378 Increase in the Complement-content of Fresh Blood-serum .
By J. Henderson-Smith , M.B. , Philip Walker Student of Pathology in the University of Oxford .
( Communicated by Professor Gotch , F.R.S. Received March 26 , \#151 ; Read May 9 , 1907 .
) Very soon after coagulation has occurred in blood withdrawn from the vessels , serum begins to be expressed .
The amount , at first small , increases rapidly , and in a few hours , at room temperature , contraction of the clot is practically complete .
Some observations made in the course of other work suggested that the composition of the serum first yielded differs from that of the serum expressed later , and a special investigation of the point led to the results recorded in this paper .
The serum given off from the clot at varying intervals of time after bleeding was tested for its content in haemolytic complement .
This body was selected for examination , not only because it can conveniently be studied in the test-tube , but also because much greater quantitative accuracy of result can be obtained with it than with , for example , a bactericidal complement .
In most of the experiments , the serum of rabbits was tested against the red blood-corpuscles of the ox , and the necessary immune-body was derived from the serum of rabbits immunised by repeated injections of these corpuscles .
In some instances other combinations were used , as is stated later , but in every case the red cells used for injection and for examination were fresh , and had been freed from serum by being washed thrice in many times their bulk of isotomic salt solution .
The experiments fall into two groups , according as the serum examined was from a normal animal or from an animal that had been previously immunised .
1 .
Normal Serum .
The method adopted was as follows:\#151 ; A rabbit was treated with injections of ox red blood-corpuscles , and when its serum had reached a sufficiently high titre , its blood was withdrawn and the serum obtained .
This was then heated for half an hour at 56 ' C. to destroy the contained complement , and served as a stock of " immune-body " ( by which name I shall refer to it in the rest of this paper ) .
The neck of a normal rabbit having been shaved and sterilised , the large vessels of the neck were cut , and the issuing blood was caught in a sterile flask .
As soon as coagulation had occurred , the clot was loosened from the sides of the flask by a sterile glass-rod .
At varying intervals of time , samples of the separating serum were taken and Increase in the Complement-content of Fresh Blood-serum .
379 rapidly centrifuged to free them from cells , and the supernatant fluid was then tested .
Each sample was examined separately as soon as it was obtained , instead of waiting for the last sample and examining them all together .
The latter method was used as a control in a few cases , but the interval between the taking of the first and last samples extended over several hours , during which time spontaneous deterioration or other change of the complement might easily occur in the earlier samples .
Such a change does , in fact , occur , as will be shown later , and it was necessary to avoid this source of error .
The examination of a sample was made in the following way:\#151 ; A series of test-tubes was prepared , and into each was measured the same amount of immune-body ( the necessary dose having been previously ascertained ) .
Diminishing amounts of the sample of serum were then added to the test-tubes , the first tube receiving a large dose , the next a slightly smaller , the third a still smaller , and sp on .
The volume of fluid in all the tubes was then made the same by adding salt solution , and finally 4 c.c. of a 1-per-cent , suspension in salt solution of ox corpuscles were added to every tube .
Control-tubes having been prepared , all tubes were shaken and placed in an Ostwald water-bath at 37 ' C. After two hours the tubes were withdrawn , and of the series the lowest tube , in which total haemolysis had occurred , was noted .
The dose of serum in this tube contained just enough complement to produce total haemolysis in that time .
The same procedure was carried out with every sample , the amounts of immune-body and red cells being kept constant throughout .
An example of such an experiment is given below .
Eocample 1.\#151 ; Dose of immune-body = 0*04 c.c. ; 4 c.e. 1-per-cent , suspension ox red blood-corpuscles .
Total = complete haemolysis .
Time , two hours at 37 ' .
Dose of Time in hours between bleeding of rabbit and withdrawal of sample .
serum in c.c. |j i 1 .
2 .
3\#163 ; .
5 .
