File:811-x men

THE URINE OF THE GOOSEFISH(LOPHIUS PISCATORIUS): ITS NITROGENOUS CONSTITUENTS WITH SPECIAL REFERENCE TO
THE PRESENCE IN IT OF TRIMETHYLAMINE OXIDE.

УРИНАТА НА ГЪСКАТА РИБА (LOPHIUS PISCATORIUS): АЗОТНИТЕ СИ СЪСТАВКИ СЪС СПЕЦИАЛНИПРЕПОРЪЧКА ЗА

НАЛИЧИЕТО В НЕГО НА ТРИМЕТИЛАМИНОВ ОКСИД.

Основната цел на настоящото разследване беше опит да се определи естеството на азотните съставки. Денис (5), в
резултат на нейните анализи на урината на Лофий, също отбеляза това съществуване на голяма част от азота в
урината в неопределена форма. In preliminary tests to determine the form in which this nitrogen existed, search was
made for the presence of such compounds as the amino acids, purines, allantoin, betaine, elc., which are common
constituents of the urine of other animal species. Although the first of these compounds was found to be present in
relatively large amount, the others were lacking or present in only slight traces. There still remained undetermined a large
fraction of the total nitrogen.

В предварителните тестове за определяне на формата, в която този азот съществува, беше направено търсене за
наличието на такива съединения катоаминокиселини, пурини, алантоин, бетаин и др., които са често срещани
съставки на урината на други животински видове. въпреки че първо то от тези съединения е установено, че
присъства в относително голямо количество, другите липсваха или присъстваха само в незначителни количества
следи. Все още остава неопределена голяма част от общия азот.

След няколко неуспешни опита за изолация това вещество, беше установено, че при определени условия могат да се
получат относително големи количества триметиламин. Това амин първоначално не присъства в урината и не може
да има образувани от някоя от познатите съставки. Улика беше така предоставени относно естеството на
неопределените нитроген. По-нататъшни експерименти показаха, че това неопределенодобитият азот присъства
като триметиламин оксид.

Trimethylamine Oxide.

Trimethylamine oxide was found to be present in the urine of Lophius in high concentration. In order to isolate this
substance for identification, the urine was evaporated to dryness on the water bath in the presence of HCI, and the residue
extracted with absolute alcohol. The alcohol was driven off on the water bath leaving a -brownish, syrupy liquid containing
the trimethylamine oxide together with some of the other nitrogenous substances of the urine. It may be separated from
these substances according to the procedures of Suwa (24) or Poller and Linneweh (22).In their method the syrupy liquid is
dissolved in 5 per cent sulfuric acid, precipitated by saturated phosphotungstic acid in 5 per cent sulfuric acid solution,
allowed to stand overnight, and the precipitate removed on a Buchner funnel and washed with 5 per cent sulfuric acid. This
precipitate is dissolved in acetone and de- composed with baryta. After removal of the excess of baryta with carbon dioxide,
the solution is evaporated to a syrup and the trimethylamine oxide precipitated as the aurichloride with 30 per cent
aqueous gold chloride. In subsequent work it was found that the trimethylamine oxide, when only slightly contaminated
with other substances, may be precipitated directly from the syrupy liquid obtained after evaporating the alcohol as
described above. After recrystallization the comparatively pure compound was obtained. The aurichloride of trimethylamine
oxide was obtained from each of the urines, except Atlantic I, in which, for reasons to described later, he oxide had been
completely decomposed to trimethylamine. A typical experiment follows. 150 cc. of Atlantic II urine were distilled in vacua
with magnesium oxide to remove all the trimethylamine and ammonia present. The distillate contained 100 mg. of nitrogen
almost entirely in the form of trimethylamine, identified by converting it to its aurichloride which was analyzed. The residue
in the distilling flask, after filtration and acidification with hydrochloric acid, was treated as described above, 3.07 gm. of the
aurichloride of trimethylamine oxide being obtained. This amount corresponded to 80 per cent of the undetermined
nitrogen, excluding that recovered as trimethylamine, in the distillate from the magnesium oxide. The aurichloride melted at
255” with decomposition. The melting point is given by previous workers as 250-257’. On analysis of the aurichloride, there
was found, Au, 47.7 percent; N, 3.39 per cent. Calculated for (CH&NO.HAuCl, Au,47.5 per cent, N, 3.37 per cent. The picrate
prepared from a portion of the above aurichloride melted at 197”. Previous workers give it as 187-202”, depending upon the
rapidity’ with which it is heated. A further identification of the existence of trimethylamine oxide was made by heating the
urine with sodium hydroxide and zinc dust and determining the trimethylamine liberated. In the case of each of the urines
studied, an amount of base was liberated corresponding to the total undetermined nitrogen present.Trimethylamine oxide,
as a chemical entity, was first prepared and described by Dunstan and Goulding (7) in 1894. They formed it by the
interaction of methyl iodide and hydroxylamine and later showed that it could be more easily prepared by treating
trimethylamine with a 3 per cent aqueous solution of hydrogen peroxide and allowing the mixture to stand at room
temperature for 24 hours. Other amine oxides were prepared in a similar manner. Since the initial work of Dunstan and
GouIding, the amine oxides have been established as definite, stable chemical compounds by a number of workers (14). The
structure of trimethylamine oxide has also been studied by several investigators, the last of whom, Noyes (21), considers it
as (CH&N: : ! with a double union between the nitrogen and oxygen atoms. Ionization studies by the same investigator
indicated that the compound may or may not be combined with water when in solution. Although the presence of
trimethylamine oxide in urine has never been heretofore recorded, this substance has been found in other biological
materials. Suwa (24) isolated a substance from the muscles of the dogfish (Acanthias oulgaris) which he definitely showed to
have the composition and properties of trimethylamine oxide. Henae (13) obtained the substance from the muscles of
cephalopods. More recently Poller and Linneweh (22) have found it in the muscle and roe of the herring (Clupea
harengus).One might expect, therefore, that trimethylamine oxide might be of quite common occurrence, and its
appearance in the urine of Lophius is thus not surprising. Although no attempt has been made to isolate trimethylamine
oxide from the muscle of Lophius, its occurrence in the sources cited above renders it very probable that it is likewise
present in the muscle of Lophius and that this is the source of the material found in the urine. It is possible also that it may
have, in part at least, an exogenous origin, being derived from the numerous Ashes and Crustacea which Lophius ingests.
Although not demonstrated to appear there, it is not unlikely that trimethylamine oxide is also a constituent of the urine of
the dogfish and herring and most probably of other animal species. In passing, it may also be noted that the hitherto
unisolated substance termed aminol, which is found in herring brine and has a bactericidal action in the presence of lime
(3), may be trimethylamine oxide.As regards the physiological significance of the occurrence of trimethylamine oxide in the
lower vertebrates, nothing definitely may be said. From the high concentration in which it occurs in the urine-representing
about half of the total urinary nitrogen and the relatively high concentration in which it has been found in the muscles, it is
but natural to conclude that it forms one of the main end-products of protein metabolism. It is possible that it may arise
from betaine which, according to the investigations of Kutscher and Ackermann (19), is of wide occurrence in the lower
vertebrates. Betaine, in fact, when fed to rabbits was found to be converted to trimethylamine by Kohlrausch (17). The
oxidation of the trimethylamine would then take place according to the schema proposed by Ackermann, Poller, and
Linneweh (2). As a result of a study of the reactions of trimethylamine oxide, the Iatter investigators have suggested that it is
a bioIogica1 hydrogen acceptor in the sense of the word as used by Wieland (29).

