Normally, all mammals, for a short or longer length of time after birth, feed exclusively on their mother’s milk.
Evidently, breast milk contains everything that is essential for the newborn to thrive.
When, in 1909, Wilhelm Stepp in Strassburg
and then Frederick Gowland Hopkins in England tried to raise young mice and young rats
on a diet consisting of highly purified protein, fat, carbohydrate and mineral salts, they found it did not work.
The animals died.
By contrast, the animals developed well when milk was added to this diet.
That is more or less what you might expect.
But the experiments also led to an unexpected observation, namely that even “astonishingly small amounts of milk”,
as Hopkins put it, were sufficient to supply the animals with what he himself called accessory nutrients,
but which soon became known as vitamins.
Milk can serve as an example to describe and explain nearly everything we have since learned about these vital substances:
The distinction between fat-soluble vitamins, which separate into the cream during centrifugation,
and water-soluble vitamins, which remain in the whey; methods of biological and chemical determination;
enrichment and purification techniques, structural elucidation, and finally the synthesis of vitamins,
their efficacy in treating deficiency disorders and the protection they offer against these and many others.
Two vitamins in milk can be directly perceived, as they are pigments: Firstly, carotene,
which separates out into the cream and is responsible for the yellow colour of butter;
secondly, lactoflavin, which is water-soluble and gives whey its greenish-yellow colour.
It is not my intention here to enumerate all the vitamins so far identified in milk individually
or to discuss, with the help of long tables of figures, the differences in milk from various species,
differences in vitamin content depending on diet, seasonal fluctuations or technical dairy problems.
I wish to attempt, by drawing on just a few scientific findings arising from problems with milk,
to show how they have gained relevance to other questions in chemistry and medicine.
Two examples are particularly close to my heart,
carotene, lactoflavin and a still uncharacterised factor found in human breast milk.
Carotene was crystallised as early as 1831 by Wackenroder, not from butter but from carrots.
However, it was not until 1928 that Hans von Euler in Stockholm showed it to be an essential raw substance,
constituting the crucial factor in the fat-soluble phase in the pioneering experiments by Stepp and Hopkins.
In an experiment with Edgar Lederer
it was then discovered in 1931 that carotene can be separated into two clearly defined components.
They are known as alpha and beta carotene.
Together with Hans Brockmann, gamma carotene was later added.
The method used for separating carotene into its constituent part is called chromatographic absorption analysis.
And the extraordinary power of this method is reflected in the fact that it separated three isomeric carbohydrates,
all of which have the same molecular formula: C40H56.
The principle of chromatography was described by Schönbein in 1861 and in another form by the Russian botanist Zwet in 1906,
but it had only been used occasionally as a qualitative tool for analytical observations.
The separation of carotene into its constituents showed
that the method is suitable for preparing organic compounds in their pure state.
Since then, there has grown an almost bewildering array of applications for this tool.
From my own institute, in particular Alfred Winterstein, Hans Brockmann
and then Theodor Wieland made important methodological contributions.
The technique has also been used on fluorescing compounds, on colourless compounds, on liquids and on gases.
It has also found use in inorganic chemistry, for example, the separation of rare earth elements,
a problem to which many outstanding chemists had devoted decades of their lives,
which can now often be accomplished relatively easily by chromatography.
And the same applies to the separation of uranium.
And the principle of the method was specifically refined during World War II in England by Consden and Martin
in the form of paper chromatographic analysis,
which now appears to be irreplaceable for many problems in chemistry as well as biology and medicine.
The question has occasionally been asked how such a simple and versatile method could be ignored for so long.
Specifically, why Richard Willstädter, in his investigations of chlorophyll, did not use Zwet’s method,
who had already described the separation of green pigment from leaf extracts into two components.
Let me answer this question as follows.
Willstädter was familiar with Zwet’s experiments,
and he carried out the separation of chlorophyll into its A and B components according to Zwet’s description.
And yet he rejected the method.
The reason is that during the absorption of chlorophyll on silicic acid,
on calcium carbonate, on aluminium oxide and on other adsorbants that Zwet had recommended,
chlorophyll A and chlorophyll B undergo changes of a chemical nature that Zwet had not noticed, but which Willstädter did.
During adsorption, the chlorophylls lost the property of turning brown after the addition of alkali.
They lost what is known as the brown phase, and, for this reason, Willstädter believed
that the method was unsuitable for isolating natural pigments.
Among the numerous adsorbants Zwet recommended was, however, sucrose,
on which, as Winterstein later showed, both chlorophylls remain intact.
But sucrose, of all the substances, appears to have been ignored in the investigation by Willstädter’s laboratory.
Of all the vitamins that occur in milk,
probably only one was first actually isolated in its pure form from milk, namely lactoflavin.