7 .
i 0 '4 1 total total total total total total total 0*35 i 99 99 99 99 99 99 99 0*3 \gt ; 9 ; 99 99 99 99 99 0-25 !
not total 99 99 99 99 99 .
99 0*2 less not total 99 99 99 99 99 0*15 less not total 99 99t 99 99 0-1 less not total 99 ' 99 99 0*08 less not total 99 I not total 0-06 less not total less 0*05 less 380 Mr. J. Henderson-Smith .
Increase in the [ Mar. 26 , No lysis was produced by 0*35 c.c. of the 5-hours sample without immune-body ; nor by 0 04 c.e. of the immune-body without the normal serum .
In this case , then , 0*3 c.c. of the serum , withdrawn 20 minutes after bleeding , was required to produce total lysis ; but with the successive samples the amount required steadily decreased , until of the serum which had remained five hours in contact with the clot 0*08 c.c. was sufficient .
The amount of complement contained in the later sample , or at least its activity , must be more than three times as great as that of the complement in the earliest sample .
It increased steadily for five hours , and then began to diminish ( decreasing afterwards till it eventually disappeared ) .
This example is typical .
In all the experiments ( seven ) done with normal rabbit 's serum and ox corpuscles , a similar increase in the complement has been observed .
The amount of the increase varies , but as a rule it is to double , sometimes to three times , the quantity found in the first sample .
The time during which increase occurs also varies .
The longest time observed was 14 hours ; but this is exceptional , and in most cases the rise is complete in five to seven hours .
The same phenomenon occurs when , instead of ox corpuscles and the appropriate immune-body , horse corpuscles are used with immune-body from rabbits treated with injections of horse corpuscles .
When normal guinea-pig serum is used instead of rabbit serum to supply the complement for either ox or horse corpuscles ( with appropriate immune-body from rabbit ) , the same increase to about double the amount is found .
The duration of the increase , however , is shorter , extending to two to four hours .
( In one of six experiments with guinea-pig serum the rise was very slight , and complete in one and a-half hours .
) It may be suggested that this increase in the complementing-power is only apparently a change in the serum , and that the change really occurs in the red corpuscles used as the reagent .
In the interval between the examination of the first and last samples the red cells may become more sensitive , and so give a false appearance of increased power in the serum .
This , however , cannot be the explanation of the facts observed .
The red corpuscles were kept in the cold room ( about 3 ' C. ) till immediately before use , and were always fresh .
Under these conditions they do not become perceptibly more sensitive within so short a time as seven to eight hours , when tested with other lytic bodies , such as bacterial haemolysins and chemical reagents .
Further , the results , described later , obtained with immune sera prove that no such change as is here suggested occurs in the corpuscles used .
It was of interest to determine whether this change in the complement-content of serum occurs even when the serum is removed from the clot , or takes place only when it remains in contact with the Clot .
This was 1907 .
] Complement-content of Fresh Blood-serum .
381 investigated in a number of the normal serum experiments .
A portion of the first sample was set aside in the cold after centrifuging , and examined .some hours later , at the same time as a sample which had remained on the clot during the same period .
The results show that a rise does occur in the serum , even when separated from the clot .
The amount of this rise is , however , very small , only a small fraction of the increase which has taken place in the serum left on the clot .
E.g. , in the example given above , the value of the 20 minutes sample rose in five hours , when off the clot , from 0*3 c.c. only to 0-2 c.c. , as against the rise to 0*08 c.c. of the serum left on the clot during that time .
It is not easy to suggest an explanation of this phenomenon .
We might suppose that complement exists in serum or plasma partially in an inactive or zymogen phase , which is slowly converted into the active body .
Or again , on the view that complement is derived from leucocytes , we may suggest that even after prolonged centrifugalisation , fragments of white cells remain suspended in the serum and gradually disintegrate .
Such explanations , however , have as yet no real experimental basis , and we know too little of the nature and source of complement to make speculations profitable .