Trimethylumine.

The trimethylamine was identified by distillation in 2racuo with magnesium oxide, according to the procedure of Takeda
(25)2 by collecting in acid, evaporating the distillate to dryness, extracting with alcohol, and preparing the gold salt of
trimethylamine hydrochloride. The product melted with decomposition at 228’: Au found, 49.6 per cent; N found, 3.46 per
cent. Calculated for (CH&N.HAuCla, Au, 49.4 per cent; N, 3.51 per cent.The presence of so great an amount of amine in the
urine (Atlantic I), was quite unexpected, as previous experiments had indicated the presence of only very small amounts of
this substance. The explanation of its occurrence in so high a concentration in this specimen is easily accounted for when
one considers the relative ease with which trimethylamine oxide may be reduced to the amine. It is, therefore, logical to
assume that the trimethyIamine found in this particular specimen of urine resuhed from decomposition in alkaline solution
of the trimethylamine oxide originally present.Determinations of the amount of trimethylamine present in the other urines
were also made by the method of Takeda (25).

Woods HoIe urine, several weeks after its collection, contained 9 mg. of trimethylamine nitrogen per 199 cc. of urine: after
10 weeks it contained 23 mg. per 169 cc. Atlantic I, as already stated, contained practically all the undetermined nitrogen in
the form of trimethylamine. Atlantic II contained 7 mg. of amine

* Experiments with trimethylamine oxide, prepared by the action of hydrogen peroxide on trimethylamine, showed that the
former compound underwent no decomposition when subjected to the procedure of Takeds(25).

nitrogen per 100 cc. a week after its collection and 70 mg. 5 weeks later. Aquarium urine contained 6 mg. of trimethylamine
nitrogen, a week after its collection. The trimethylamine present in the urine most probably was absent in the fresh
specimens, but resulted from the slow decomposition of the oxide, as is indicated by the increase in its content with time.
The chemical nature of the oxide is such as to lead one to expect this decomposition. It is also possible that other
compounds such as nova&e, which Kutscher (18) isolated from crab extracts and which give trimethylamine on heating with
alkali, are also present in the urine although no attempt was made in this investigation to isolate these substances. Since at
least 80 per cent of the undetermined nitrogen of the urine of Lophius could always be recovered in the form of
trimethylamine plus trimethylamine oxide, we may conclude that the latter substance constitutes the greater part of the
hitherto undetermined nitrogen. The large and increasing amounts of the trimethylamine found in the urine after some
weeks, but not present in the fresh material, result, no doubt, from the reduction of the oxide.