It occurs in nature, partly in a free state, dialyzable, and fluoresces yellowish-green,
partly bound to proteins in an undialyzable and nonfluorescent form known as flavoproteins.
Studies carried out jointly with Paul György and Theodor Wagner-Jauregg showed
that lactoflavin has the same vitamin activity whether it is bound to protein or not,
the amount of lactoflavin required to achieve one growth unit in young rats,
that is, a weight gain of 40 grams in 30 days, is 10 grams lactoflavin.
For lactoflavin later synthesised with Friedrich Weygand, the exact same activity was found.
At the time, Otto Warburg had already showed that a yellow ferment occurs in yeast
which acts within a system that oxidises glucose-6-phosphoric acid ester to gluconic acid-6-phosphoric acid ester,
so that the activity of the ferment can be tracked by observing the oxygen consumption.
By exposing his still crude ferment solution to light, Warburg obtained a chloroform-soluble fragment
and the same fragment is obtained when the vitamin we isolated by crystallisation from milk is exposed to light.
This observation led to the concept that vitamins are building blocks of ferments.
Subsequently, this concept has been proved more precisely and in greater detail.
The concept of a vitamin as a building block of a ferment explained for the first time
why vitamins are active in such “astonishingly small amounts”.
Because ferments, as catalysts, even in small quantities, are able to convert a large number of organic substrates in cells,
and if a vitamin is a building block of such a ferment, it too, in small amounts, would result in high substance conversion rates.
Theorell had found that Warburg’s yellow ferment could be split by the action of dilute acids into a colourless protein component
and a phosphoric acid ester of lactoflavin.
Together with Friedrich Weygand and Hermann Rudy, a complex multi-step process was used
to obtain lactoflavin-5-phosphoric acid in a fully synthetic form,
and it was possible to combine this synthetic lactoflavin-5-phosphoric acid ester with the protein component
to obtain a fully catalytically active yellow ferment.
Other flavins that do not occur in nature
but can be synthesised have similarly been converted to synthetic yellow ferments by combining them with protein.
The importance of the phosphoric acid residue for linking the vitamin to the specific protein body
was illustrated in these experiments.
The next picture shows you that the vitamin itself, the lactoflavin, has a certain affinity for the protein body.
But this affinity is much smaller than that of lactoflavin-5-phosphoric acid ester.
Here time is plotted, here oxygen consumption in cubic millimetres^2 in the enzymatic reaction.
whereas 30 and 150 grams of lactoflavin bound only to protein, which dissociates much more strongly,
is needed to even approach similar reaction speeds.
The concept of a vitamin as the building block of a ferment was also recognised
two years later by Karl Lohmann of the Meyerhof Institute in Heidelberg for vitamin B1,
the antineuritic vitamin, in the form of chlorophosphoric acid ester, which forms a building block of carboxylase,
an enzyme that plays a key role in splitting carbon dioxide from alpha-ketocarbonic acid.
Vitamin B6 was also shown to have a similar relationship to ferments that split CO2 from amino acids.
The generalisation of this concept became so widely accepted that as soon as a new vitamin was discovered,
the question was asked in which ferment it acts as a building block.
But this concept has not been proved in detail for the entire group of fat-soluble vitamins.
It is questionable whether it will prove generally valid for this area.
Although we know from George Wald’s elegant experiments that,
for example, a cys-aldehyde of vitamin A forms visual purple in our retinas after combining with a specific protein body.
But we also know that this function is definitely not the only one vitamin A has in the body.
Lactoflavin is a vitamin not only in higher animals and for humans, it is also vital for many microorganisms,
a growth factor, for example, of many lactic acid bacteria, whose activity is manifested when milk turns sour.
This formation of lactic acid from carbohydrate also plays a key role in our muscles whenever work is performed.
And the mechanism of lactic acid production was the object of a long series of admirable experiments by Otto Meyerhof.
If I dedicate a few words to Meyerhof’s work here,
I do so to convey the words this great physiologist said to me shortly before his death.
I also do so to pave the way for an understanding of a specific lactic-acid-forming organism that we will now focus on.
According to a long-standing principle formulated by Adolf von Baeyers,
when oxygen combines with organic compounds, it tends to go where there is already oxygen.
In the case of fructose with its 6 C atoms, we might therefore expect the C1 atom, which is at the highest oxidation state,
namely the aldehydic state, to be the one to form CO2
during alcoholic fermentation or a carboxyl group of lactic acid during lactic-acid fermentation.
But according to experiments by Otto Meyerhof, that is not the case.
Consider the 6 C atoms of fructose here on the left, the first C atom of which is an aldehyde group,
while each further C atom carries an alcoholic hydroxyl group.
