576. 809. 737 : 615.361 .018. 55 THE EFFECT OF INJECTIONS OF SERUM ON

HBMOLYTIC COMPLEMENT, WITH PARTI- CULAR REFERENCE TO BORDET'S ANTI- I1MMUNE-BODY EFFECT.

SCOTT THOMSON. From the Department of Bacteriology, University of Edinburgh.

IN a previous paper (Thomson, 1938) it was shown that about a week after the subcutaneous injection of horse serum, the serum of a rabbit loses its a bacteriolytic activity for B. typhosus, P. cholera? and B. aertrycle. The p bacteriolysin does not appear to be affected. It was concluded that this result is due to an effect on the heat-stable antibody-like factor of normal serum. Injections of horse serum also produced an apparent inactivation of haemolytic complement when tested with sheep red cells sensitised by the addition of a hzemolytic immune body prepared in a horse (H.v.S. I.B.) but none with an immune body prepared in a rabbit (R. v. S. I.B.). This paper deals with an analysis of this phenomenon.

METHODS.

Complement was titrated by adding amounts of rabbit's serum ranging from 0.01 to 0.2 C.C. to 0-5 C.C. of a 3 per cent. suspension of thrice-washed sheep red cells sensitised by the addition of an excess of antiserum. The immune body and red cells were incubated a t 37" C. for, half an hour a t least and the complement added, after which the mixtures were incubated for 16 hours at 37" C. The " titre " was recorded as the reciprocal of the smallest amount of rabbit serum producing complete lysis of the test amount of cells.

Precipitins for horse serum were detected by means of a " ring " test, varying dilutions of horse serum (antigen) being layered on undiluted rabbit serum (antibody). The highest dilution of antigen producing a precipitate was recorded.

RESULTS. Experiments with horse serum.

The serum of a rabbit which has been previously injected twice with 10 C.C. of horse serum subcutaneously loses its ability to lyse sheep red cells sensitised with H. v. S. I.B. but is still active with R. v. S. I.B. on the 13th and 27th days after injection (table I).

Rabbits which showed this phenomenon had presumably developed antibodies to horse serum, and the absence of haemolysis

3 95

396 S C O T T T H O M S O N

Horse serum injected (c.c.).

with horse haemolysin could be explained in terms of a fixation of complement due to the interaction of these antibodies with the horse serum of the sensitising immune body. The phenomenon must be considered to be essentially the same as the ‘‘ anti-immune- body” effect (Bordet, 1904), which is that an antiserum to the serum proteins of a given animal species will inhibit the lysis by complement of cells sensitised with a haemolytic antibody prepared in that species. In the present experiments the fresh serum from a rabbit inoculated with horse serum contains antibodies thereto and so the complement it contains is unable to lyse red cells sensitised by a haemolytic antibody prepared in the horse.

TABLE I.

Effect of injection of horse serum on the hoemolytic complement of the rabbit.

Complement titre. .

H. w. S. I.B. I R.w.S. I.B. 1 Dar.

1 6 7

13 27 36

Day of experiment . . 1

Titreofcomplement (withH.v. S. I.B.) . 10

3 6 10 11 13 19

10 (5 0 0 0 10

10

10 ...

...

...

...

20 ( 5

0 0

10

... 13 30

25 20 25

...

Bordet considered that the serum of species A contained a natural hzmolytic immune body for the red cells he was using. An antiserum to the serum proteins of A contained an “ anti-immune- body ” which could prevent the action of even an artificially prepared haemolytic immune body from species A by keeping complement out of combination with the sensitised cells. But the results are more simply explained in terms of complement fixation by an antigen-antibody reaction.

The samples of horse serum used in these experiments were examined with excess of guinea-pig complement for the presence of a natural haemolysin for sheep’s cells and none was ever found.

Complement inactivation appeared about a week after a single subcutaneous injection of 20 C.C. of horse serum, and persisted for a varying period, usually a little more than a week (table 11).

TABLE 11.

T i m e of onset of complement inactivation.

COMPLEMENT A N D SERUM INJECTIONS 397

Complement titre.

H. w. S. I.B. R. W. S. I.B.

13 10 16 10

0 10 13 13

... ...