SUMMARY.
A study was made of the urinary constituents of the goosefish, Lophius piscatorius. The chief nitrogenous constituents of
this urine were found to be creatine, creatinine, and amino acids. Ammonia, urea, and uric acid were present only in small
amounts. The remainder of the urinary nitrogen was found to be made up chiefly of trimethylamine oxide which was
isolated and identified. Specimens of urine obtained from the fish freshly caught in the Atlantic were found to contain much
more nitrogen and much less inorganic salts than the specimens reported previously by other observers.

METABOLISM OF AMINES. Не
BY WILSON D. LANGLEY AND RICHARD J. WEBER.*
(From the Department of Biochemistry, university of Buffalo Medical School,
Bu$alo.)
(Received for publication, September 4, 1930.)
File:66957

METABOLISM OF AMINES.
I. TRIMETHYLAMINE.
BY WILSON D. LANGLEY.
(From the Department of Biochemistry, University of Buffalo Medical School,
Buffalo.)
(Received for publication, August 12, 1929.)
[1] W. D. Langley, “Metabolism of Amines,” J. Biol. Chem., vol. 84, no. 2, pp. 561–570, 1929, doi: 10.1016/s0021-
9258(18)77015-4.
[2] “Supplement to N A T U R E of February 18 , 1939,” p. 1939, 1939.
[1]

Обикновено е известно, че простите алифатни амини са токсични

вещества и да има малка терапевтична употреба. Те се образуват
по време на бактериално разлагане на повечето животински или растителни вещества,
и могат да бъдат открити лесно поради тяхната променливост и
характерни миризми. Аминът, който най-често се среща при разлагане
органичната материя е триметиламин. То също е освободено
при деструктивна дестилация на определени тъкани. Trimethylamine
has been thought to be present in small amounts in the blood and
urine of mammals (1). Takeda (2) and Erdmann (3) have shown
by working with fresh blood and urine that this is not the case.
Trimethylamine oxide has been isolated several times from fish
muscle and urine (4-9). Suwa (10) has injected trimethylamine
oxide into rabbits, and, has isolated both trimethylamine and the
unchanged oxide from the urine. This finding would lead one to
suspect that trimethylamine is not readily metabolized by mammals,
but is excreted unchanged. Neither the free amine nor the
amine oxide has been found in fresh mammalian tissue. The
occurrence of trimethylamine combined in the form of quaternary
ammonium bases among the extractives of various tissues is well
known. ‘Such bases are betaine, carnitine, choline, ergothioneine,
neosine, and trimethylaminocrotonic acid.
The principal chemical decomposition which characterizes a
quaternary ammonium base is the liberation of a tertiary amine on
heating with alkali. Kohlrausch (11) has shown that betaine,
when fed to rabbits, gives rise to trimethylamine, and it is conceivable
that each of the quaternary ammonium bases named
above may decompose similarly in the body so as to liberate trimethylamine. It is not improbable that
the physiological functions of the substances may be associated with such a chemical
transformation in which trimethylamine is liberated. In view of these considerations, it is surprising that
quantitative studies of the metabolism of trimethylamine itself have not been made.
Триметиламин
Смята се, че присъства в малки количества в кръвта и
урина на бозайници (1). Такеда (2) и Ердман (3) се показаха
чрез работа с прясна кръв и урина, че това не е така.
Триметиламин оксид е изолиран няколко пъти от риба
мускули и урина (4-9). Suwa (10) е инжектирал триметиламин
оксид в зайци и изолира триметиламин и
непроменен оксид от урината. Това откритие би довело до
подозирате, че триметиламинът не се метаболизира лесно от бозайници,
но се екскретира непроменен. Нито свободният амин, нито
амин оксид е открит в пресни тъкани на бозайници. The
поява на триметиламин, комбиниран под формата на кватернер
амониеви основи сред екстрактивите на различни тъкани е добре
известен. Такива основи са бетаин, карнитин, холин, ерготионеин,
неозин и триметиламинокротонова киселина.
Основното химично разлагане, което характеризира a
кватернерна амониева основа е освобождаването на третичен амин върху
нагряване с алкали. Kohlrausch (11) показа, че бетаинът,
когато се дава на зайци, се образува триметиламин и това е възможно
че всяка от назованите кватернерни амониеви бази
по-горе може да се разложи по подобен начин в тялото, така че да освободи триметиламин. Не
е изключено физиологичните функции на веществата да бъдат свързани с такъв химикал
трансформация, при която се освобождава триметиламин. С оглед на тези съображения е
изненадващо, че не са направени количествени изследвания на метаболизма на самия
триметиламин.
A review of the analytical procedures for the separation and
determination of each of the methylamines and of ammonia in a
mixture has been made by Weber and Wilson (12). These workers
proved that the procedures described in the literature prior
to 1918 led to inaccurate results, and they developed a new quantitative
method for the determination of each component in a
mixture of the four bases. Woodward and Alsberg (13) have since
developed a method for the determination of trimethylamine only
in the presence of the other amines and ammonia, but the method
offers no improvement over that of Weber and Wilson.
The present work is an application of the procedure of Weber
and Wilson to a study of the metabolism of trimethylamine by
rabbits. The rabbits were maintained upon a diet of carrots,
and the nitrogen intake was reduced to less than 1 gm. per day.
Weighed amounts of the amine hydrochloride dissolved in small
amounts of water were given through a stomach tube, and sodium
carbonate in amounts sufficient to prevent the depletion of base
in the tissues was given immediately after. The urine of the
rabbits was then collected in 24 hour periods and was analyzed
for ammonia, the amines, and its total nitrogen content. In
order to avoid loss of the volatile amines from the urine, and in the bottles used for collecting the urine.
After the measurement of the volume of the urine and the removal
of samples which were used for the determination of total
nitrogen, magnesium oxide was added, and the alkaline urine
was distilled in vucuo. Octyl alcohol was used to prevent frothing.
The distillate was collected in standard acid contained in a receiver
which was cooled with ice. When about 200 cc. of distillate
were obtained, the excess of standard acid was titrated
with sodium hydroxide. The acid which was neutralized during
the distillation was a measure of the total volatile base. The solution of bases then was analyzed for
ammonia and the amines
as follows :l
The solution was made alkaline with a mixture of sodium
hydroxide and sodium carbonate, and was diluted to a known
volume. Yellow mercuric oxide was added so as to give a fine
suspension in the alkaline solution. The flask was covered to
exclude light and was shaken for 1 hour; the suspension then was
allowed to settle for several hours. The solution was filtered
through cotton by use of air pressure, and aliquot portions of the
filtrate were distilled into standard acid. The acid neutralized
during the distillation was a measure of the total amines. The
difference between the total volatile base and the total amines was
considered to be ammonia.
The aliquot distillates containing the volatile amines were
combined, and the solution was acidified with sulfuric acid and
evaporated on a steam bath to about 25 to 30 cc. It was treated
with an excess of sodium nitrite and glacial acetic acid. The
nitrous acid produced decomposed the primary amine, converted
the secondary amine into the nitrosoamine, and left the tertiary
amine unchanged. The solution containing the nitrous acid was
allowed to stand for about 12 hours, and then was made alkaline
with sodium hydroxide and distilled into standard acid. The
acid neutralized was a measure of the tertiary amine present.
Any nitrosoamine which was formed from secondary amine was
now to be found mixed with the tertiary amine in the distillate.
The nitrosoamine, however, was indifferent toward the acid and
alkali used in the titration, so that the values for the tertiary
amine were not affected by its presence.
To the solution containing nitrosoamine and tertiary amine
after the titration were added concentrated hydrochloric acid and
metallic zinc, in order to reduce the nitrosoamine to the secondary
amine again. The solution was allowed to stand for about 24
hours while the evolution of hydrogen continued. It then was
made alkaline by the addition of sodium hydroxide, and the
amines were distilled again into standard acid. The acid neutralized
was a measure of both the secondary and tertiary amines.
The difference between the titration values before and after reduction represented secondary amine.
The primary amine was
assumed to be the difference between the total amines and the
sum of secondary and tertiary amines. This method of estimating
the primary amine was different from the procedure of Weber and
Wilson, and it made the analyses less accurate than theirs. However,
by use of this modification, the analyses could be made in
much less time, and by a more simple process

