vol. 28, No. 3
`
`Journal of THE was HINGTON ACADEMY of sciences
`84
`CHEMISTRY.-Biochemistry
`by analogy:
`the sulfur of cystine."
`BEN H. NicoLET, Bureau of Dairy Industry.
`We are, I think, all agreed on the proper way to attack a chemical
`problem. One should set up a crucial experiment with the substances
`supposed to be concerned, carry the experiment to completion, and
`see what happens. But often, particularly in biochemistry,
`this is
`hard to do. It may be necessary to carry out an analogous reaction,
`under supposedly more or less analogous conditions. This is known
`as a “model” experiment, though it is very far from being a model
`of what one might wish to do. It frequently amounts to thinking what
`certain molecules should do, and trying to establish, a bit indirectly,
`I shall talk to you this
`whether such a reaction is really plausible.
`evening of “model” experiments, and I intend to see whether I can
`convince you of the value of the conclusions, as yet incompletely con
`firmed, which I shall draw from them.
`Cystine (I) and methionine (II) are the two best known amino
`acids containing sulfur which go to make up proteins. Both are
`readily synthesized by plants, but with regard to animals (including
`outselves) both are essential amino acids—or nearly so.
`An “essential” amino acid is one which must be supplied in the
`diet if an animal
`is to live and grow normally.
`In other words,
`the
`animal body cannot synthesize essential amino acids, or cannot do
`so in sufficient quantity. According to the latest data,” animals can
`not synthesize methionine under any known conditions. On the other
`hand, a cystine deficiency can be corrected either by cystine, methio
`nine, or homocystine (III). It i
`accordingly, a
`t least very probable
`that animals can synthesize cystine, although only, s
`s we yet
`o far a
`know, when methionine o
`r homocystine i
`s fed. The problem tonight
`npart, how they
`how cystine and methionine get their sulfur, and,i
`t again.
`losei
`2(−SCH2CH(NH2)CO2H)
`
`s,
`
`is,
`
`CH-SCH2CH2CH(NH2)CO2H
`II
`
`I
`
`2(−SCH2CH2CH(NH2)CO2H)
`III
`There are two syntheseso
`f cystine on record. Erlenmeyer
`
`treated
`benzoylserine with phosphorus pentasulfide. Emil Fischer converted
`serine ester hydrochloride t
`o 3-chloroalanyl ester hydrochloride, and
`January 13, 1938.
`
`f retiring president, Chemical Society of Washington,
`1 Address o
`Received January 28, 1938.
`* Rose, Science 86: 298. 1937.
`
`Eton Ex. 1026
`1 of 10
`
`

`

`nicolet: biochemistry
`
`85
`
`BaS2
`
`Mar. 15, 1938
`then allowed this to react with barium disulfide. Hydrolysis gave
`cysteine and cystine, respectively.
`PS,
`
`HOCH,CHCO,H
`*HSCH2CHC02H
`I
`I
`NHBz
`NHBz
`PCi8
`-* ClCH2CHC02Et
`HOCH2CHC02Et
`>2(-SCH2CHC02Et)
`I
`I
`I
`NH,HCi
`NH2HCl
`NH2
`Surely neither of these syntheses approaches the possibility of being
`a biological synthesis.
`A few years ago I talked to you once before34 about cystine. I
`asked you then to recall the reactions of aldol formation and dehydra
`It is particularly necessary to remember
`tion, and their reversals.
`that
`all these reactions are reversible.
`0
`OH
`O
`O
`CH,CHCH2CH ?± CHSCH + CHSCH
`jr
`
`0 I
`
`I
`
`CH,CH = CHCH + H20
`I presented to you at that time the idea that the desulfurization of
`cystine by alkali was essentially analogous to the dehydration of an
`aldol, and should obey the same rules,
`including reversibility. The
`work of Dr. H. T. Clarke6 and his students on the hydrolysis of
`It
`cystine was entirely in accord with the first part of this notion.
`showed very clearly that the chief direction of cystine decomposition
`led to hydrogen sulfides, ammonia, and pyruvic acid. In the
`by alkali
`following equations cysteine is used,
`for simplicity of presentation,
`in place of cystine to show the formal analogy to the aldol reactions
`shown above. Under suitable conditions, reactions analogous to most
`of these types can be demonstrated.
`HS H,N O
`HSN 0
`CH21,-CH-COH ^ CH2 = C-COH + H2S
`Jt
`0 0
`HS
`O
`H2N
`I
`II
`I
`II
`II
`CH3C-COH + NH,
`HOCH2 + CH2-COH
`J. Am. Chem. Soc. 53: 3066. 1931.
`3 Nicolet,
`J. Biol. Chem. 95: 389. 1932.
`4 Nicolet,
`J. Biol. Chem. 94: 541. 1931. 102: 171. 1933. 106: 667. 1934.
`« Clarke and others,
`
`V.
`
`Eton Ex. 1026
`2 of 10
`
`

