Immunology, I958, I, 365.

Estimation of Tetanus Toxoid by Different Methods,

Including Haemagglutination Inhibition
A. J. FULTHORPE
The Wellcome Research Laboratories, Biological Division, Beckenham, Kent
Summary. Comparative quantitative estimations of tetanus toxoid by different
methods have been made. The in vivo total combining power test is considered
likely to give a true assessment of the combination of toxoid with antitoxin since
it is based on the biological action of toxin. The haernagglutination inhibition
test has been found to give good agreement with the in vivo test and it is cheap,
sensitive and readily repeatable. Since the haemagglutination inhibition test can
be performed with toxin or toxoid by the same method, it is possible to relate the
quantitative estimation of toxoid to the L + dose of a routine test toxin, and thus to
a standard antitoxin, which is convenient.
Considerable discrepancies have been found between results in the flocculation
test and the in vivo total combining power test, particularly with toxoids denatured
by heat and phenol, or modified by formalin. Discrepanicies between the floccula-
tion and in vivo tests with many preparations have been related to the concentration
of residual free formalin.
INTRODUCTION
IT is now generally accepted that the flocculation reaction with toxins and antitoxins
represents visible combination between several antigen-antibody systems. Carefil
examination of such systems by the gel diffusion technique for instance, has demonstrated
their complexity. The quantitative estimation of tetanus toxoid has usually been carried
out by the flocculation method, but this test in tetanus antigen-antibody systems is known
to be complicated by the appearance of several zones of flocculation at different points
at the same time. Specially selected sera must be used to obviate this difficulty. When
titrated with suitable sera however, tetanus toxoids show considerable variation in anti-
genic efficiency even when they have the same toxoid content as estimated by flocculation.
The possibility arises therefore that the flocculation test gives an incorrect estimate of the
toxoid content of a preparation.
For diphtheria toxoids Barr and Glenny (I949) showed that an in vivo total combining
power method of estimating toxoid was a better indication of antigenic efficiency than the
Lf dose.
There are many objections to the total combining power test. The values obtained by
this method are affected by alterations in the time allowed for antigen-antibody combina-
tion, since more toxoid combines with antitoxin with time, and toxin has the ability to
turn out toxoid from combination with antitoxin and this effect increases with time.
It is therefore advisable in comparative work always to use the same antitoxin and test
toxin since the results can also be affected by the avidity of the antitoxin and the toxicity
of the toxin (Barr, Glenny and Stevens, I954). The test cannot therefore be used to com-
pare preparations containing a variable ratio of toxin and toxoid. It is in fact comparative
only when carried out under strictly standard conditions, and the results obtained are
365
366 A. 1. Fulthorpe
in no way absolute and will vary from one laboratory to another. The in vivo total
combining power test also has the disadvantage that it is time-consuming and that many
tests must be carried out in order to get a satisfactory assessment of the end point, which is
always liable to be rather diffuse, and this limits the application of the method for routine
use. The total combining power test has however one signal advantage, some of the test
animals undoubtedly die with symptoms of tetanus, and it is therefore reasonable to sup-
pose that, properly handled, it gives a measure of combination of toxoid with antitoxin,
and not some other unknown antigen-antibody reaction.
It is possible to estimate the volume of a toxoid necessary to inhibit agglutination of
toxoid sensitized sheep cells by a fixed dose of antitoxin, i.e. a simple haemagglutination
inhibition test. This test in common with the flocculation test can be used to estimate the
specific content of a preparation of toxin, toxoid or a combination of the two, by the same
technique. However, the haemagglutination inhibition test has the same disadvantage
as the in vivo total combining power test in that conditions of the test with regard to the
time and temperature of combination of toxoid and antitoxin must be kept constant, as
must the sensitivity of the cell suspension. The test is, however, quick, cheap and readily
repeatable and the end-point is comparatively sharp.
An investigation of the antitoxin combining power of tetanus toxoids of different types
has been undertaken by flocculation, in vivo total combining power and haemagglutination
inhibition. The aim of the survey was to determine the most satisfactory method for
preliminary assay of tetanus toxoid and if possible determine the cause of discrepancies
between the different methods.
METHODS
THE FLOCCULATION TEST

Constant volumes of a suitable dilution of the toxoid under test were pipetted out into
a series of 3/8 X 3 cm. round-bottomed tubes. A suitable working standard antitoxin
(23 units/ml.) was added in volumes differing by io per cent. The tubes were incubated
at 500. and inspected at intervals. Estimation of the toxoid content of a preparation
was based on the principle that the mixture which flocculates most rapidly contains an
equal number of Lf doses of toxoid and of in vitro units of antibody. Any figures given for
toxoid content, as determined by flocculation, have been obtained in relation to the
provisional in vitro unit in use in these Laboratories.