Trimethylamine Formation in Relation to the Viable

Bacterial Population of Spoiling Fish Muscle
[2]
IN view of the recent communication in NATURE on
this subject\ the following comments may be of
interest.
Experiments with haddocks stowed in ice 2 have
shown that there is no close correlation between the
viable count and the trimethylamine content of the
flesh. Indeed, the content of dimethylamine, also
produced from trimethylamine oxide, follows the
general curve of bacterial growth much more closely 2 •
Whilst it is most probably true that orily certain
organisms are capable of reducing the oxide, the
empirical fact remains that in haddocks stowed in
ice--the usual industrial preservative-the increase
in content of trimethylamine has always been found
to follow the organoleptic phases of spoilage in a
sufficiently consistent way to render the amount of
this substance present a useful index of the extent
of spoilage. It would seem that, as Dr. Tarr suggests
may be the case, the proportion of trimethylamineforming
to non-trimethylamine-forming bacteria is
fairly constant, at least under the conditions of
stowage mentioned above. The content of dimethylamine
has similarly been found to be a useful index
of spoilage. Indeed, reference to these two indexes
together serves satisfactorily to characterize the condition
of haddocks at any stage right up to the point
of definite staleness.
Finally, whilst it is reasonable to assume, until
the contrary is proved, that the flora of the external
surfaces of all sea fish is much the same, it has been
found that spoilage may pursue different courses in
different species. Thus in the dogfish the contents
of diamine and triamine never rise during stowage in
ice to the extent that might be expected from the
much larger amount of oxide present as compared
with the haddock. Indeed, diamine is present only
in traces during stowage for thirty days ; whilst the
amount of triamine is only of the same order as that
found under similar conditions in the haddock. The
explanation of these results may be that the large
amount of ammonia formed from urea, present in
much larger amounts in the dogfish than in the
haddock, causes the trimethylamine oxide reducers to
be overgrown by other bacteria. Certainly it has
been noticed that after stowage the flora of the dogfish
is considerably restricted in types as compared
with that of the haddock.
In this connexion, it is of interest that Dr. Tarr's
work, as reported', was done with smoked fish, since
it is probable that the constituents of smoke effect
considerable alteration in the original flora of the fish.
Torry Research Station,
Aberdeen. Jan. 19.
JAMES M. SHEWAN.
' Tarr, H. L. A., NATURE, 142, 1078 (1938).
2 Shewan, J. M., Report of the Food Investigation Board (London),