The time of onset would support the view that it was dependent upon the production of antibodies to the injected horse serum, but the phenomenon did not persist for as long as circulating antibodies were detectable. Table I11 shows that in an animal which had been injected twice intravenously at an interval of 5 days the hzmolytic complement had returned to its original state on the 15th day, although precipitins were still present to a high titre.

TABLE 111.

Relationship of precipitins to complement inactiwation.

Precipitins for horse serum.

0 0

512 1024

...

Day of expt.

","f" expt.

1 5 6 9

15

Horse serum injected.

Horse serum injected.

0.01.

5 C . C .

5 C . C . ... ... ...

0.02 _ _

0 0

0 0 0 1 0 1 0 0

. . . . . .

0.04.

Not infrequently, instead of complete absence of complementing action, a zone phenomenon was obtained, as shown in table IV.

TABLE IV.

Zone phenomenon in the titration of complement of a rabbit injected with horse serum.

0.05 _ _

2 4

1 2 2 3 3 4 1 2

. . . . . .

0:l.

Precipitins for

horse serum. 0.15

-_ 5 5

4 4 4 3 5 5 3 5

. . . . . . 1 6 7

12 21 35

5 C.C. 5 7 1 ...

...

...

...

0

256 2048 8192 256

...

Lysis produced by varying amounts (c.c.) of rabbit serum with H. W. S. I.B.

- 0.03 -

1

1 2 2 1

...

- 0.06 -

5

4 4 5 2

...

- 3.075. -

5

4 4 5 3

...

-

0.2. -

5

4 2 5 5

...

- 0 = no lysis ; 1 = trace of lysis ; 2 = distinct lysis ; 3 = marked lysis ;

4 = very marked lysis ; 5 = complete lysis.

A zone effect was present on the 12th day, and the table also shows that precipitins were present to a high titre on the 21st day, yet the complementary activity had returned to its former state. The difference in activity of complement on the 21st and 35th days is not regarded as significant, but as due to the use of different samples of sensitised cells.

Table V shows how repeated injections of serum may maintain the phenomenon over long periods.

JOUBN. OF PbTE.-VOL. XLWL 2 c 2

398 SCOTT THOMSON

Dayofexperiment . . Horse serum injected .

Complement titre with H. v. S. I.B.

1 7 8 14 16 21 24 28 30

5 C.C. 5 C.C. ... 5 C.C. ... 5 C.C. ... ... ... 16 0 0 0 0 0 0 0 0

A serum which has lost its power to lyse sheep cells sensitised with H. w. S. I.B. can also inhibit lysis by an active complement. In one experiment 0.1 C.C. of an experimental serum produced complete inhibition of 5 M.H.D. of normal complement. This degree of inhibition, however, was exceptional. Furthermore, the serum of an experimental rabbit was more inhibitory after heating to 55" C. than when used unheated. The observed differences may have been due to lysis by complement still present in the unheated serum.

A 3 per cent. suspension of red cells was sensitised by the addition of 5 M.H.D. of H. w. S. I.B. and the mixture incubated for an hour at 37" C. to allow fixation to occur. A portion of the cells was centrifuged and washed three times in a large excess of normal saline before being reconstituted into a 3 per cent. suspension and the serum of a rabbit previously injected with horse serum was titrated for complement against washed and unwashed suspensions of sensitised cells. With unwashed cells no lysis was obtained whereas with washed cells the serum was actively hzemolytic. This would clearly indicate that the inhibition of complement was largely dependent upon the presence of uncombined horse serum proteins of the sensitising antiserum. Washing with excess of saline was not uniformly successful in removing the inhibitory substances, which raises the question of the relationship between an antibody and the globulin of the serum with which it is associated. The antibody molecules fixed to the red cells may still behave sero- logically as serum globulin.

The ability of the heated serum of a rabbit previously injected with horse serum to inhibit the lytic action of complement from a normal rabbit was examined with an ordinary suspension of sensitised cells and with one which had been washed in saline after sensitisation. With washed cells no inhibition of lysis was produced, but the washing was not uniformly successful in removing the inhibition, which would indicate a certain difficulty in washing the sensitised cells free from horse serum proteins.

It was to be expected that the addition of normal horse serum to the sensitised cells would interfere with the action of complement from a rabbit injected with horse serum. Table V I gives the results of an experiment designed to show this.