THE in vivo TOTAL COMBINING POWER TEST

Volumes of the toxoid under test, suitably diluted to contain about i unit equivalent
of toxoid and differing by Io or 20 per cent, were measured in a series of tubes and
constant volumes of standard antitoxin containing 2 units added; the mixtures were gently
shaken and allowed to stand for one hour. Constant volumes of a test toxin equivalent
to I unit of standard antitoxin were then added and mixed and allowed to stand for half an
hour. The whole of the triple mixtures (restricted to a reasonable total volume) was then
injected into mice. The mice were observed twice daily and the end-point of the in vivo
test was assessed on a basis of death of the test animal on the fourth day or definite symp-
toms of tetanus on the third day. If an animal was dead on the second or third day and
the next animal in the series was alive, the end-point was arrived at by interpolation.
 Estimation of Tetanus Toxoid 367
Several estimations were made on each preparation and the value allotted was the arith-
metic mean ofindividual results. It will be noted in Table 2 that the actual dose of toxoid
equivalent to I unit, as determined by this method, is lower than those obtained by
other methods. This fact merely illustrates the point that total combining power tests
are only comparative.

THE HAEMAGGLUTINATION INHIBITION TEST

Graduated volumes of the toxoid under test differing by IO or 20 per cent were mixed
with o I unit of working standard antitoxin, in 3/8 X 3 cm. round-bottomed tubes, allowed
to stand for one hour and o-i ml. of a toxoid-sensitized suspension of sheep cells added,
dispersed through the mixture and the test allowed to stand overnight at room temperature.
The tests were read on the following day, and the end-point calculated by a standard
method (Fulthorpe, I957).

PREPARATION OF SENSITIZED SHEEP CELLS

Washed sheep erythrocytes were treated with tannic acid and sensitized with purified
tetanus toxoid by Stavitsky's method (1954). After sensitization the cells were preserved
with formalin (Fulthorpe, I957). All cells were sensitized with the same batch of toxoid,
the suspension diluted to a. standard opacity, and only those preparations with the same
degree of sensitivity used.
@: 00
0
0
1000 . ; 1000
0-

0~~~~~~~~~~0
0

C
.00 -
0~~~~~~~~~@
EIOO.

C
0~~~~~~0

10 , 10

10 100 1000 10 100 1000

Unit equivalents by flocculation
FIG. I. Estimation of tetanus toxoid, content (unit
equivalents/ml.) of several preparations by the in iwo
total combining power method plotted against the
values obtained for these materials by flocculation
(Lf/ml.), log scale.
Unit equivalents by haemaqqlutination
FIG. 2. Estimation of tetanus toxoid content (unit
equivalents/ml.) of several preparations by the in viw
total combining power method plotted against the
values obtained for these materials by haemagglutina-
tion inhibition (L./ml.), log scale.

RESULTS
COMPARISON OF DIFFERENT METHODS OF TOXOID TITRATION

A number of tetanus toxoids prepared by different methods and of widely different

toxoid content were titrated by flocculation with a suitable antitoxin not given to producing
VOL. I. 4.
5
368 A.J. Fulthorpe
multiple zones. The same toxoids were also titrated by the total combining power test
and the two sets of results plotted (Fig. i). While the general correlation between the
in vivo and flocculation results is obvious there are considerable discrepancies on this very
wide scale.
The same group of toxoids were then titrated by haemagglutination inhibition, using
cells sensitized with a purified tetanus toxoid and a volume of serum containing o-I unit
of antitoxin (LA/IO level of test). When these results were compared graphically with the
results of the total combining power test the agreement was found to be far better than
with the flocculation method (Fig. 2). These two sets ofresults suggest that the flocculation
method had produced many discrepancies which were not due to the development of
multiple zones since these were looked for.