`

`86
`
`JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES
`
`VOL. 28, NO. 3
`
`At the same time, attention was called to the fact that modification
`of the cystine molecule in the direction of peptide formation, or, to
`put it more simply,
`in the direction of combining the carboxyl group
`with an amide grouping (as in peptide formation) and the amino
`group with an acyl group (which could be an aminoacyl group)
`facilitated the removal of hydrogen sulfide (or disulfide) and should
`facilitate its addition.
`It might be argued that the dehydration of an aldol is a very easily
`occurring reaction, and is due to the activating influence of the
`aldehyde group. Cystine, on the other hand, offers considerable re
`the -CO- grouping
`is well known that
`sistance to alkali; and it
`retains very little of the typical "carbonyl"
`present in carboxyl,
`properties. It was, however, shown quite definitely at that time that
`mercaptans add very readily to a,^-unsaturated ketones, such as
`benzalacetone.
`PhCH = CHCOCH, + RSH ^ PhCH(SR)CH,COCH3
`These products lose mercaptans with extreme ease. A quite small
`fraction of the full "carbonyl" activity would therefore be sufficient
`to account for the results obtained.
`You will perhaps not ask me to repeat my earlier talk further. Let
`us assume that the reactions eliminating sulfur occur as suggested,
`and that they are reversible. The simplest reaction by which cystine
`(or cysteine) could be formed, would be the addition of hydrogen
`disulfide (or hydrogen sulfide) to a-aminoacrylic acid (IV), which we
`may also call dehydroalanine, since it represents the removal of two
`hydrogen atoms from alanine.
`But we shall not expect to be able to demonstrate this reaction, as
`for two reasons. Aminoacrylic acid itself is so unstable that it
`such,
`has never been isolated. And, secondly, we should have much better
`hope of success if the carboxyl or amino group, or both, were suitably
`modified. What we should prefer would evidently be a peptide (V)
`in which the dehydroalanyl group was not at
`(at least a tripeptide)
`either end.
`
`CH,
`CH2 = C(NH2)CO,H
`NH2CHR'CO-NH-C-CO-NHCHRCO2H
`IV
`V
`Such a compound was not itself available. For the tentative test,
`a compromise had to be made. The simplest known derivative of
`It
`aminoacrylic acid is Bergmann's acetylaminoacrylic acid (VI).
`
`Eton Ex. 1026
`3 of 10
`
`

