Defensive Ferments of the Animal Organism/Methods in Use/Sources of Error in the Dialysation Process

by Emil Abderhalden

Sources of Error in the Dialysation Process.

There are many possibilities leading to erroneous results. It is best to consider them from the point of view of utensils employed and manipulations adopted, and to refer again to the sources of error mentioned in the description of the method.

(i) Tubes.—We take it for granted that all tubes are thoroughly and accurately tested before anything else. On the average about 20 to 30 per cent, of the dialysing tubes supplied by the firm Schleicher and Schüll will be useless, because there are nearly always some which allow the passage of albumen.^[1] Or they may become useless subsequently, generally becoming permeable to albumen. This, however, only occurs when they are handled improperly. They must not be cleaned with a rough brush, nor must they be boiled for too long a time. Tubes may become impermeable to peptones through over-boiling, so that, though they should be thoroughly washed, they should be boiled but slightly. They must be kept in sterilized water with a thick layer of toluol (see p. 162), and must never be left for a long time unemptied of their contents.

A great source of error which is, however, impossible with proper manipulation, is due to tubes being insufficiently cleaned. The result of this is, that the wall of the tubes will contain traces of substances, which react with ninhydrin if sufficiently concentrated. They may be present in such minute quantities as to be unable of themselves to produce a coloration; yet they will, when added to the analogous substances that are present in the serum, convert a negative reaction into a positive one. Therefore the utmost possible care must be exercised in the manipulation of the tubes.

Tubes must be tested again about every four weeks. Should any error in diagnosis have occurred before this time, and other possible errors have been excluded, then the tubes must be immediately tested for permeability to albumen and for even permeability to peptones.

(2) Serum.—Here we have to deal only with its age, the possibility of an infection, hæmolysis, and the contents of the serum in respect of red blood corpuscles and of other form-elements. (See pp. 173-175-)

(3) The Organ.—This is nearly always the cause of errors in diagnosis. It is nearly always forgotten that, in the arrangement and execution of the experiment, we are dealing with quantitative conditions. Two cases have to be distinguished:—

(a) The Biuret Reaction.—The serum alone does not give off substances which diffuse and produce a biuret reaction; so that, as regards compounds which give a biuret reaction, it must be reckoned as completely indifferent. It is comparatively easy to boil the organ in such a way, that the water in which it was boiled will not give any biuret reaction. If the ninhydrin reaction turns out negative, one can never obtain a biuret reaction. If such an organ be mixed with serum, and the dialysate now gives a positive biuret reaction, then we may be sure that decomposition has taken place. The conditions here are very simple.

(b) The Ninhydrin Reaction.—In order to understand the propositions that follow, we must be clear concerning the fact, that blood serum always contains, in varying quantities, substances which are to be found within the peptone group, and therefore react with ninhydrin. After a meal at which albumen has been taken, the quantity of such substances appearing in the serum immediately increases, in consequence of which the blood must be taken during a state of hunger.

A great many experiments have been necessary to determine what quantity of serum, in general, will give off to the dialysate only so much of the substances referred to, as is required for a negative reaction with ninhydrin. An insufficient quantity of serum must not be used, if the decomposition of the organ's albumen is to be as complete as possible. It has been found that, in general, 1.5 c.c. of the serum may be used. It is obvious that, under certain circumstances, an even greater quantity of serum may give off so few substances reacting with ninhydrin that the reaction of the dialysate still remains negative. Conversely it may happen that 1.5 c.c. of serum alone will give a positively reacting dialysate, which is the reason why a control test with serum alone is absolutely essential. The latter test indicates whether the serum in use answers the condition of not giving off, of itself, a sufficiency of substances to react with ninhydrin. It is obvious, for the reasons mentioned, that exactly the same quantity of serum must be added to the organ, as has been used for the control test with serum alone. We must never, on the strength of the fact that the test with the serum alone gives a positive reaction, jump to the conclusion that, during the test, proteins have been decomposed in the serum. The substances producing this reaction were present from the beginning. If the reaction with serum alone turns out negative, then it simply means that the dialysate contains those compounds, which react with ninhydrin, in a state of concentration insufficient to produce a coloration; and this is the only conclusion we are entitled to draw from the result. It certainly does not indicate that there are no such compounds present. If one concentrates a dialysate of this kind, it eventually gives a positive reaction.