Sec. 4, p. 75 (1937)

Studies of Fish Spoilage

III. The Trimethylamine Oxide Content of the Muscles
of Nova Scotia Fish
By S.A. BBemv
Atlartic Fisheries Erperitnental Station
(Receivedf or pwblicationM a'ch 24,1938)
i,i-ettyt"mine oxide was roundi:t;::t:"e press juice or alt salt water fish
examineda nd in the anadromousfi sh Pontolobusp seudoharengutask en from the sea. Traces
weref oundi n Anguillat akenf rom salt water,b ut nonei n Angui.llafr om freshw atef

The presence of trimethylamine oxide in dogfish 'uvas demonstrated by Suwa

( 1909). He showed also that bacteria reduce the oxide to trimethylamine and
concluded that the particular odour of spoiling sea fish is due to the increase in the
amine. Beatty and Gibbons (1937) and Beatty (1938) have shown that, in spoiling
cod muscle press juice, trimethylamine is produced only as a result of the action
of the bacteria producing the spoilage, and that it is derived from the reduction of
trimethylamine oxide.
The literature previous to 1933 has been reviewed by Kutscher and Ackermann
(1933). The oxide has been demonstrated by the isolation and analysis of typical
salts, in selachians and salt water teleosts, but it was not found in Angwilla vulgari,s,
Salmo salar, Perca fl,wviatilis, or Cyprinws carpio

Because of the
importance of trimethylamine oxide in studies of fish spoilage, and beiause of the
conflict of evidence as to its oCcurrence in nature, a survey (not yet completed)
has been undertaken as to its occurrence in fresh water and sea fish of Nova Scotia

Trimethylamine oxide was determined by Lintzel's method (1934). Since the

primary purpose of the investigation was to aid in the investigation of fish spoilage,
and since previous studies were done mainly on press juice, the trimethylamine
oxide content of the muscle press juice rather than of the muscle itself was determined.
Analyses of press juice and muscle of Ctupea harengus showed the press
juice to be approximately 19 per cent higher in oxide than the muscle itself. The
results of the analyses are shown in table I.
Grollman (1929) showed that the urine of Lophius pi,scatorius is high in the
oxide. It is very improbable that this compound is used as a means of nitrogen
elimination
The marked difference between sea fish and fresh water fish points possibly to
a different mechanism of spoilage. Trimethylamine is one of the earliest compounds
produced during the spoilage gf sea fish, and is definitely responsible for
some of the spoilage odour. Previously, the development of spoilage, as determined
by the production of trimethylamine, was followed only in members of the cod
family. Since the oxide is so widely distributed in sea fish, its reduction by spoilage
bacteria should be a general reaction, although the great differences in the oxide
contents of various species may mean that different spoilage threshold values may
be necessary
Muscles of all sea fish examined contained trimethylamine oxide.
Trimethylamine oxide was not found in muscles of fresh water fish.
The oxide was present in muscle of Powolobus psewd,oharengutsh at migrates
into fresh water to spawn, but was absent in muscle of Anguilla while in fresh
water.
The data throw no light on the physiological importance of trimethylamine
oxide.
production of trimethylamine by spoilage bacteria is likely to be a general
phenomenon during the spoilage of all sea fishes examined.

BLOOD AND CRANFIELD : THE DETERMINATION OF BETAINE 829

The Determination of Betaine in

Sugar Beet By-Products
BY J. W. BLOOD, A.I.C., AND H. T. CRANFIELD
(Read at the Meeting, October 7, 1936)

От самото начало се разбра, че петното, което често се описва

като „рибка“ се свързва с бетаин, важна съставна част на захарта
растение цвекло. Следователно определянето на това съединение стана въпрос
важност. Този документ дава, първо, общ преглед на известните методи за
определянето на бетаин и, второ, описание на експерименталната работа
върху определянето на това съединение и адаптирането на разработения метод
към страничните продукти, получени при производството на захарно цвекло.
124. THE METABOLISM OF BETAINE AND ALLIED
TERTIARY NITROGENOUS BASES IN THE RUMINANT
BY WILLIAM LEWIS DAVTES
National Institute for Research in Dairying, University of Reading
(With 1 Figure)
ПРЕЗ последните няколко години се обръща значително внимание на спорадичните
поява на рибен вкус в млякото на крави, хранени със странични продукти от захарно
цвекло, в които бетаинът в значителни количества е обичайна съставка. Има,
средно около 1-5 процента бетаинов азот в меласата от цвекло,
0-5 процента в меласирана каша от цвекло и 0-03 процента в пресни върхове от цвекло
(0-2 процента, в сухото вещество); млечни крави на нормални дажби, съдържащи
подходящи количества от тези продукти могат да се приемат до 100 g. на бетаин
на ден. Горните фуражи също съдържат следи от третичен азот основи, холин,
триметиламин и триметиламин оксид.

