents of complement in amounts comparable to un- Heart Institute, National Institutes of Health, U. S. treated serum. The RP used in these studies was re- Public Health Service; and by a grant from Lederle quired to meet the criteria previously established (3, 9). Laboratories Division, American Cyanamid Company, Purified properdin was prepared by a method described Pearl River, New York. elsewhere (3). It was concentrated to contain 100 to 700 2 U. S. Public Health Service Postdoctoral Research units per ml. in phosphate or barbital buffer and was di- Fellov. luted with isotonic sodium chloride immediately prior 453 454 CARL F. HINZ, JR., WILLIAM S. JORDAN, JR., AND LOUIS PILLEMER to use. Such preparations retained potency for at least TABLE II 6 months at - 20'C. The effects on PNVH and immune hemolytic systems of Resin-treated serum was prepared by the addition of removing properdin from serum an equal volume of Amberlite IRC-50 (sodium cycle) to serum for ten minutes at room temperature. In this way Hemolysis Hemolysis in whole in RP 99.9 per cent of calcium and magnesium ions were re- Type of hemolysis serum serum moved (10). Such serum was used immediately after PNH Present Absent preparation because of the instability of C'1 in the ab- Paroxysmal cold hemoglobinuria Present Present sence of calcium. Anti-A, Anti-B Present Present Thrombin: Lyophilized thrombin (Parke-Davis and Sensitized sheep cell Present Present Co.) was dissolved to a concentration of 1000 interna- Anti-human red cell Present Present Hemolysin from patient with tional units per ml., in isotonic sodium chloride contain- acquired hemolytic anemia Present Presenit ing 0.0005 M magnesium chloride. The solution was stored at - 20'C and thawed immediately prior to use. Fifty units (0.05 ml.) was added to 1 ml. of serum in the level of complement or properdin in that serum. PNH test system (11). However, the assay methods would not detect RESULTS small changes in either complement or properdin. 1. Effect of removing properdin on PNH he- 2. Restoration of the ability to hemolyze PNH molysis cells by the addition of purified properdin to The removal of properdin from serum rendered RP serum that serum inactive for the hemolysis of PNH The addition of purified properdin to RP to a erythrocytes (Table I). Fresh normal serum was final concentration of 5 units per ml. caused found to be more hemolytic at pH 6.8 than 7.4. marked PNH hemolysis, whereas RP or properdin At the optimum pH of 6.8, serum from which alone caused no hemolysis (Table III). Pro- properdin had been removed, (RP) was com- perdin prepared from cow or hog serum was also pletely inactive for hemolysis of PNH cells, even effective. Properdin did not restore hemolytic though it contained normal amounts of all the activity to Rl, R2, R3, R4 and heated serum. components of complement. Other serum sam- Comments: Removal of properdin from serum ples from which one component of complement renders the serum inactive for hemolysis of PNH had been removed, Ri, R2, R3, and R4, were cells. No other measurable substance is removed, also inactive for PNH hemolysis, although R2 for such sera are essentially unchanged as regards and R4 contained normal amounts of properdin. complement and the clotting factors. The sub- However, as indicated in Table II, the removal stance removed by zymosan behaves functionally from serum did not inactivate sys- of properdin as a single entity that is fully capable of restoring tems dependent on the action of a specific hemoly- hemolytic power to RP. The failure of properdin as paroxysmal cold hemoglobinuria, sin, such to restore sera inactivated by other methods again anti-A or anti-B isohemolysin, amboceptor sheep- or cell hemolysin, anti-human immune erythro- TABLE III cyte hemolysin. The hemolysis of PNH erythro- Restoration of ability to hemolyze PNH cells by the cytes in a serum did not result in a decrease in the addition of properdin to RP serum