FORMALIN TREATMENT
A toxin was treated with formalin and samples taken at intervals and tested for toxicity,
and the combining power estimated by flocculation and haemagglutination inhibition.
The values obtained did not change as toxoiding proceeded, at least for a time (Table i).
No in vivo titration could be carried out on these samples since the ratio of toxin to toxoid

TABLE I
EFFECT OF PROGRESSIVE TOXOIDING ON COMBINING POWER
ESTIMATIONS BY FLOCCULATION AND IIAEMAGGLUTINATION
IIBITION

M.L.D. (ml.)
Floccua
Lf dose (ml.)
HaTmagglutination
LA/io dos (ml.)
3 I
ustte I/IO
io
Toxin o-oooooo, o-oo87 OOOI2
0,0012
0-000005 0-0093
00001 o-o087 00012
0-0005 00090 0-00I3
0-005 0-0090 00012
0-05 0-0090 000I3
Toxoid >0-5 00090 00012

* Provisional flocculating unit of antitoxin.

was changing progressively. Since the haemagglutination inhibition test for toxoid will
estimate toxin or toxoid or mixtures of the two, the test dose so determined can be related
to- the L + dose of a routine test toxin in conjunction with a standard antitoxin.

PROLONGED FORMALIN TREATMENT

An autolysate toxin was divided into two portions and one portion was treated with
formalin. Both were left for a period of twelve months in the cold room. After this time it
was found that the flocculation value of the toxin was the same as that of the toxoid pre-
pared from it, but that the value of the toxoid by haemagglutination inhibition was half
that of the parent toxin. In addition a batch of autolysate toxoid was divided up and
 Estimation of Tetanus Toxoid 369
part thoroughly dialysed and then ultrafiltered down to its original volume. Both the
dialysed and undialysed samples had the same flocculation value after eighteen months
but the combining power value in vivo of the undialysed portion had fallen to half that
of the dialysed portion. These observations suggested that prolonged contact with formalin
may be responsible for some of the changes in combining power ratios.

FORMALIN IN DIFFERENT CONCENTRATIONS

An ultrafiltered tetanus toxin was treated with four different concentrations of formalin
at 370 for a period of eighteen days and the material tested by flocculation, total
combining power and haemagglutination inhibition at the end of this time, when the
material was non-toxic to mice in a dose of o-5 ml. The four samples had the same
flocculation value, but the toxoid content by the other methods was found to be lower
TALE 2
EFFECT OF FORMALIN ON TOXOID INCUBATED AT 370 FOR I8 DAYS

Test dose estimations

Flocculation Haemagglutination In vivo
Lf dose (ml.) LA/Io dose (ml.) T.C.P. dose (ml.)
= I unlit*
1/IO Petit I tuit
Toxin .. 0-0O93 0OOI2

Toxoid:
rO-I5% 00I0 OOOI2 o0oo45
03%
0 oIOOI3
0010 o0oo46
o-6% o0oIO O-OOI5 o oo6
, I*2% OIo o ooi8 o0oo7
* Provisional flocculating unit of antitoxin.

when higher concentrations of formalin had been used (Table 2). The fall in combining
power as measured by methods other than flocculation suggests that formalin has a de-
naturing effect on toxoid, and that this denaturation process does not involve other
antigen components entering into the flocculation reaction. While the Lf dose was the
same for all samples, the LA/IO dose was 40 per cent higher and the in vivo T.C.P. dose
50 per cent higher for the last sample than the first.

DENATURATlON BY HEAT AND PHENOL

Progressive denaturation of a partially purified toxoid was carried out by heating for
different periods at 500 in the presence of 0-5 per cent phenol, to see if denaturation by a
standard method would result in divergence of flocculation and other combining power
values. The results of flocculation tests on the successive samples showed that the toxoid
content of the last sample was about half that of the unheated control, but the haemag-
glutination values of the last sample fell to one-fifteenth of the control value (Fig. 3).
Since it seemed possible that this result might be due to an error in the estimation by
haemagglutination the samples were retitrated in vivo by the total combining power test.
The results were strictly comparable with those obtained by haemagglutination inhibition
(Fig. 3).
VOL. I. 4. 5*
370 A. J. Fulthorpe
COMBINING POWER RATIOS AND PURITY
A comparison of the total combining power values to the flocculation values of a number
of different types of toxoid showed that the ratios tended to be greatest when the material
tested was of the highest purity, and vice versa. Since the flocculation results were of
doubtful value, the purity of the preparations was calculated on a basis of unit equivalents
of toxoid in vivo per mg. protein nitrogen (Fig. 4).