In a nutshell, Otto Meyerhof’s experiments showed
that during lactic acid formation this first C atom is reduced to a methyl group of a lactic acid molecule
and that the carboxyl group of both moles of lactic acid come from the third and fourth C atoms of fructose.
The accuracy of Meyerhof’s concept can be easily proved today with the help of radioactive indicators.
Mr. von Hevesy explained the principle to you this morning.
The result was a resounding confirmation of Otto Meyerhof’s views.
In this case it is possible to start with a non-radioactive pentose and to add the first C atom in radioactive form.
In this way you obtain a fructose in which only the C1 atom is radioactive.
On the other hand, if, for example, you grow sugar beets in an atmosphere of radioactive CO2,
you obtain a sucrose from which you can extract fructose and glucose
in which nearly all the radioactivity is located on the third and fourth C atoms.
Thus, if we use glucose in which only the C1 atom is radioactive,
we will find practically the entire activity in the one methyl group of lactic acid.
In the case of alcoholic fermentation, the relationships are such
that these two methyl groups from two molecules correspond to ethyl alcohol, and these two are released in the form of CO2.
But last autumn a paper by Gonzales in Texas appeared which Mr. Meyerhof discussed with me in detail.
The experiment was superbly conducted.
It showed that another microorganism, Lactobacillus mesenteroides, breaks down fructose in such a way
that 50% lactic-acid fermentation occurs,
namely one mole of lactic acid is formed, as here, but instead of a second molecule of lactic acid,
one mole of CO2 and one mole of ethyl alcohol are produced.
And it was found that even when only the C1 atom in the fructose was radioactive, the entire radioactivity was found in the CO2.
That is entirely at odds with Meyerhof’s scheme.
It would have been understandable
if the ethyl alcohol here had arisen from the first and second C atoms and the CO2 from the third.
But that is definitely not the case.
Meyerhof therefore concluded that his scheme is probably generally valid for so-called homofermentative lactic-acid producers,
that is, for those microorganisms that form two moles of lactic acid from one mole of sugar.
And that the fundamentally different mechanism used by heterofermentative microorganisms to break down carbohydrates,
of which I’ve listed three, requires further research.
Here, the breakdown scheme described has been clearly proved.
For III and IV it has not been proved, an example of aldehyde, which produces, in addition to one mole of lactic acid,
one mole of acetaldehyde and one CO2.
The table shows another microorganism, Lactobacillus bifidus, which is also heterofermentative.
It produces one mole of lactic acid and one mole of acetic acid.
Lactobacillus bifidus was first isolated by Tissier in Paris in 1899 from the faeces of a breast-fed infant.
Science is greatly indebted to the Heidelberg pediatrician Moro for his research.
It is a highly sensitive microorganism.
Its name is meant to express its shape, as it is branched, or bifurcated, at the ends, hence bifidus.
It is Gram-positive, taking on Gram stain readily.
The faeces of a breast-fed infant contains almost only Lactobacillus bifidus.
When an infant is fed cow’s milk, the quantity of Lactobacillus bifidus rapidly declines;
when human milk is given again, their number rebounds.
It is an anaerobe, meaning that it has a poor tolerance of oxygen
and this has created numerous problems with regard to its isolation.
According to Hotchkiss, it stains bright red.
This staining is achieved by first exposing the microorganism to a solution of sodium periodate,
which splits carbohydrates into aldehyde groups.
In the second step it is exposed to fuchsin-sulfuric acid, which makes the aldehyde groups visible by staining them red.
Under the microscope, Lactobacillus bifidus cultures look like alphabet soup with nothing but capital Y's floating in it.
But this structure can be altered, as I mentioned, quite easily, not only under the influence of oxygen,
and the changes are not only expressed morphologically.
Rather, it has been shown that along with many other changes, a characteristic physiological property can also change suddenly.
Instead of optically active lactic acid, only racemic lactic acid is produced.
Just three years ago researchers in the United States
succeeded in establishing a clone of Lactobacillus bifidus for growing cultures,
in which all the cells are derived from one and the same parent cell.
Subsequently, Paul György isolated a mutant Lactobacillus bifidus strain from the faeces of a breast-fed infant.
This strain had the property of growing only if human milk was added to the nutrient medium or solution.
When cow’s milk was added, the mutant strain did not develop.
This mutant has long been referred to by its laboratory name, 212A.
It was later registered under the name Lactobacillus bifidus var. Penn.
Penn is short for the University of Pennsylvania in Philadelphia, where the strain was first isolated.
The next picture shows some known differences between the composition of human milk and cow’s milk.
Most notably, human milk contains only half as much protein as cow’s milk, so that to prepare infant formulas,
cow’s milk is often diluted by around 50% so that it contains the same protein content.