This point will be discussed later.

C O M P L E M E N T AND SERUM I N J E C T I O N S 399

TABLE VI. Effect of horse serum on complement activity of serum of injected rabbit.

added.

Nil

H. v. S. 7 M.H.D.

Yonnal horse serum

Sheep red cells sensitised with

7 M.H.D. of I.B. 0.01. 0.02. 0.03. 0.04.

- ~ - _ _

0 0 1 1

0 1 1 2

0 0 1 2

0 0 1 1

1. H. v. S.

2. R. v. S.

3. R.v. S.

4. R. v. S.

0.1.

Haemolysis produced by varying amounts (c.c.) of serum of rabbit injected with horse serum. Horse serum

0.2. --

0 0

4 5

0 0

0 0

0.06. --

0.05. -

1

2

2

1

0.075

1 0

3 3

2 1

1 0

0.1. 0.2. -__

0 0

5 5

5 5

The amount of normal horse serum added in row 4 was equal to the amount of immune horse serum present in rows 1 and 3.

The inhibition of lysis was practically the same with normal horse serum as with H. U. S. I.B., whether the latter was used alone (1) or with R. U. S. I.B. (3), suggesting that the inhibition was dependent upon the horse serum proteins and not on any immune constituent. Although obtained in most cases, this result was by no means regular, as shown in table VII.

TABLE VII. Effect of normal horse serum on the complement activity of the serum

of a rabbit injected with horse serum.

Sheep red cells treated with

1. H.v.S.1.B. . 2.R.v.S.I .R. .

3. R. v. S. I.B. +normal horse serum

Haemolysis produced by varying amounts (c.c.) of serum of rabbit injected with horse serum.

0.075.

The part played by the proteins of the haemolytic antiserum in fixing complement was further studied by mixing the .various reagents in different orders. Since there is fixation of complement by combination of the horse serum proteins with the antibodies thereto present in the rabbit serum, one would expect the degree of fixation to be greater if the two sera were mixed and allowed to stand for an interval before the addition of the sensitised cells. This would allow the antigen-antibody reaction to proceed and fix complement before the sensitised cells were added. This experiment was carried out with normal horse serum and horse

400 SCOTT THOMSON

0.05. -~

serum containing hemolytic antibody for sheep cells. The results are shown in table VIII.

0.06

5 5 2 2 0 0

0 0

5 5 2 3

TABLE VII1. Effect of normal and hmmolytic horse serum on complement activity of serum

of rabbit injected with horse serum.

1.1.

Titration no.

0.2. -- 5 5 0 0 0 0 I 0 0

5 5 0 0

Cells sensitised

with

...

... 30mins.at

room

R. v. S. I.B. H. v. S. I.B. R. v. S. I.B.

R. v. S. I.B.

R. v. S. I.B. R. c. S. I.B.

0 1 0 0 0 0

Horse serum added to

serum 632.

Nil

Normal

H. 21. S. I.B.

H. v. S. I.B. Normal

temperature 30 mins. a t

room l o

Hremolysis produced by varying amounts (c.c.) of serum 632 from rabbit injected with horse serum.

Interval.*

O

10.01.1 0.02.1 0.03 I-i-I-

3 1 0

0

1 ;

0.04 1,075.

5 1 0

0

5 1 -

The amount of horse serum added was the same in all instances. * Interval = time between addition of horse serum and of sensitised cells.

There was clearly less lysis when horse serum, either normal (3) or as H. v. S. I.B. (4), was added to the complement before the addition of sensitised cells than when it was added along with them (5 and 6). The discrepancy in the readings of 5 and 6 is not disturbing, because serum 632 was able to produce some lysis of cells sensitised with H. v. S. I.B. (2).

Although the phenomenon has been explained in terms of complement fixation between the horse serum proteins of the sensitising immune body and their antibodies in the rabbit serum, it does not follow that this would be independent of the quantity of sensitising immune body used. It is shown in table IX that the complement inhibition was absent when larger amounts of H. v. S. I.B. were used.