0
0\. * p

E
50
A
A o~A00A 0
0 o 2-0 0 P P

P P
0 * * IA A A A 0 ~ ~ ~ ~ 0

W- ~0 A 1-5

._
.0~~~ 0
0
0
o e

0 10
0 ** 0

0 M,1
0 .0

40
0 *
*
Vs
0 0

CT
07i

2 0
1O 3 6 12 24 48 96 192
Dencturation time in hours
FIG. 3. Estimation of tetanus toxoid content (unit
equivalents/ml.) of samples of toxoid progressively
denatured by heating at 50° in the presence of
0-5 per cent phenol. Toxoid content estimated by
(a) flocculation-A*; (b) haemagglutination inhibi-
tion-*; and (c) by the in vivo total combining power
method-O; (log scale). The arrow indicates the
points obtained for the unheated control by the
different methods.
30 100 300 1000
Purity
FIG. 4. Correlation of combining power ratios with
purity. The combining power ratios have been
calculated from estimations of toxoid content by
the in vivo total combining power method (unit
equivalents/ml.) divided by the values obtained by
flocculation (Lf/ml.).
The purity of the preparations has been calculated
in terms of unit equivalents of toxoid by the in vivo
total combining power test per mg. protein
nitrogen. The materials consists of (a) filtrate toxoids
-0; (b) chemically purified filtrate toxoids-P; (c)
autolysate toxoids-O; and (d) purified toxoids
denatured by heat and phenol-A.

The actual ratios were found to range from o-8 to 2-2, a considerable difference which
means virtually that when injecting animals with I Lf of purified toxoid, the quantity of
toxoid injected may be twice that represented by Lf of a cruder preparation.
i

There appear Fig. 4. One has a gradual slope, comprising

the results obtained with filtrate toxoids and chemically purified fractions obtained from
filtrate toxoids. The other which has a much steeper slope represents results with autoly-
sate toxoids and purified material denatured to different degrees by heat and phenol.
Filtrate toxoids are prepared from day broth cultures of the organism, filtered to
io-i i

remove bacterial debris. Such toxoids contain a large amount of non-specific nitrogenous
material, with all the products of bacterial metabolism. Autolysate toxoids are prepared
from broth cultures grown for 2-3 days, at which time little free toxin is present, the organ-
isms are separated and then autolysed with molar salt solution; this permits toxin contained
 Estimation of Tetanus Toxoid 371
in the organism to diffuse out. Autolysate preparations contain proportionately less
non-specific material and start with comparatively good purity because the culture medium
and products of metabolism are discarded before autolysis.
It is suggested that the differences obtained by flocculation and in vivo titration and the
relation of these differences to purity result from two factors. Firstly toxoid is susceptible
to denaturation if left in contact with free formalin, and secondly, toxoid preparations
contain certain other antigens not so susceptible and capable of flocculating with their
respective antibodies.
When toxoid in any formalized filtrate is denatured the toxoid protein is still recogniz-
able chemically as protein, so that the apparent purity of the preparation is reduced by
an amount proportional to the degree of denaturation. This tends to be related to a change
in the ratio of the in vivo values to the flocculation values for the toxoid, since the denatura-