The form of the free un-ionised tertiary base found in plants is betaine while, in the tissue and urine of
some aquatic animals, a simpler form, trimethylamine oxide, exists. In the present work,
the latter base has, so far as the ruminant is concerned, been found to be the main metabolite of all the
tertiary bases tested.
Suwa(7) isolated trimethylamine and unchanged trimethylamine oxide from the urine of rabbits
injected with the oxide, while Kohlrausch(8)
found that betaine fed to rabbits gave rise to trimethylamine. Langley(9) has
suggested that bases such as betaine, choline, carnitine and trimethylaminocrotonic
acid may decompose in a similar manner in the body to give rise to
trimethylamine, and that the physiological functions of these substances may
be associated with chemical changes involving the liberation of free trimethylamine.
Langley fed trimethylamine to rabbits, and his maximum
recovery of unchanged amine in the urine amounted to 19-3 per cent. He
concluded that the bulk of the trimethylamine (80-96 per cent.) was wholly
or partly demethylated, part being excreted in minute amount as secondary
amine, none as primary amine, and the residue as urea. From his methods
of analysis it can be gathered that the possible excretion of trimethylamine
in the form of a detoxicated tertiary base, such as trimethylamine oxide, was
not explored.
Trimethylamine oxide is of wide occurrence in the muscle and urine of
some marine animals (Suwa(7), Grollmann(iO)) and undoubtedly plays an important
part in the excretion of tertiary nitrogen in fish generally. No reference
has been made to its occurrence in the urine of higher animals, although
Grollmann advances that possibility. As a free base, the oxide is feebly
dissociated and it is doubtful if it is ionised at the pH of blood and urine.
It can thus be considered an ideal form for the detoxication and excretion of
highly ionised trimethylamine in the animal body, and it is possible that it
enters into the physiological economy of tertiary nitrogen excretion in a
similar way to urea in ammonia excretion. It seems surprising that this line
of investigation has not been previously explored in studying the metabolism
of tertiary bases. Further, the close similarity in properties of the oxide and
of betaine may have led previous workers to mistake the excreted oxide in
the urine for betaine itself.
It was originally intended in the present work to investigate the metabolism
of betaine to the extent only of determining the amount of trimethylamine
formed.
It was evident, therefore, that the main tertiary metabolite was a substance
easily reduced to trimethylamine. This substance was found to be trimethylamine
oxide, a fact which was confirmed in many instances by its isolation
and by the analysis of its derivatives. The investigation was therefore widened
to include the metabolism of choline, the methyl ester of dimethylaminoacetic
acid (an isomer of betaine), trimethylamine oxide and trimethylamine.
The results justified this step, since they showed (a) that the main tertiary
metabolite of betaine, choline and the ester was trimethylamine oxide, (b) that
after feeding with trimethylamine oxide, it -was quantitatively recovered unchanged
in the urine, and (c) that when trimethylamine was fed, it was
quantitatively recovered from the urine as trimethylamine oxide.

Characterisation of the trimethylamine oxide in the urine

Trimethylamine oxide was isolated as picrate from samples of urine taken
at the peak excretion of the base. The base was precipitated with phosphotungstic
acid and, after liberation with baryta, treated with Neuberg's reagent
to separate it from other bases. The base was then precipitated as its picrate
(fine needles, M.P. 197° C). Ether extraction of the picric acid from acid
solution yielded a solution which, on reduction with zinc-copper couple, gave
all its nitrogen as the volatile base, trimethylamine. The aurichloride, prepared
in two cases, gave a melting point of 249-251° C.
Owing to the difficulties of separating the relatively small quantities of
trimethylamine oxide from urine, the yields of picrate were not quantitative,
but the above isolations of crystalline derivatives, together with the formation
of trimethylamine on reduction with zinc-copper couple, were considered
sufficient evidence of the metabolite being trimethylamine oxide.