`

`Mar. 15, 1938
`
`nicolet: biochemistry
`
`87
`
`it does work.
`to work very well, but
`should not be expected
`Heated for ten hours at 100° with benzyl mercaptan and a little
`piperldine as catalyst, it gave a small but definite yield of N-acetyl-
`S-benzylcysteine (VII). This was hydrolyzed to S-benzylcysteine
`(VIII).
`
`CH^CC02H + PhCH^H -+ PhCH2SCH2CHC02H
`I
`I
`NHAc
`NHAc
`VII
`HSCH8CH(NH2)C02H
`
`VI
`PhCH2SCH2CH(NH2)C02H
`VIII
`Dr. Loring was kind enough to apply to this Dr. du Vigneaud's
`process of reduction by means of sodium in liquid ammonia, for the
`removal of the benzyl group from sulfur, and obtained a 93 per cent
`yield of cysteine, as shown by the specific Sullivan method.
`This is a new synthesis of cysteine, and therefore of cystine. But
`if it is to be considered as
`certain other requirements must be met
`even a possible model for a biochemical synthesis of
`these amino
`acids.
`tripeptide such as V was not available,
`Since a dehydroalanyl
`benzyl mercaptan was next added to benzoyldehydrophenylalanyl-
`glycine ester (IX). The resulting mercapto derivative (X) was formed
`some 50 or 100 times more readily, as estimated by the much better
`yield obtained in a much shorter time.
`PhCH
`PhCHSCH2Ph
`PhCONHCCONHCH2C02Et
`PhCONHCHCONHCH2C02Et
`IX
`X
`Now peptides containing alanine are very common. It is extremely
`in their metabolism,
`in the presence of
`probable that,
`they pass,
`suitable enzymes and of suitable hydrogen acceptors
`(or oxidizing
`I have
`through the stage of dehydroalanine derivatives.
`agents),
`tried to show elsewhere6 that this dehydrogenation should take place
`more readily when the alanine was a component of a peptide chain,
`and particularly when it was not terminally located. Thus the most
`commonly formed dehydroalanyl derivatives should be just of the
`type most suitable for sulfide addition. We have thus acquired,
`through model experiments, the basis for a picture of the biological
`synthesis of cystine which is at least somewhat credible.
`As a sort of parenthesis,
`it might well be remarked here that a
`• Nicolet, Science 81: 181. 1935.
`
`Eton Ex. 1026
`4 of 10
`
`

`

`88
`
`JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES
`
`VOL., 28 NO. 3
`
`natural synthesis of serine should be based on just the same organic
`It is here merely a case of
`intermediates as the cystine synthesis.
`adding water, instead of sulfide.
`Silk proteins contain much more serine than any others. They are
`also conspicuously richer in alanine than most other proteins, and
`therefore should offer a richer source of dehydroalanyl derivatives.
`This is not, I think an accident
`I should now like to extend the ideas already advanced to the con
`sideration of the biochemical synthesis of methionine. Only plants
`can make methionine, for Rose has found it an essential amino acid
`for animals.
`The logical intermediate for its synthesis appears to be methylene-
`pyruvic acid (XI). It is a well known fact that plants have at their
`disposal for synthetic purposes formaldehyde and pyruvic acid. The
`condensation of these to form methylenepyruvic acid would be most
`orthodox. Here again we meet a substance which has never been
`isolated, presumably on account of its considerable lability.
`This is a hindrance to its use by a chemist, but not to its use by
`a plant. Formaldehyde has been condensed with pyruvic acid by
`various investigators, and under various conditions. The process has
`always gone too far, but in such a way as to indicate that methylene-
`acid, which would
`pyruvic acid, or possibly hydroxymethylpyruvic
`perhaps serve equally well, has been an intermediate.
`CH2O + CH3COC02H ?± CH2 = CHCOCO*H ^
`XI
`
`SCH,
`SH
`SCH3
`NH2
`! UCH2COC02H -> CH2CH2COC02H — CH2CH2CHC02H
`XII
`XIII
`Whether the addition reaction involves methylmercaptan, or hydro
`gen sulfide with subsequent methylation,
`is not at present con
`in contrast to animals, have a conspicuous capacity
`sidered. Plants,
`for such methylations.
`[ Since the desired substance was not available, recourse was had to
`another model experiment. Benzalpyruvic acid (XIV) was found to
`add mercaptans7 with the greatest ease, under the simplest conditions
`possible. With no added catalyst whatever, fairly quantitative addi
`tion was obtained in five minutes at 100°. It is considered that this is
`an obviously adequate rate of reaction to justify this stage of the
`J. Am. Chem. Soc. 57: 1098. 1935.
`7 Nicolet,
`
`i
`
`Eton Ex. 1026
`5 of 10
`
`