We therefore arrive at the fact, that we can only determine whether there are sufficient compounds present to give the coloration, but not what the quantities actually are. If, however, the following conditions are complied with, then this circumstance offers no difficulties. The organ must be absolutely free from substances, reacting with ninhydrin, which can be boiled out and so passed over to the filtrate. When the tubes are rinsed, no water should be allowed to enter them. The organ must be perfectly dried, before the tube is filled with it. During storage in the incubator no evaporation must take place. Further, when boiling the actual samples, uneven ebullition must not be allowed. An example may help us to make these conditions clear. We will assume that twelve experiments have been made with serum obtained from non-pregnant individuals, and that the serum has in every case given negative results. We conclude that none of the dialysates have attained the necessary concentration in compounds that produce coloration with ninhydrin. Only from a certain concentration onwards is the coloration possible; and this limit we designate by the number 1. Then, to give an example, the cases quoted in the annexed table are possible:—

Case	Test with serum alone. Ninhydrin test.	Contents in the serum of compounds which, with ninhydrin at a sufficient concentration, react so as to produce a coloration	Test with organ+serum organ = 0. Ninhydrin test.	Test with organ+serum organ = 0.10. Ninhydrin test.	Test with organ+serum organ = 0.50. Ninhydrin test
01	-	0.12	-	-	-
02	-	0.45	-	-	-
03	-	0.84	-	-	+
04	-	0.65	-	-	+
05	-	0.89	-	-	+
06	-	0.98	-	+	+
07	-	0.87	-	-	+
08	-	0.99	-	+	+
09	-	0.42	-	-	-
10	-	0.86	-	-	+
11	-	0.78	-	-	+
12	-	0.75	-	-	+

Three series of experiments were conducted with the same sera, and with equal quantities of these. In the first experiment the organ was = 0, i.e., it was absolutely free from substances which could be boiled out and filtered, and which, under the strictest conditions, would produce a coloration with ninhydrin. In every case we had to add to the quantity of substances emanating from serum alone, and passed into the dialysate, 0 gr. of these compounds. Then, in the experiment serum + organ, the ninhydrin reaction obviously remains negative.

For the second experiment an organ was taken, which passed over to the boiled water just a trace of reacting substances. We will assume that it contained 0.10 gr.^[2] of these compounds. This quantity is added to that which the serum gives off, and we have the positive reaction of Cases 6 and 8. The limital value, 1, has been exceeded. Thus, by means of a simple addition, a positive reaction has been obtained and, in consequence, two errors in diagnosis. The third column shows us how the ninhydrin reaction results, when we use an organ prepared in a still more imperfect manner.

Exactly the same position is reached, if the dialysate evaporates unevenly in the incubator. Take, for instance, Cases 6 and 8. In both cases the serum alone nearly reaches the limit, 1. Then, should the dialysate, in the experiment organ + serum, evaporate more strongly, or should the corresponding dialysate become more strongly concentrated, during boiling, than that of the relative control experiment, then we shall get a positive reaction owing, entirely, to the concentration; in which case we shall get a wrong result. These examples may be a warning to those who make use of a particular technique in an imperfect manner.

It is easy to understand that errors in diagnosis have often occurred, and that, on the other hand, brilliant results have been reported.

As a matter of fact our limit value, 1, is seldom attained. Unfortunately, this occurs just when carcinoma, myoma, salpingitis, exudates, suppurations, or hæmorrhages are present, that is, just when the method should diagnostically give the most valuable differential results. It is obvious that the investigation of such cases requires double care.

The performance, under absolutely equal conditions, of a particular experiment, and its control, is of decisive importance in regard to the results obtained. In the first place, absolutely pure distilled water must be used. Water, which gives an acid or alkaline reaction, leads inevitably to erroneous results. Ninhydrin reacts not only with albumen and albuminous decomposites, but under certain conditions with other compounds as well, for instance, sugar.^[3] No trouble can be caused by these, if distilled water be used. The organ cannot give off any non-albuminous substances, which will affect the reaction of the fluid, if it is boiled in the manner prescribed. There could not possibly be any carbohydrates left, and we have the control test with serum to fall back on, in any case. Were this to contain much sugar, and, in consequence, to interfere with the reaction of the outer fluid, then it is conceivable that a coloration might take place, which could not be referred to albuminous decomposites. This result, however, would appear in the test with serum alone, and also in the one with serum substrate. Even the blood serum from cases of diabetes does not show any positive reaction ascribable to the presence of sugar. Non-compliance with the directions respecting water generally manifests itself in the fact, that a really positive reaction turns out negative; the reaction being, in fact, very sensitive towards acids and alkalies, i.e., towards H and OH ions.

For the reasons laid down we must always boil the organs in distilled water, and preserve them, as well as the tubes, in this medium. The rinsing of the dialysing tubes must also be done with distilled water.

Finally, we must bear in mind another source of error, which we have not yet specially referred to. It may sometimes happen, that the substrate added to the serum absorbs some constituents of the latter, and retains them. Such a case would manifest itself in the fact, that the serum alone would react positively, while the dialysate of the experiment, organ + substrate, would give a negative reaction. Further, a reaction might give a negative result, although decomposition had actually taken place. The optical method would easily detect such sources of error.