TABLE I Hemolysis of PNH The effect on hemolysis of PNH erythrocytes of Serum Added properdin erythrocytes removing properdin from serum Normal 0 +++ RP 0 0 Hemolysis RP 5 units human +++ of PNH RP 5 units hog ++ Serum pH Properdin Complement erythrocytes RP 5 units cow +++ Normal 7.4 Normal Normal + Ri 5 units human 0 Normal 6.8 Normal Normal + + + R2 5 units human 0 RP 6.8 0 Normal 0 R3 5 units human 0 RI 6.8 Normal Lacks C'1 0 R4 5 units human 0 R2 6.8 Normal Lacks C'2 0 Heated 56°C 5 units human 0 R3 6.8 0 Lacks C'3 0 Isotonic sodium 5 units human 0 R4 6.8 Normal Lacks C'4 0 chloride PROPERDIN SYSTEM IN PNH HEMOLYSIS 455 demonstrates that other factors similar to or TABLE IV identical with the components of complement are The relation of serum properdin level to hemolysis of essential for hemolysis of PNH cells. Although PNH erythrocytes the inactivated sera were prepared by accepted Hemolysis Properdin of PNH methods, one cannot exclude the possibility that Serum content erythrocytes there are removed other unidentified substances units per ml. per cent unrelated to complement, but required for PNH Normal-No. 1 4-8 33 Normal-No. 2 1-2 14 hemolysis. The ability of properdin obtained from RP 0 0 several animal species to restore the human PNH RP + 3 units properdin 3 23 system strengthens the initial hypothesis that this RP + 5 units properdin 5 36 factor is a common and basic constituent of serum. tween properdin level and PNH activity. This in- 3. The requirement for magnesiumt in PNH consistency, presumably due to other limiting vari- hemolysis ables such as complement components, precludes the use of PNH hemolysis as a means of routine Removal of magnesium and calcium from serum measurement of properdin in serum at this time. by cationic exchange resin rendered that serum inactive for hemolysis of PNH erythrocytes. Ad- 5. Relation of properdin to the thrombin effect on dition to resin-treated serum of calcium to a PNH hemolysis final concentration of 0.0015 M did not restore the hemolytic property. Addition of magnesium in a In normal serum with moderate hemolytic ac- concentration of 0.0005 M resulted in complete tivity for PNH cells, addition of thrombin has restoration of the hemolytic system. Excesses of been noted to result in increased hemolysis. This magnesium resulted in inhibition of hemolysis, has been the basis of a diagnostic test for PNH probably because of the significant increase in total (11), and forms an important part of a theory ionic strength. Addition of 0.0015 M calcium to proposed by Crosby relating PNH hemolysis to 0.0005 M magnesium did not result in an increase the clotting system. Accordingly, the possible re- in hemolysis greater than that resulting from ad- lationship between thrombin and properdin effect dition of magnesium alone. was investigated. Thus, as previously demonstrated (1), there Both commercial and "pure" thrombin (Parke- is an absolute requirement for magnesium in the Davis and Co.) were found to contain properdin hemolysis of PNH erythrocytes. activity, varying from one unit to four units of properdin in every 50 units of thrombin, which is the amount used in the thrombin test. Neither 4. Relation of serunt properdin level to the PNH properdin nor thrombin was hemolytic alone, but hernolytic actiz'ity of serum the addition of equivalent amounts of thrombin or The properdin content of human sera correlated properdin to serum resulted in equal augmentation with the hemolytic activity of those sera against of hemolysis of PNH cells (Table V). Thus, a PNH erythrocytes. In a typical experiment serum initially containing 2 units per ml. of pro- (Table IV) a normal serum No. 1 containing 4 to perdin caused 21 per cent hemolysis. Addition of 8 units of properdin per ml. caused hemolysis of 50 units of thrombin "A", a preparation con- 34 per cent of the test cells, while another normal taining 2 units of properdin in each 50 units of serum No. 2 containing 1 to 2 units of properdin thrombin, caused 45 per cent hemolysis; addition per ml. caused only 14 per cent hemolysis. RP of 2 units of properdin resulted in 48 per cent he- alone failed to cause hemolysis, but the addition of molysis. On another occasion, using thrombin 3 units of properdin per ml. gave 23 per cent and "B", a preparation containing approximately 4 i units caused 36 per cent hemolysis. There was units of properdin in each 50 units of thrombin, a relationship between properdin content and he- addition of 50 units of thrombin caused increase molytic activity in many untreated sera and on in hemolysis in serum from 11 per cent to 54 per adding properdin to RP. However. some un- cent. Five units of properdin gave an increase in treated sera failed to show good correlation be- hemolysis from 11 per cent to 56 per cent. 456 CARL F. HINZ, JR., WILLIAM S. JORDAN, JR.. AND LOUIS PILLEMER