Excess of H. Y. S. I.B. could therefore be used to reveal the presence of complement in this rabbit serum. This result was not unexpected, because Dean (1911-12) has emphasised the fact that when complement is fixed by an antigen-antibody reaction the total amount of complement fixed need not depend upon the total amount of precipitate formed. If a fine precipitate is formed slowly, much more complement is fixed than if the same total amount of precipitate appears quickly as a coarse deposit. Further, the rate of formation of a precipitate depends upon the relative amounts of antigen and antibody, and in the experiment with 15 M.H.D. of H. v. S. I.B. (table IX) the rate of formation of precipitate may have been altered because of a relative antigen excess and some of the complement thus left unfixed.

C O M P L E M E N T A N D SERUM I N J E C T I O N S 40 1

0.01.

TABLE IX.

Effect of varying the sensitising dose of hcemolytic horse serum.

0.02. ~-

0 0

0 0

0 1

Amount of H.v.S. I.B.

added. 0.03. 0.04. 0.05. 0.06. ~ ~ ~ - - 0 0 0 0 0

1 1 1 1

2 2 2 3 4

3M.H.D. .

0.075.

2

Lysis produced by varying amounts (c.c.) of serum of rabbit injected with horse serum.

0.1. -___

1

0.2.

0 0

0

5 5

Mixed with serum of injected rabbit and washed 4 times

Washed 4 times .

0

. o

Reference has already been made to the difficulty of washing sensitised sheep cells free of the horse serum proteins used in sensitisation. In a further experiment 1 C.C. of an experimental rabbit serum which would not lyse sheep cells with H. U. S. I.B. was added to 5 C.C. of a 3 per cent. suspension of cells so sensitised. After an hour the cells were centrifuged and washed four times in a large excess of saline and reconstituted into a 3 per cent. sus- pension. The fresh serum of a normal rabbit was titrated for complement with (1) the suspension treated as above and (2) a suspension of cells similarly sensitised and washed but untreated with the serum of an experimental rabbit (table X).

TABLE X.

Persistence of the effect of the preliminary exposure of sensitised cells (H.v.S. I.B.) to the serum of a rabbit injected zcrith horse serum.

Lysis produced by varying quantities (c.c.) of normal rabbit serum. Preliminary treatment of

sensitised cells.

This indicates that it is difficult to wash away the product of the interaction between the horse serum and the antibodies thereto present in the serum of the injected rabbit.

Experiments with ox serum. Subcutaneous injections of ox serum in rabbits produce anti-

bodies which react with horse serum and vice versa (Kolmer, 1923 ; Thomson, 1938). They were found to produce a fall in the activity

402 SCOTT THOMSON

of hEmolytic complement with cells sensitised with H. U. S. I.B. but not with cells sensitised with R. v.S. I.B. The effect was less marked than with horse serum but was quite definite (table XI).

In one animal a complete inactivation of haemolytic complement was found ; in the other a zonal phenomenon appeared indicating some inhibitory mechanism.

TABLE XI.

Effect of injections of ox serum on the activity of rabbit complement.

Expt. 1 Day of experiment . Ox serum injected . Titre of complement (H. U. S.

Titre of complement (R. o. S. I.B.)

I.B.)

Expt. 2. Day of experiment . Ox serum injected . Titre of complement (H. o. S.

Titre of complement (R. 'U. S. I.B.)

I.B.)

1 5 C.C.

20

15

1 20 C.C. 13

13

8 5 C.C. 5

10

4

13

15

...

12

0

15

...

7

10

15

...

__

23

20

15

...

11

3 ;zone

15

...

~

...

...

...

...

18

6

13

...

__

...

...

...

...

26

20

15

...

Experiments with sheep serum. Six rabbits were injected twice with 10 C.C. of sheep serum

a t an interval of one week. There was no reduction in haemolytic complement activity. Presumably, therefore, the suspension of sheep cells had been washed free of sheep serum, otherwise it would have produced some degree of complement fixation through inter- action with antibodies in the rabbit serum. This would indicate that the sheep cells were easily freed from the homologous serum by washing three times in saline. Attention has been drawn to the difficulty of washing sensitised cells (sheep) free of the proteins of the antibody-containing serum (horse). In this case the wash- ing was much more thorough because a 3 per cent. suspension of cells containing less than 1 per cent. of horse serum was washed with a large excess of saline, whereas the unsensitised cells were washed with only three or four volumes of saline to one volume of deposited cells. It would appear that either cells can be more readily washed free from their own than from a foreign serum, or else the specific antibody molecules of the haemolytic antiserum which combine with the cells retain the serological properties of the serum proteins. Interference with precipitate formation and complement fixation by the presence of a large antigen (sheep serum) excess when sheep cells are washed free of homologous serum has not been entirely excluded. This latter factor probably plays no part, because this particular haemolytic system (sheep

COMPLEMENT A N D SERUM INJECTIONS 403

cellsfH. v. S. I.B.) is universally used in routine complement fixation tests and the cells do not have to be washed with particular care.