There is no single point in the rules which lacks a definite foundation. Researches have generally been wrecked owing to trifling details. A glance at the literature, however, shows that at present the method is properly used in many places, and leads to surprisingly beautiful results.

Further sources of error are: The use of vessels that are not dry, and of boiling-sticks that have been touched by the hands, soiling the pipettes with saliva, inaccurate measurement of the ninhydrin solution, the use of infected water, the cultivation of bacteria in the same incubator as is used for experiments on the action of ferments, covering the contents of the tubes, and the outside fluid, with an insufficient layer of toluol, changes of temperature in the incubator, and working in places where acid or alkaline vapours are developed.

These sources of error ought, properly, to occur very seldom.

On the other hand, the following point is often overlooked. After the tube has been filled with the organ, and the serum and the toluol have been added, it is absolutely necessary to make sure that the whole of the tissue is covered with the serum and toluol. Should the slightest portion of the tissue project above the toluol, it is then liable to decay in the course of sixteen hours, and so become a source of serious error.

In conclusion, we will add the following supplementary details, which are not at present in general use, because they are not considered as being absolutely necessary. We may, instead of the control with serum, use a control with organ + inactivated serum. The serum is heated for thirty minutes at a temperature of 60° C. This kind of control is capable of indicating an insufficiently prepared organ.

Starting with the idea that a certain limital value must be present, in order to give a colour reaction with ninhydrin, one might conclude that it would not suffice to test the filtered water, in which the organ was boiled, with 1 c.c. of ninhydrin solution. We have therefore produced a solution of silk-peptone, which has been so strongly diluted, that 5 c.c. of the solution just fails to show any coloration with 1 c.c. of ninhydrin solution. 2.5 c.c. of this solution were then added to 2.5 c.c. of the filtrate obtained from the water in which the organ was boiled, and 2 c.c. of the ninhydrin solution were added to this. The mixture was boiled in the usual way for one minute, and the reaction remained negative. It would always have been possible for the limital value to be attained by means of additions. Further, a volume of 10 c.c. was reduced to 5 c.c. After the addition of 1 c.c., and later of 2 c.c. of the ninhydrin solution, no coloration appeared.

Finally, we may once more insist on the fact that an organ containing blood frequently fails to act, even when it fully complies with the conditions with reference to the water, in which the organ has been boiled (see pp. 164-168).

A desire has often been expressed, that we might have a special colour-scale for estimating the results of the ninhydrin reaction, with a view to recording the strengths of the reaction in a generally equivalent manner; but this cannot well be effected, because the ninhydrin reaction does not allow of sharp delimitation. With experience, each observer will soon be able to judge whether the reaction is strong, medium, slight, or very slight. Besides, we must not lay too much stress upon the intensity of the reaction. It is quite possible, for instance, that in any given case a quantity of highly molecular peptones is present in the dialysate. The biuret reaction is surprisingly strong, while the ninhydrin reaction, on the contrary, is very weak. Conversely, we can imagine the extreme case, in which the decomposition lies below the peptone limit. We obtain a deep blue ninhydrin reaction, which means that many compounds, having the structure of amino-acids, are present; whilst the biuret reaction gives a negative result. These facts show clearly, that the ninhydrin reaction enables us to recognize many more compounds of the group of albumen decomposites than the biuret reaction.

Certainly many points in the whole method of the dialysation process might be modified. In the first place, the apparatus used could be improved. One might consider, for instance, the possibility of constructing an apparatus, which would enable us to boil the solutions of the ninhydrin reaction simultaneously and equally, and at the same time to prevent any evaporation. We have purposely made no propositions in this direction, because it seemed to us, that the great advantage of the present method is just that it is simple, clear and concise. We have also made experiments for simplifying the preparation of the organs, and more particularly for shortening that process. Studies on organs, that had been dried and pulverized at 37° C. with special precautions, gave good results, but the risk of infection is great. In any case, organs prepared in this way have also to be tested each time before use. The boiling process has this advantage over the other, that the tissues are loosened, and in this way are more easily acted upon by the ferment.

Layout 2

1. We have recently observed up to 80 per cent. of useless tubes. It would be very desirable if a dialysing tube could be produced which was, at the least, indubitably impermeable to albumen.
2. We take this here merely by way of an example. Obviously, in actual tests, the same quantity, i.e., 0.10 gr., would never be transferred to the dialysate, if the organ can only give off that amount; some lesser quantity would pass over.
3. Vgl. W. Halle, E. Loewenstein und E. Pribram: "Bemerkungen über Farbreaktionen des Triketohydrindenhydrats (Ninhydrin)," Biochem. Zeitschr., lv, 357. 1913.