`

`nicolet: biochemistry
`Mar. 15, 1938
`89
`synthesis. It was further shown that the products had the structure
`indicated, since on oxidation they yielded the known /3-sulfones of
`£-phenylpropionic acid (XVII).
`
`I
`
`/
`
`SR
`PhCH = CHCOCO2H ^ PhCHCH2COC02H
`XIV
`XV
`NOH
`SR
`SO2R
`PhCHCH2C02H
`PhCHCH2CC02H
`XVII
`XVI
`This was the only really doubtful
`reaction in the series postulated.
`acid is an entirely reasonable product of plant
`Methylenepyruvic
`metabolism, and after the addition of the sulfur in whatever form,
`there results an a-keto acid such as, according to Knoop, should be
`readily converted to a methionine derivative by standard biological
`routines.
`The addition products (XV) of benzalpyruvic acid have been made
`with benzyl and p-tolyl mercaptans, and with thioacetic acid, and
`(XVI). Addition of hydrogen sulfide also
`identified as their oximes
`occurs readily, but no serious attempts have yet been made to isolate
`the product.
`You are asked to consider the hypothetical methionine synthesis
`thus briefly presented, not as a phantasy that begins and ends no
`where, but as a definite and coherent outgrowth of the more definite
`results already reported for cystine. It is at least possible that plants
`actually do make methionine in some such way.
`I now wish to change the subject again, and return to cystine.
`There remain to be discussed the questions as to why cystine is not
`freely synthesized by animals, and how it may be so synthesized.
`if we accept the picture offered a few minutes ago, it be
`Clearly,
`from
`the whole organic skeleton
`to assume that
`comes necessary
`which cystine is produced, exists regularly, and quite as a matter of
`routine, in animals as well as in plants. In plants, a ready synthesis of
`cystine results. In animals, the synthesis takes place, if at all, to a very
`there is fed either methionine or homo-
`insufficient extent, unless
`cystine. And yet, serine is not an essential amino acid. If the theory
`is even roughly correct, the difference is simply that the animal body
`lacks a suitable source of sulfur. The simplest possible source would
`be sulfide or disulfide ion— the time has not come to discuss whether
`
`Eton Ex. 1026
`6 of 10
`
`

`

`90
`
`JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES
`
`VOL. 28, NO. 3
`
`cystine or cysteine is first formed. But sulfides are definitely toxic
`rather efficient
`to animals, and the animal body has developed
`mechanisms for preventing their accumulation in any considerable
`concentration.
`them to
`Naturally, one cannot feed inorganic sulfides and expect
`arrive at the site of synthesis, and with the animal normal. But there
`might be some other method by which the necessary sulfur could be
`supplied.
`As already pointed out, we know two substances which are, up to
`the present, unique in being able to supply cystine deficiencies. We do
`not actually know that methionine and homocystine replace cystine
`by causing cystine formation, but this is certainly much the simplest
`assumption. The fact that homocystine is as effective as methionine
`the idea that it is an intermediate in the metabolism of
`encourages
`methionine. The further stages in the degradation of homocystine
`must be somewhat hypothetical. But Knoop's well known oxidative
`deamination would lead to the a-keto acid. This, according to the
`principles already discussed in detail, should yield (see, for instance,
`XII— *XI) reversibly hydrogen disulfide and methylenepyruvic acid.
`It must be recalled that this oxidative deamination should take
`just the time and place (perhaps the liver) at which the
`place at
`dehydrogenation of the alanyl peptides is going on, to produce the
`intermediates necessary for cystine formation. And here it is assumed
`that the cystine formation occurs.
`A purely inorganic deficiency would thus have been overcome
`through the addition of a rather complex organic compound. The
`thesis at this moment is essentially this: that methionine or homo
`cystine can cause the animal body to produce cystine, but without
`contributing
`the
`anything whatever
`to the organic skeleton of
`resulting cystine.
`Dr. Brand,8 who has also interested himself in this problem, has
`made certain suggestions,
`some of which he has tested by feeding
`experiments. None of the tested compounds have as yet produced
`cystine when fed. The most ingenious of his suggestions,
`to my mind,
`is that the intermediate between homocystine and cystine might be
`a mixed
`H02CCH(NH2)CH2CH2SCH2CH(NH2)CO2H
`sulfide,
`(XVIII). This substance has not yet been synthesized for feeding
`experiments, but I shall ask you to note two things about it. It is
`possible for it to break down according to a mechanism essentially
`like that already suggested,
`to produce cystine; and in that case the
`* Brand, Block, Kassell, and Cahill, Proc. Soc. Exp. Biol, and Med., 35: 501. 1936.
`
`Eton Ex. 1026
`7 of 10
`
`