TABLE V ment to nornmal serum caused increased hemolysis. Effects on hemolysis of adding properdin and However, serum complement did not appear to be thrombin to serum depleted during the hemolysis of PNH cells, and Hemol- subsequent studies indicated that the hemolytic ysis of PNH activity of a serum could be destroyed without loss eryth- Date Thrombin Sertumi mnixture rocytes of conventional complement (12). Thus, com- per cent plement alone did not appear to be responsible for 9-25-54 lThrombin "A" hemolysis of PNH cells, although factors similar (50 units con- Seriiim 21 to complement appeared to be necessary for the tains 2 units of Serlim+50 ULlitS thronibi n 45 properdin) Sertim + 2 uinits properdini 48 hemolytic reaction. Crosby and Dameshek (13) proposed that the 1-14-55 TIhrombin "B" PNH hemolytic factor was related to ac-globulin, (50 units coin- Seruml 11 tains 4 units of Seruim+50 Iunlits thrombi n 54 but Harris, Jordan, Pillemer, and Desforges (1) properdin) Serum+ 5 Utiits properdiii 56 clearly showed that ac-globulin was not essential RI" 0 RP+50 units thrombini 27 for hemolysis of PNH cells. More recently, RP+5 uinits properdin 24 Crosby (14) has proposed a complex of four ac- R3+50 uiniits thrombini 0 R3+5 units thrOlllbiil 0 celerators and inhibitors of the PNH system related to the clotting mechanism. This theory had its origin in the observed clinical tendency to The addition of 50 units of thrombin "B" or 5 thrombosis in patients with PNH, and it is based units of properdin to RP resulted in 27 per cent in large part on the action of thrombin in causing hemolysis, but neither caused hemolvsis when augmentation of PNH hemolysis in vitro. None added to R3 or heated serum. of the four proposed factors has been identified in Larger amounts of thrombin have an anticom- support of this theory. plementary effect on serum and could not be tested. The current study offers evidence that the pro- Comments: The detection of properdin in perdin system, a natural defense mechanism of thrombin preparations andI the evidence that 1) human serum, is required for PNH hemolysis. the hemolytic property of thrombin corresponds In everN- manner tested the PNH system is coIml- to its properdin content, and 2) thrombin specifi- parable in requirements and behavior to the reac- cally restores an RP serum to hemolytic activitv, tions of the properdin system with zymosan, bac- suggest that properdin is the active principle re- teria and viruses. All these reactions require sponsible for the effect of thrombin. Accordingly. magnesiunm and factors indistinguishable from the activity of thrombin may be due to the addi- complemiient along with properdin, and all are de- tion of properdin contained in thrombin rather pendent on a specific temperature and pH. than to the inactivation of a proposed inhibitor by These observations support the findings of thrombin per se. The coexistence of properdin Ham and Dingle (2) that factors resembling comii- with thrombin in preparations of the latter is not plement are necessary for the hemolytic reaction surprising, for both are p3-euglobulins with similar against PNH erythrocytes. The demonstration iso-electric points. It was not possible to remove of the participation of properdin in PNH hemoly- properdin from thrombin preparations with zy- sis and its relation to complement may explain the mosan because of the requirement for magnesium hitherto confusing observations that 1) there is and the components of complement. no utilization of complement during hemolysis, and 2) hemolytic activity may disappear without DISC USSION depletion of total complement. That there are Both complement and the clotting system have additional undescribed factors affecting hemolysis been implicated in the hemolysis in vitro of PNH is entirely possible. erythrocytes. Originally Ham and Dingle (2) These studies may identify one factor of the showed that destruction of any component of com- complex proposed by Crosby, for it is probable plement resulted in loss of hemolytic activity of that properdin is one of the heat labile factors of serum, and that addition of guinea pig comple- earlier authors and may be the heat labile acceler- PROPERDIN SYSTEM IN PNH HEMOLYSIS 457 ator of Crosby. Further, properdin activity is 5. The PNH hemolytic system of normal serum probably responsible for the effect observed when includes properdin, factors resembling the com- thrombin is added to serum. The existence of a ponents of complement, and magnesium. heat stable inhibitor which is inactivated by thrombin was proposed on the evidence