In the normal sera examined , then , haemolytic complement increases by 100 to 200 per cent , during the first few hours after the drawing of blood , the duration of the increase extending in the case of the rabbit over a period of five to seven hours , in the case of the guinea-pig over a period of two to four hours .
It is probable that this variation is not peculiar to the haemolytic complements .
Ainley Walker ( 1 ) found the bactericidal action of serum to increase during the first few hours after the blood was shed , and attributed this to an increase in the amount of complement .
According to Macdonald ( 2 ) , whose results have been confirmed by Craw , the opsonic power of serum undergoes a similar increase .
J. H. Craw has pointed out that all these bodies are relatively unstable , while the haemolytic immune-body , which does not vary , is comparatively stable ; but whether this fact is of significance can only be determined when further evidence has accumulated on the subject .
The results obtained in the experiments recorded here have no immediate bearing on Metchnikoff 's view that alexine does not exist as such in circulating plasma .
Sufficient serum for examination does not separate until 15 to 20 minutes have elapsed from the time of bleeding .
In the serum then obtained the amount of complement , though much less than is found later , is still always considerable , and there was nothing to indicate that at the moment of bleeding the blood contained no active complement .
" Example 2.\#151 ; Dose of immune-body : 0'025 .
Horse red blood-corpuscles : 4 c.c. of 1-per-cent , suspension ; two hours at 37 ' .
Quantity of normal VOL. LXXIX.-\#151 ; B. 2 F 382 Mr. J. Henderson-Smith .
Increase in the [ Mar. 26 , rabbit serum necessary for total haemolysis was , of 1 hour sample , 0 6 c.c. 2| hours sample , 0*5 c.c. ; 5 hours sample , 0*3 c.c. ; 7\#163 ; hours samplo , 0*325 ; 9 hours sample , 0*35 c.c. ; 12 hours sample , 0*4 c.c. Hence a rise to double the amount in 5 hours .
In the control tubes , 0*5 c.c. of the 1 hour sample gave no lysis , but the same dose of the 5 hours sample gave a trace .
Example 3.\#151 ; Dose of immune-body , 0*04 c.c. Ox corpuscles : 4 c.c. 1-percent .
suspension ; 2 hours at 37 ' .
Quantity of normal guinea-pig serum necessary for total haemolysis , of \#163 ; hour sample , 0*05 c.c. ; of 1 hour sample , 0*025 c.c. ; of 4 hours sample 0*03 c.c* 2 .
Immune Serum .
The immune sera examined were those of animals , mostly rabbits , injected with blood-corpuscles , usually of the ox .
In some cases the serum came from rabbits , which had been injected with horse corpuscles , and in one case from a goat injected with ox corpuscles .
The results in all the cases were concordant .
Two factors in such sera require to be examined : the immune-body and the complement .
The immune-body is easily tested , since it does not appreciably , if at all , deteriorate spontaneously .
The immunised animal was bled as before , and samples of serum withdrawn in the usual way .
Each sample , after centrifuging , was at once heated 30 minutes at 56 ' to destroy the complement , and , when the last was ready , all were examined together .
The method was similar to that described in the last section .
With a constant quantity of corpuscles and a constant quantity of serum from a normal animal as complement , the minimum dose of each sample necessary to produce total lysis was determined .
This was found to be the same for all the samples ; and in all the experiments this proved to be the case .
The immune-body , therefore , does not vary , but is present in its full amount from the first Example 4.\#151 ; Dose of normal rabbit serum as complement , 0*15 c.c. ; 4 c.c. 1-per-cent , suspension ox corpuscles ; 2 hours at 37 ' C. Immune serum from rabbit injected with ox corpuscles .
The dose required of this serum after heating was for the 1 hour , 2 hours , 4 hours , 6 houra , 8 hours , and 12 hours samples the same , viz. , 0*04 c.c. The dose of the 10 hours sample was 0*045 c.c. , which is probably an experimental error .