DISCUSSION OF RESULTS
The rapid and almost quantitative excretion of trimethylamine oxide in
the unchanged form and also the rapid excretion of trimethylamine as its
oxide prove that this is the mode of excretion of tertiary nitrogenous bases
in the ruminant, and that with these simple bases no attack on the methyl
groups attached to the nitrogen occurs. But when one of the groups attached
to the nitrogen is a substituted methyl group (in which one H atom is substituted
by a CH20H or COOH group), there is a varying and considerable
metabolism of both methyl and substituted methyl groups; from 14 to 33 per
cent, only of the nitrogen appears in the urine in the tertiary form, while the
amounts of other forms of methylated nitrogen are insignificant. The metabolic nitrogen must
therefore appear in the urine in molecular form, i.e. either as
trimethylamine oxide or as urea. The greater utilisation of betaine by animals
fed regularly on a betaine-containing food is explained by their greater capacity
to metabolise the nitrogen to urea and not to trimethylamine oxide, but this
will naturally depend on the level of intake of betaine nitrogen.
The trimethylamine obtained by distillation of the urine samples with
magnesia varied generally with the content of trimethylamine oxide in the
urine. It is possible that all the trimethylamine determined by this method
did not exist as such in the urine, but that a small amount of reduction of
the oxide by the reducing agents in cow's urine (sugars, etc., in alkaline
solution) contributed to the value. This matter has not been further investigated.
In most cases the ratio of N(CH3)O nitrogen to N(CH3)3 nitrogen was
from 12 to 15 to 1. Large amounts of trimethylamine were formed by bacterial
action on the stored urine samples. This proved that the betaine
content of the oil, if any, was very low. From a number of observations the
conclusion was reached that there could be very little unchanged betaine in the
experimental urine. It can also be seen from Tables I and II that the
control urine samples contained residual "Stanek" nitrogen in amounts which
were of the same order as those for the experimental urine. The rates of excretion of trimethylamine
oxide show well-defined peaks
at various times after feeding, depending on the amount and form of tertiary
nitrogen fed to the animal. With the simpler bases, N(CH 3)3 and N(CH3)3O,
the maximum rates of excretion of the oxide occur at about 4£-6 hours after
feeding. When feeding small amounts of tertiary nitrogen as choline or the
methyl ester of dimethylaminoacetic acid, the peak again occurred about
4^-6 hours after feeding. But when comparatively large amounts of betaine
were fed, the peak of excretion was pushed forward to about 12 hours after
feeding (Table II); the feeding of lower amounts of betaine caused the peak
to be reached in a shorter time after feeding (6-1 hours, Table I). For an
intake of 100 g. of betaine daily, it can be inferred that the peak of excretion
would be reached in 4—4£ hours.
The importance of a peak of excretion of the oxide rests in the fact that
the maximum concentration of the metabolite would occur in the blood of
the animal shortly before the peak of excretion in the urine would be reached.
The maximum concentration of the oxide in the "residual nitrogen" fraction of milk secreted would
therefore be expected to occur at this time. Since the
trimethylamine oxide in milk is in all probability the precursor of the fishy taint,
it is evident that the possibilities of the development of the taint are greatest
when the cow is milked just previous to the peak of oxide excretion in the
urine. The importance of the time interval between feeding the betainecontaining
food and milking is obvious, and with this is also associated the
amount of betaine fed. With an intake of 100 g. of betaine the peak is reached
in roughly 4£ hours, and it is therefore advisable to feed beet by-products
as far away from the subsequent milking time as possible, namely, during or
soon after the milking time.
The mechanism of the development of the fishy taint in milk by interaction
of the precursor with one or more of the milk constituents is being at present
investigated. The results of this work will be reported in a later paper

SUMMARY
In the ruminant, the main tertiary nitrogenous metabolite of all the tertiary
nitrogenous bases examined is trimethylamine oxide. Small traces of trimethylamine
and insignificant traces of mono- and dimethylamine also occur in
cow's urine. The nitrogen of simple bases (trimethylamine and trimethylamine
oxide) is almost quantitatively and rapidly excreted as trimethylamine oxide.
With betaine, choline and the methyl ester of dimethylaminoacetic acid, only
from 14 to 43 per cent, of the nitrogen is excreted as trimethylamine oxide,
the amount so excreted depending on the level of nitrogen intake, the nature
of the base fed and the degree of accommodation of the animal to the base fed.
The rates of excretion of the tertiary metabolite show well-defined peaks
depending on the nature and amount of the base fed. The feeding of 100 g>
of betaine would show a peak of excretion at about 4£ hours after feeding.
The importance of the time interval between feeding and milking is discussed
from this aspect. No unchanged betaine was be detected in the urine.
The author is indebted to the Sugar Beet Research and Education Committee
(Ministry of Agriculture and Fisheries) for a grant to carry out this
work.
REFERENCES

198. THE DIFFUSION OF TRIMETHYLAMINE

OXIDE FROM THE UDDER
BY W. L. DAVIES
National Institute for Research in Dairying, University of Reading
PREVIOUS investigations(i) have shown that, in the ruminant, the main
tertiary nitrogenous metabolite of betaine and allied tertiary bases is trimethylamine
oxide. Experiments showed that on feeding betaine and methyl
dimethylamino acetate about 16 % of the nitrogen fed appeared as tertiary
nitrogen in the urine. The feeding of choline gave 46 %, while the feeding
of trimethylamine and trimethylamine oxide gave quantitative recoveries
of nitrogen in the tertiary oxide form in the urine. It was also found that a
peak period of excretion occurred at a certain time after feeding, the time
taken to reach this point, and the level of maximum excretion, depending on
the amount and nature of the tertiary base fed.
It was considered that this short period of maximum excretion in the urine
followed closely the period of maximum concentration in the blood. Analysis
of milk samples taken periodically after feeding any of the bases also tended
to show that the concentration of trimethylamine oxide in milk could be
correlated with its concentration in simultaneously drawn samples of urine.
It appeared that the oxide was a substance which diffused rapidly through
blood and tissue and that, as far as the mammary gland was concerned, there
was a constant ratio of concentration of oxide in milk to that in blood.
In order to consolidate these observations it was necessary to investigate
the diffusibility of trimethylamine oxide from the udder into the blood stream
and to attempt to account quantitatively for the material in the urine
from the udder to the blood stream and its later excretion
in the urine.
SUMMARY
On injecting a neutral solution of pure trimethylamine oxide into two
quarters of an udder, the quarters increased in turgidity up to 2-5 hr. and then
relaxed. The oxide diffused rapidly out of the quarters during the third and
fourth hours after injection and a maximum concentration was shown in the
urine at the end of the third hour. In a period of 8-3 hr., 76 % of the oxide
available for diffusion was recovered from the urine.
The milk obtained 7 hr. after injection contained 7-8 mg. % of tertiary
nitrogen, but the secretion had not in that time regained normal composition
owing to the interference with the natural functioning of the quarters.