20

1.5 * FIG. 5. Correlation of combining power ratios

00 with percentage residual free formaldehyde in
the preparations. The combining power ratios
O have been calculated from estimations of toxoid
*< content by the in viw total combining power
cX
1-0
o
10
* 0
*
K method (unit equivalents/ml.) divided by the
values obtained by flocculation (Lf/ml.). Esti-
mations of residual (uncombined) formaldehyde
in the preparation were made in terms of
* gm./foo ml.
0*7 07 _ *~~
02 0.1 0-05 0025
Percentaqe Formaldehyde
tion is not correctly assessed by flocculation. The progressive change in combining power
ratios as a result of denaturation, with a corresponding fall in the apparent purity, can
be most readily seen in respect of the purified material denatured by heat and phenol.
The difference in slope of the two association curves may be explained by the different
purity of the starting material of the two types of toxoid. The autolysate material starts
with a relatively high purity and even after serious denaturation, enough to cause a
considerable change in the combining power ratios, the material still retains a compara-
tively high value per mgm. protein nitrogen. It may be also that autolysate material
contains relatively more of the heat- and formalin-stable flocculating material.
The residual free formalin after toxoiding and storage has been found to be greatest
in autolysate toxoids possibly because they do not contain so much non-specific material
capable of combining with it. For this reason denaturation might be expected to be
greatest with autolysates and it was in fact such material which produced the lowest
ratios between the in vivo and flocculation results.

RESIDUAL FREE FORMALIN

The combining power ratios of a number of toxoids were plotted against the results
of chemical tests for free formaldehyde (Fig. 5). A rough correlation can be seen to exist
even though the period of storage of these materials in the presence of free formalin varied.
372 A. 3. Fulthorpe
This again suggests that the discrepancies obtained in the flocculation test are due to
a differential denaturation of toxoid, in which some flocculating component of these
preparations does not take part.
DISCUSSION
Much work has been done on the complex nature of bacterial filtrates, some of which
are known to contain many toxins, with recognizable biological properties, such as the
toxic filtrates of Clostridium welchii. Other toxic filtrates however, such as those obtained
from Corynebacterium dipktheriae, even after considerable chemical fractionation may
still contain many antigens lacking obvious biological characteristics but capable of
producing visible precipitates with antibody (Pope, I957). Controlled digestion of
crystalline diphtheria toxin with trypsin gives rise to a considerable change in the in
vivo/in vitro combining power ratios with eventually the production of additional floccu-
lation zones (Pope, I957), and progressive denaturation with tribasic phosphate results
in the destruction of biological properties with an actual increase in the flocculation
value (Pope and Stevens, 1958).
In the course of this investigation it was evident that denaturation of tetanus toxoid
by heat and phenol and modification by formalin had caused a similar divergence of
flocculation and total combining power values. The results of the haemagglutination
inhibition method of toxoid estimation were in general agreement with those of the total
combining power method, when titrating such material.
The variations revealed in the combining power ratios of different types of preparation
underline the necessity for a satisfactory method of quantitative estimation of toxoid.
The total combining power method is probably used in many laboratories to ensure that
the values obtained by flocculation are not due to the presence of a false zone. The method
has many disadvantages however and for routine use the haemagglutination inhibition
technique appears to be more convenient.
It is essential that combined prophylactics should have a proper balance of antigens
to produce an adequate response to all components, and this is difficult without a satis-
factory quantitative measure of total toxoid in a preparation.
It is not suggested that formalin denaturation is entirely responsible for the qualitative
difference between toxoids. Non-specific antigens, that is antigens other than tetanolysin
and tetanospasmin may well affect the results of antigenicity tests. It seems advisable
however to assume that this denaturation is a disadvantage since the modified protein,
although lacking specificity, may still evoke a useless antibody response which competes
with antitoxin production. For this reason it seems advisable to limit formalin treatment
to conditions consistent with safety and remove excess formalin as soon after toxoiding as
possible.
REFERENCES
BmAR, M. and GLENNY, A. T. (1949). 'Qualitative POPE, C. G. and STEVENS, M. F. (1958). 'The proper-
differences among toxins and toxoids.' J. Hyg., ties of crystalline diphtheria toxin protein: its
Camb., 47, 107-I14- antigenic composition as determined by gel diffusion
BARR, M., GLENNY, A. T. and STEVENS, M. F. (1954). methods.' Brit. J. exp. Path., 39, 150-157.
'The combination of antitoxin with toxin and STAVTs5KY, A. B. (I954). 'Micro methods for the study
toxoid.' 3. Hyg., Camb., 52, 379-396. of proteins and antibodies.' . Immunol., 72, 36o-
FULTHORPE, A. J. (I957). 'Tetanus antitoxin titration 367, 368-375.
by haemagglutination.' 3. Hyg., Camb., 55, 382-401.
PoPE,C.G. (I 957).'Observations on the diphtheria toxin-
antitoxin reaction.' Brt.j3. exp. Path., 38, 207-2 i6.