The investigation of the complement in immune serum is more difficult .
We have no means of destroying immune-body , and leaving complement in the serum , and Ehrlich 's ( 3 ) method of absorption of the immune-body in the cold is hot suited to quantitative estimations of this kind , and has been 1907 .
] Complement-content of Fresh Blood-serum .
383 shown by Mainwaring ( 4 ) to be open to serious objections .
On the other hand , we cannot simply ignore the immune-body contained in the serum .
By taking varying doses of the serum we should be taking not only varying quantities of complement , but also varying quantities of immune-body ; and Morgenroth and Sachs ( 5 ) have shown that the amount of complement necessary for haemolysis depends in some measure on the amount of immune-body present .
Now , in the process of immunising , the animal responds by a great increase of immune-body in its serum , but there is no corresponding increase of complement .
Of the serum of a highly immunised animal , therefore , it is possible to find a dose which contains enough immune-body to cause total laking of the chosen quantity of corpuscles , but does not contain enough complement .
Complete haemolysis , therefore , does not occur with that dose , unless we make good the lack of complement by adding some from another source .
By making use of this fact , a method was devised of estimating the complement in immune serum .
An immunised animal was bled and samples of serum were withdrawn from the clot in the usual way , each sample being tested at once after centrifuging .
Such a dose as has been just described , containing enough immune-body but not enough complement , was taken , and the same dose was taken for all the samples .
Since , as we have just seen , the immune-body does not vary in the different samples , we are thus taking always the same amount of immune-body .
The deficiency in complement was supplied by adding the serum of a normal rabbit , and the quantity of this serum which had to be added was carefully determined for each sample .
The greater the quantity of normal serum required , the less complement must the dose of immune serum contain .
The amount of normal serum that would be required , if the dose of immune serum contained no complement , is readily found by first heating the latter and so destroying its contained complement .
The difference between this amount and the amount we had to add when the sample was unheated gives us a measure of the value of the complement contained in the unheated dose , this value being expressed in terms of the normal serum .
We can thus determine the amount of complement contained in each sample , and so ascertain the extent of the variation , if any .
Example 5.\#151 ; Four cubic centimetres of 1-per-cent , suspension of ox corpuscles ; 2 hours at 37 ' C. Immune serum from rabbit , injected with ox corpuscles ; of this serum , when heated , 0*02 c.c. required the addition of 0*25 c.c. normal rabbit serum ( i.e. , complement ) to produce total haemolysis , and this quantity of the normal serum by itself produced no lysis .
2 f 2 Mr. J. Henderson-Smith .
Increase in the [ Mar. 26 , c.c. c.c* To 0*02 unheated immune serum , I hr .
after bleeding , must be added 0*225 normal serum .
99 99 i 99 99 0*2 tf 99 2 99 99 0*225 " 9i 99 3f 99- 99 0*25 99 99 5 99 99 0*25 " 0*02 c.c. unheated serum by itself , no perceptible lysis .
Since 0*25 c.c. normal serum was required , when the immune serum had been heated , there was contained in the dose of the unheated immune serum\#151 ; c.c. c.c. c.c. Of the \ hr .
sample complement equivalent to 0'25\#151 ; 0*225 = 0*025 of the normal serum .
\#187 ; 1 \gt ; \gt ; \#187 ; \#187 ; \gt ; 0*05 , , , , \gt ; \#187 ; ^ j\gt ; \#187 ; \#187 ; 0*025 , , " , , 3f " " " No appreciable amount .
\#187 ; \#174 ; \#187 ; \#187 ; \#187 ; \#187 ; \#187 ; The amount of complement increased to double the amount by the end of the first hour , but had decreased again after two hours , and by four hours had become inappreciable .
Example 6.\#151 ; Immune serum from rabbit injected with ox corpuscles ; of this when heated 0*05 c.c. required 0*13 c.c. normal rabbit serum .