`

`Mar. 15, 1938
`
`nicolet: biochemistry
`
`91
`
`latter would again have been formed without the contribution of any
`carbon atoms from the homocysteine portion. Furthermore,
`the only
`important change required in the reactions already outlined would be
`one of order. The sulfur of homocysteine would merely have to be
`come a part of the molecule which was to yield cystine, before the
`degradation of the homocysteine, instead of after it.
`I cannot at present offer you any further evidence
`Unfortunately,
`as to the course of these reactions. The work has not been completed,
`and you must share the responsibility for having made me discuss it
`at this stage. The current plan is to label the homocystine chain by
`If one of these,
`the synthesis of 7-methyl or 7-phenylhomocystine.
`when fed, should show the power to make good cystine deficiencies in
`a diet, it would be rather clear that its carbon skeleton was not being
`used for the synthesis.
`Dr. du Vigneaud has imagined a mechanism by which three carbon
`atoms of the chain in homocystine would be retained in the cystine
`to the extent that
`produced. We are in most
`friendly disagreement,
`I consider his mechanism for this particular
`reaction as possible
`rather than probable. He would perhaps make a similar statement
`about mine.
`He is now engaged in a subtler attack on this problem, which in
`volves labeling the homocystine or methionine to be fed, not with
`methyl or phenyl groups, but with strategically located atoms of
`deuterium. Neither of us would really ask for anything better than
`to have the question definitely settled, by whatever process, and with
`whatever
`result.
`Exactly how this information, when obtained, may be applied to
`the economic supplementing of deficient diets,
`is by no means clear.
`On the other hand, such a possibility exists, and there is also the
`chance of developing new tools for the study of metabolism.
`With the background which I have laid, I should like to shift the
`field of discussion once more. Dr. Csonka9 has somewhat recently
`made a very interesting generalization with regard to the behavior of
`the various amino acids occurring in proteins, when fed to animals
`which have been treated with phloridzin. He suggests the rule that
`those amino acids which are freely produced by animals, cause sugar
`formation under these conditions. On the other hand, "essential"
`amino acids, which animals are apparently unable to synthesize, or to
`formation.
`synthesize in adequate amount, should not cause sugar
`at Am. Soc. Biol. Chem., Washington, April
`• Csonka (unpublished): presented
`1936.
`
`Eton Ex. 1026
`8 of 10
`
`