DISCUSSION.

The serum of a rabbit previously injected with horse serum is unable to act as complement and to lyse a suspension of red cells sensitised by a hemolytic immune body prepared in a horse. This phenomenon has been shown to be essentially the same as Bordet’s “ anti-immune-body ” effect, and can be simply interpreted in terms of complement fixation between the serum proteins of the hemolytic horse serum and the antibodies thereto which are present in the serum of the injected rabbit. If the sensitised cells were thoroughly washed to remove the free proteins and leave the specific antibody fixed to the cells the rabbit serum frequently produced lysis, though the effect of washing was somewhat irregular owing to the fact that the horse hemolysins adhered to the cells more firmly than did homologous serum and might subsequently react with antibodies in the serum of the injected rabbit as normal serum proteins. A sensitised particulate antigen is said to behave in some respects as though coated with a layer of antibody-serum- globulin (Marrack, 1934) and it may be difficult to remove this globulin from an antigen-antibody combination.

Although the explanation advanced of the inhibition of lysis is quite a simple one, there are a few interesting applications to be considered. It might be expected that the serum of an animal injected with horse serum would give an anticomplementary result in a complement fixation test in which the serum used to sensitise the suspension of red cells was derived from a horse. This was investigated in rabbits whose sera showed no fixation of complement with the Wassermann antigen. Samples of sera were collected before and after injections of horse serum and tested in a com- plement fixation test with Wassermann antigen. The tests, which were all done at the same time, were performed in duplicate with H. v. S. I.B. and R. v. S. I.B. In none of the tests did the serum appear to be anticomplementary, i.e. the serum controls never fixed 2 M.H.D. of complement (the lowest amount used). A result has been quoted where 0.1 C.C. of serum from a test rabbit was able to fix more than 5 M.H.D. of a fresh complement, and no doubt this serum would have given an anticomplementary result in a complement fixation test but this serum was recognised to be unusually active in this respect.

Of perhaps greater interest is the fact that the phenomenon soon disappeared although the precipitins for horse serum were still present in the rabbit serum. The relationship of complement fixation to precipitation has already been discussed, and this result

404 SCOTT THOMSON

does not seem to us to invalidate the opinion advanced to explain the inhibition phenomenon, for Dean has shown that a reaction of this nature depends on optimal proportions of the reagents, excess of one over the other yielding a diminished effect.

Xummary . 1. The serum of a rabbit previously injected with horse serum

cannot lyse red cells sensitised by a haemolytic immune body from a horse.

2. This is due to the fixation of complement by the interaction of the horse serum proteins of the sensitising immune serum and antibodies present in the injected rabbit’s serum.

3. The phenomenon does not run parallel to the amount of antibodies for horse serum proteins present in the rabbit serum.

4. It is essentially the same as Bordet’s “ anti-immune-body ” effect, which can be interpreted in terms of complement fixation.

5. It is more difficult to wash red cells free from the proteins of a sensitising antiserum than from homologous serum.

I wish to thank Professor T. J. Mackie for his interest in this work, the expenses of which were met by a grant from the Earl of Moray Fund.

REFERENCES.

BORDET, J. . . . . . . 1904. Ann. Inst . Past., xviii. 593. DEAN, H. R. . . . . . 1911-12. Proc. Roy. Soc. Med., v., Sect.

Path., 62. KOLMER, J. A. , . . . . 1923. Infection, immunity and biologic

therapy, 3rd ed., Philadelphia and London, p. 313.

MARRACK, J. R . . . . . . 1934. Medical Research Council, Special Report Series, no. 194, London, p. 50.

THOMSON, S. . . . . . 1938. this Journal, xlvii. 131.