To 0*05 of the immune serum unheated\#151 ; c.c. From the \#163 ; hr .
sample must be added 0*1 normal serum .
99 1 99 99 0*085 99 99 99 99 0*08 99 99 ^ " 99 0*1 99 99 \#174 ; 99 99 0:1 99 C.C. Therefore the dose of the I hr .
serum contained the equivalent of 0*03 normal serum* 99 " 1 99 99 0*045 " 99 " H 99 99 0*05 " 99 " 3 99 99 0*03 99 \#187 ; 5 99 99 003 " The complement increased by about 66 per cent , in l\#163 ; hours .
By this method we can demonstrate the occurrence of change in the quantity of complement contained in the immune serum , and can determine the time during which such change is occurring .
It is true that we supplement an " immune " complement by a " normal " complement ; and it is probable that partial complements do not exist in the same proportions in the two sera .
It is even possible that they are not identical , even when derived from animals of the same species , although there is no evidence of this , and such evidence as we have is against this view .
Even if it were so " however , the fact would not invalidate the method , in so far as it indicates the occurrence of change and the duratipn of that change .
The amount of 1907 .
] Complement-content of Fresh Blood-serum .
385 change which it shows to occur is , however , more open to question .
It is assumed that the immune-body undergoes no alteration in the process of heating , and this , probable as it may be , has not been proved .
Further , the existence of different proportions of partial complements in the two sera would affect the quantitative accuracy of the results obtained .
In any case , great stress should not be laid on particular figures , since the experimental error is quite appreciable , though remarkable regularity and accuracy are obtainable with haemolytic sera .
While we may rely , then , on the method showing us whether any change occurs , and the duration of such change if present , the amount of change indicated is probably only an approximate measure of what really occurs .
In example No. 5 given above , the amount of complement had risen by the end of one hour to double the amount first observed .
As a rule , the increase is rather less .
In some of the experiments ( 10 were made with immune sera ) no increase at all was found , the complement diminishing steadily from the first observation .
Further , in all the experiments , where any rise took place it was complete in one and a-half to two hours after the blood was drawn .
There is a difference , therefore , between normal and immune serum in this respect .
In normal serum we found the complement to increase by 100 to 200 per cent , and the duration of the increase to be from five to seven hours .
In immune serum the full amount of complement is often present from the first , and where an increase does occur , it is complete in a short time and is not so great as the increase in normal serum .
By the process of immunising , therefore , complement , or the source of complement , has been altered in such a way that it is developed in the serum with greater readiness than is the case with normal blood .
The repeated injections of red cells may have induced repeated new formations of complement , with the result that the leucocytes or other cells have been " educated , " in Metchnikoff 's sense , to secrete complement , or its precursor , more quickly than where such stimuli have not been given .
But here , again , explanation becomes pure hypothesis .
On the suggestion of Professor Muir , an experiment was made to see whether the change produced by the process of immunising affected all the haemolytic complements of the animal immunised or only those specially concerned in the laking of the red cells used for the immunisation .
A rabbit immunised against ox corpuscles was bled , and samples of serum withdrawn in the usual way .
Two series of experiments were then carried out .
In the first , ox corpuscles and the appropriate immune-body were used , and the variation of the complement determined by the method for immune serum .
386 Increase in the Complement-content of Fresh Blood-serum .
The rise was complete in one and a-half hours , and amounted to under 70 per cent. In the second series , guinea-pig corpuscles were used along with immune-body derived from a rabbit immunised to these cells .
Here the samples were treated as if they were samples of a normal serum , and the minimal dose giving total haemolysis with constant amounts of immune-body and corpuscles was determined for each .
The rise of complement was not complete after three hours , and amounted to more than 100 per cent. , i.e. , it followed the rule of variation of a normal serum .
The change produced by immunising with ox corpuscles in this case , therefore , did not affect all the complements laking guinea-pig corpuscles , but was confined to the ox-cell complements .