`

`92
`
`JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES
`
`VOL. 28, NO. 3
`
`Rather generally, the results of experiment fit the prescribed pat
`tern. But cystine and methionine blur the picture somewhat. Both
`produce sugar. Now methionine is rather definitely, by present data,
`essential. Cystine I have already spoken of as "nearly essential," for
`I have given, though Dr. Csonka prefers to call
`it non
`reasons
`for reasons with which I cannot very fully agree.
`essential,
`The logic behind this generalization of Dr. Csonka's i
`seems to
`a given amino acid, in the course of its biological break
`me, this.I
`down, produces any essential substance in the chain of reactions by
`which the body normally breaks down and builds up sugars,i
`t will,
`in phloridzinized animals, cause sugar production.
`If one assumes, not unreasonably (although the logical necessity i
`not entirely complete), that most or all of these reactions are bio
`logically reversible, then the odds are very good that,i
`fagiven amino
`acid produces sugar under these circumstances, the normal metab
`olism of sugar will produce also the organic intermediates required
`to allow the body to produce the amino acid in question. Stated thus,
`tto be moder
`the principle hasasimple basis, and we should expecti
`ately general, as indeedi
`t appears to be.
`But therei
`s no reason at all whyi
`t should be completely general.
`To my notion, one might well expect a limited number of exceptions.
`These would become, not arguments against the generalization, but
`to explain why the observed variation exists.
`simply challenges
`Our earlier discussion of mechanisms of formation of cystine and
`methionine has already made cleara possible explanation of these
`two exceptions—the sulfur fragment required for synthesis. In the
`s an additional point involved, but therei
`case of methionine there i
`t here.I thinkI should personally have had
`not time to discuss i
`f these two acids, specifically, had not been
`definite cause to worry,i
`formers.
`sugar
`f we assume them to be thus explained or an
`These exceptions, i
`occasional additional exception, may perhaps add to the value of
`Dr. Csonka's rule as much as they detract from it. While the rule will
`have to be used somewhat tentatively in deciding whether a given
`amino acid i
`
`s oris not "essential," the exceptions themselves may
`occasionally help to decide among various mechanisms assumed for
`the synthesis, by animals, of various amino acids.
`If you now have the impression thatIhave discussed the problems
`of cystine sulfur exhaustingly, you may be right. Buti
`f you thinkI
`have considered them exhaustively, you are altogether wrong. There
`just for instance, the theory of Schoberl, buttressed bya certain
`
`it
`s,
`
`s
`
`s
`
`f
`
`is,
`
`Eton Ex. 1026
`9 of 10
`
`

`

`Mar. 15, 1938
`
`Mansfield: oligocene faunas
`
`93
`
`number of facts, and based almost entirely on "model" experiments.
`I shall later have arguments about that with some of you—but not
`tonight.
`Of those of you who have followed me thus far, I am sure that
`many will agree with me that biochemistry by analogy may be a most
`dangerous adventure. Certainly, any conclusions reached by such a
`process should be most carefully controlled by more direct experi
`ments, as soon as these become possible.
`But I can at least hope to have some of you agree that processes of
`the type outlined possess a considerable fascination, which is perhaps
`not merely that of their danger, and that they may at the very least
`be profitable for their power to suggest theories which must later be
`tested more rigorously.
`
`PALEONTOLOGY.—Oligocene faunas from the lower and upper beds
`on the A. L. Parrish farm, Washington County, Florida.1 W. C.
`Mansfield, U. S. Geological Survey.
`The locality from which the fossil material, on which this paper is
`based, was obtained is in a small ravine, the stream in which dis
`appears in the bottom of a sink, back of the house on the A. L. Par
`rish farm, about 3£ miles southeast of Wausau, Washington County,
`Florida. The first reference to this locality is given by Mossom (1).
`He reports twenty feet of cream-white soft limestone in a sink at this
`locality. He states that the contained fossils indicate, based on a
`letter of Miss Julia Gardner, that it is older than the Chipola marl
`and probably belongs to the Tampa. An analysis of the limestone as
`reported by Mossom, which probably is from the lower part of the
`showed a percentage of 94.7 of calcium carbonate. The
`exposure,
`limestones were not differentiated.
`The second reference is by Cooke and Mossom (2) . They state that
`the lower part of the limestone is white, finely granular, apparently
`pure, and contains few fossils; the upper part is more argillaceous and
`carries an abundant fauna, which appears to be of Tampa age.
`The third reference is by Dr. T. W. Vaughan (3a). Doctor Vaughan
`states :
`"The third collection, sink on A. L. Parrish farm, 3$ miles southeast of
`Wausau, Washington County, Florida, contains poorly-preserved specimens,
`but I think a definite opinion as to the geologic age is justified. One specimen
`the same species as Cushman's L. chattahoocheensis
`undoubtedly represents
`1 Received January 24, 1938.
`
`Eton Ex. 1026
`10 of 10
`
`