And then sugar is added to bring it up to 6.6%.
And in dozens of variants, you can also add other vitamins, mineral salts and so forth.
For Lactobacillus bifidus there appears to be some hitherto unknown essential difference
in the chemical composition of human milk and cow’s milk.
And although it was not initially clear
what impact the nutrient requirement of such a mutant has on an infant’s overall intestinal flora,
I believe that Paul György’s discovery has enriched and invigorated an important chapter in pediatrics by raising new questions.
In any case, it led to a test, by means of which it became possible to chemically track down an active substance in human milk.
And the test is carried out as follows: The gas atmosphere is a mixture containing 90% nitrogen and 10% CO2.
These bifidus strains are CO2-dependent.
In addition, some hydrogen is added to the atmosphere and then a glowing platinum mesh burns off the remaining oxygen,
so that growth takes place under strictly anaerobic conditions.
The next picture shows the first half of the synthetic nutrient solution that György developed together with Dr Rose.
A litre of this semisynthetic nutrient solution contains
Also a number of amino acids, which are listed here.
But the nutrient medium has so far never been fully synthesised.
You still have to add a casein hydrolysate, and not an insignificant amount of it,
in order to carry out the following determinations.
The next picture shows more mineral salts and vitamins required for this test:
potassium phosphate in relatively large quantity, iron, manganese and the vitamins shown here.
All of this is sterilised, and only then is one gram of ascorbic acid added to one litre.
Now everything is ready for the test.
To determine the activity, the total acid,
that is the sum of lactic acid and acetic acid produced by the growing bifidus cells, is potentiometrically titrated.
Here you see a comparison of acid production in 40 hours at 37° after the addition of human milk and cow’s milk.
Here it states how many tenths of a cubic centimetre of normal acid are produced.
This word could also have been written with two g’s.
You can see that in the case of human milk just 0.023 cubic centimetres is enough to produce 2.2 cubic centimetres of acid,
whereas with cow’s milk one cubic centimetre, fifty times more, is needed to produce the same quantity of acid.
Based on these figures, we would say that cow’s milk has only a fiftieth of the activity of human milk.
In total, under the conditions used,
approximately 18 tenths of a cubic centimetre of normal acid is produced from the lactose present,
so that half that quantity, half the maximum possible lactic acid, or 9 tenths of a cubic centimetre,
has been defined as one unit of activity.
In other words, one bifidus unit is the quantity of bifidus-active substance needed
to produce 9 tenths of a cubic centimetre of normal acid.
The next picture compares the activity of milk from various animal species.
You can see the activity is very low in all strict herbivores.
One unit for such and such cubic centimetres of milk, the smaller the number, the greater the activity,
and the last column shows the relative activities, taking 0.06 cubic centimetres for average human milk as equivalent to 100.
You can see that guinea pig milk is practically inactive, cow’s milk, sheep’s milk and goat’s milk have very little activity,
just 2 to 3% the activity of human milk.
Pig’s milk is somewhat more active, and the most active of all the animals listed here is rat’s milk and rat’s colostrum,
which is the very first milk the animal secretes after giving birth, and is not only more active,
but around twice as active as average human milk.
The most active we know is human colostrum, the first milk of a woman.
There is also a value for cow’s colostrum.
The case is as follows.
At the time of birth, cow’s milk also has considerable activity.
But it falls from this value of 40 to around 2 to 3% in the space of just two to three days.
The activity of human colostrum declines much more slowly to reach, over the course of many, many weeks and months,
a relatively constant level of 0.06 cubic centimetres of human milk per unit.
We have also studied women who have breast-fed for a whole year or longer.
And we found that during such long lactation, the activity continues to decline to 0.08, 0.12 cubic centimetres.
But what I also want to mention is that precisely in those women who have breast-fed for so long,
a bifidus activity of 0.02 cubic centimetres per unit was still found six to ten weeks postpartum,
a value that is otherwise found only in human colostrum.
Besides milk and colostrum, human sperm is also highly bifidus-active.
Sperm from a bull or bulls’ testicles is completely inactive.
The next picture shows examples of the distribution of the bifidus-active substance in the human body.
At the top is again human colostrum at 0.01 to 0.2 cubic centimetres, with a mean of 0.015.
Then milk with 0.02 to 0.15, or 0.06 on average.
Meconium, the intestinal content of a newborn infant, in a 10% suspension is also highly active.
Sperm 0.07 cubic centimetres on average.
Gastric juice, saliva, urine, I’ll have something special to say about them later. Tears too.
If you want to compare this figure with those for milk, you have to take into account the fact
that milk contains approximately 12% dry substance, compared to only 0.5% organic substance in human tear fluid on average.
Calculated on the basis of dry substance … [stops here]