As only one such experiment was made , it is impossible to say whether this is a general rule or not , or in what degree the complements , which are able to take part in lysis of either ox or guinea-pig corpuscles , are affected .
Summary .
1 .
The serum first separating from shed blood may differ in composition from that separating later .
2 .
With normal blood of rabbit or guinea-pig , the amount of haemolytic complement contained in the first serum to separate is comparatively small , and progressively increases in serum separating later for a period of 5 to 7 hours in the case of rabbit , 3 to 4 hours in the case of guinea-pig , to reach a maximum 2 to 3 times as large as the amount first observed .
3 .
Even after removal from the clot , the amount of complement may increase in serum which separated soon after the blood was drawn , but this increase is slight .
4 .
This increased activity of serum is not peculiar to the haemolytic complements .
5 .
With the blood of immunised animals , immune-body does not increase in this manner .
6 .
With the blood of immunised animals , the haemolytic complements may increase in amount , but this increase is not always present , and when it does occur is less , and reaches its maximum more quickly than is the case with normal blood .
7 .
The alteration thus produced by the process of immunising probably affects only the haemolytic complements related to the immunising agent , and not other haemolytic complements which may be present in the serum .
A portion of this work was done at Queensbury Lodge , Aldenham , and I desire to express my thanks to the Directors of the Lister Institute of Preventive Medicine , and also to Dr. Dean for the facilities afforded me The Correlation of the Ovarian and Uterine .
387 there .
I am glad also to have this opportunity of thanking Professor Ritchie , of Oxford , for the interest he has taken in my work and for suggestions made in the course of it .
REFERENCES .
1 .
Ainley Walker , " On some Factors in Bacteriolytic Action , " 'Journal of Hygiene , vol. 3 , p. 52 , Jan. , 1903 .
2 .
Macdonald , G. G. , " On Immunity in Pneumococcal Infections , " in ' Studies in Pathology , 'edit .
Bulloch , Aberdeen , .1906 .
3 .
Ehrlich and Morgenroth , ' Berlin Klin .
Wochenschr .
, ' 1899 , No. 1 .
4 .
Mainwaring , 'Journal of Infectious Diseases , ' vol. 2 , pp. 446 and 460 .
5 .
Morgenroth and Sachs , ' Berlin Klin .
Wochenschr .
, ' 1902 , No. 5 .
The Correlation of the Ovarian and Uterine By E. S. Carmichael , M.B. , F.R.C.S.E. , and F. H. A. Marshall , M.A. , D.Sc .
( Communicated by Professor E. A. Schafer , F.R.S. Received April 10 , \#151 ; Read May 30 , 1907 .
) ( From the Physiological Laboratory , University of Edinburgh .
) A considerable body of evidence has been adduced in support of the hypothesis that the uterus is dependent upon ovarian influence for the maintenance of its structure and the discharge of its functions , and that the removal of this influence at any time during reproductive life causes the uterus to atrophy , besides producing a profound effect upon the entire metabolism of the organism .
Some writers , on the other hand ( e.g. , Blair Bell ) , contend that it is the uterus which is of prime importance , and observations have recently been cited pointing to the conclusion that this organ exercises a control over the growth and activity of the ovaries .
Others , again ( e.g. , Bond ) , adopt the view that the ovaries and uterus are to a very large extent inter-dependent upon each other .
The clinical evidence brought forward by surgeons after removal of the uterus and ovaries in the human subject has supported to a great extent the experimental evidence of the existence of such an inter-relation .
The modern tendency to conserve these organs as far as possible during surgical interference has been brought about largely by the belief in the dependence of the one organ upon the other .
Important work bearing upon this subject has been published recently by Zweifel ( 1899 ) , Mandl and Burger ( 1904 ) , Bond ( 1906 ) , and many other surgeons .
It is the purpose of this paper to record a series of experiments dealing