552 ABSTRACTS OF CHEMICAL PAPERS

ABSTRACTS OF PAPERS PUBLISHED IN OTHER JOURNALS
Food and Drugs
Investigation of Fishy Flavour. W. L. Davies and E. Gill

Certain tentative conclusions as to the nature

.of "fishiness" have been formed, and these are supported by experimental work.
Experiments show that increase of total nitrogen and of organically combined
nitrogen in fish oils and ethereal extracts of fish products is accompanied by
increase of fishiness and of brown colouring matter, and is associated with autoxidation.
As much as 30 per cent. of the total nitrogen may be liberated and distilled
from fishy oils after treatment with various reagents. A number of oils, particularly
linseed, will enter into organic combination with nitrogen if kept for a period of
weeks with a source of nitrogen such as betaine, casein, etc., or much more quickly
if heated at 105" C. with trimethylamine oxide (considerable reduction to trimethylamine
occurs). This combination is accompanied by a development of
fishy odour. On the other hand, such odour is not produced by heating cholesterol
or the unsaponifiable matter of vegetable or animal fats with trimethylamine
oxide to 107" C., although the fatty acid fraction of linseed oil acts like the oil
itself. An aqueous solution of maleic acid (at 100" C.) does not react with
trimethylamine oxide, but in glycerol solution (at 120" to 130" C.) some reduction
.occurs, although there is no development of fishy flavour. Traces of peroxides,
formaldehyde and tertiary nitrogen (either in the form of trimethylamine or its
oxide or both), appear to be associated with fishiness, and positive reactions are
given for formaldehyde and peroxides after fishy oils, their extracts or steam
distillates have been treated with various reagents.
Betaine-content and Nitrogen Distribution of Beet Molasses and
Other Beet By-products. W. L. Davies and H. C. Dowden ( J . SOCC. hem. Irtd., 1936, 55, 175--
179~.)-Beet molasses and molassed beet pulps were collected
from 11 factories in Great Britain and analysed. The method used for determining
betaine in molassed pulp was to shake 60 g. of the finely-ground pulp with
600 ml. of 5 per cent. milk of lime at 60" C. for 1 hour, to leave overnight, and to
filter by suction. The volume of filtrate was determined, and for calculation the
volume V was taken to correspond with V/10 g. of pulp. The calcium is precipitated
by addition of sufficient sodium carbonate and filtered off; after acidification
with sulphuric acid the filtrate is concentrated in vacuo to 100 ml. Total nitrogen
is determined in an aliquot portion, and the remainder of the solution is treated in
the same way as molasses, as follows:-Twenty g. of molasses are stirred with
10 ml. of water, and 100 ml. of 20 per cent. phosphotungstic acid in 5 per cent.
sulphuric acid are added. After settling, the clear liquid is decanted through a
hardened filter in a Buchner funnel, the precipitate is washed by decantation
with 5 per cent. phosphotungstic acid in 2.5 per cent. sulphuric acid, 50ml. in
several washings being used; as much as possible of the precipitate is transferred
to the filter with the last washing, and sucked dry. The precipitate is washed into the original
beaker, and the paper is moistened with baryta water and further
washed. Powdered baryta is added and stirred in until permanent alkalinity is
reached. The mixture is filtered, the barium salts washed, and the filtrate acidified
with hydrochloric acid and evaporated in vacuo to 25 ml. Saturated solutions of
sodium carbonate and mercuric acetate are added in turn until the precipitate is
of a brick-red colour, when 150 ml. of alcohol (free from pyridine) are added, and
the mixture is shaken and left overnight. The solution is filtered, acidified with
hydrochloric acid, concentrated in vacuo to remove the alcohol, and made up to a
convenient volume, and the total nitrogen is determined in an aliquot portion,
excess of sodium sulphide being added to precipitate the mercury before distillation.
Under these conditions, 1.75 mg. of betaine nitrogen escapes precipitation, and a
correction of 0.009 is therefore added to the percentage found. The average
betaine-content of molasses was found to be 5-4 per cent., and that of molassed
pulp 1.8 per cent., the betaine accounting for approximately 38 per cent. of the
soluble nitrogen in each material. Trimethylamine is present only to a very
small e x t e n t 4 - l per cent. of the oxide in beet pulp, which is 5 to 8 times the
amount in molasses. Ten lb. of molassed pulp contain the same amount of betaine
as 601b. of fresh beet tops.