that augmented REFERENCES hemolysis resulted from addition of thrombin to 1. Harris, J. W., Jordan, W. S., Pillemer, L., and Des- serum and its subsequent removal by barium sul- forges, J. F., The enzymatic nature of the factor fate. Since properdin is present in thrombin in normal serum which hemolyses the erythrocytes and is not removed by barium sul- of paroxysmal nocturnal hemoglobinuria. J. Clin. preparations Invest., 1951, 30, 646. fate, the "thrombin effect" could well be due to 2. Ham, T. H., and Dingle, J. H., Studies on destruction the addition of properdin, a procedure which has of red blood cells. II. Chronic hemolytic anemia been shown to augment PNH hemolysis. Pro- with paroxysmal nocturnal hemoglobinuria: cer- perdin has no demonstrable relationship to the tain immunological aspects of the hemolytic mecha- by the fact that nism with special reference to serum complement. clotting system, as evidenced J. Clin. Invest., 1939, 18, 657. properdin preparations contain no activity similar 3. Pillemer, L., Blum, L., Lepow, I. H., Ross, 0. A., to the described clotting factors, and a serum or Todd, E. W., and Wardlaw, A. C., The properdin plasma from which properdin has been removed system and immunity: I. Demonstration and isola- is normal for known clotting factors when com- tion of a new serum protein, properdin, and its role in immune phenomena. Science, 1954, 120, 279. pared to an untreated control (3). 4. Wardlaw, A. C., Blum, L., and Pillemer, L., Bac- Thus, these studies indicate that the properdin tericidal activity of the properdin system in hu- system, consisting of properdin, components in- man serum. Federation Proc., 1955, 14, 480. distinguishable from complement, and magnesium, 5. Wedgewood, R. J., Ginsberg, H. S., Seibert, R. H., is required for the hemolysis of PNH erythrocytes and Pillemer, L., The inactivation of Newcastle role disease virus by the properdin system. Am. J. Dis. in vitro. The earlier observations on the Child., In press. of the properdin system in bacterial destruction 6. Hinz, C. F., Jr., Weisman, R., Jr., and Hurley, T. H., and virus neutralization (4, 5) and the require- In preparation. ment of properdin for hemolysis in vitro of PNH 7. Mertz, E. T., and Owen, C. A., Imidazole buffer: its cells, suggest that hemolysis of PNH erythrocytes use in blood clotting studies. Proc. Soc. Exper. Biol. & Med., 1940, 43, 204. may be due to the interaction of properdin with 8. Ecker, E. E., Pillemer, L., and Seifter, S., Immuno- the stroma of the intact erythrocytes. It has al- chemical studies on human serum. I. Human com- ready been suggested that the final destruction of plement and its components. J. Immunol., 1943, 47, PNH cells is due to a proteolytic enzyme system 181. (1), and the relation of properdin to such a system 9. Pillemer, L., Blum, L., Lepow, I. H., Wurz, L., and Todd, E. W., The properdin system and immunity. awaits further investigation. III. The zymosan assay of properdin. J. Exper. Med., 1956,103, 1. SUMMARY 10. Lepow, I. H., Pillemer, L., and Ratnoff, 0. D., The 1. Similarities between the factors required for influence of calcium ions on the inactivation of hu- the man complement and its components by plasmin. PNH hemolysis and the factors comprising J. Exper. Med., 1953, 98, 277. properdin system are pointed out. 11. Crosby, W. H., Paroxysmal nocturnal hemoglobi- 2. Removal of properdin from serum renders nuria. A specific test for the disease based on the the serum inactive for PNH hemolysis but not for ability of thrombin to activate the hemolytic fac- immune hemolytic systems; readdition of pro- tor. Blood, 1950, 5, 843. restores PNH 12. Wagley, P. F., and Hickey, M. D., Susceptibility of perdin derived from several species red cells and serum factor in the mechanism of hemolytic activity to such serum. hemolysis in paroxysmal nocturnal hemoglobinuria. 3. A relationship exists between the PNH he- J. Clin. Invest., 1948, 27, 559. molytic activities and properdin titers of sera. The 13. Crosby, W. H., and Dameshek, W., Paroxysmal noc- addition of properdin to weakly lytic sera also turnal hemoglobinuria. The mechanism of hemolysis and its relation to the coagulation mechanism. augments hemolysis. Blood, 1950, 5, 822. 4. The ability of thrombin to affect PNH he- 14. Crosby, W. H., Paroxysmal nocturnal hemoglobinuria. molysis may be due to the presence of properdin in Plasma factors of the hemolytic system. Blood, samples of thrombin. 1953, 8, 444.