STRUCTURAL STUDIES ON ANTIBIOTIC NETROPSIN
May 20, 1956
[CONTRIBUTION FROM THE
DEPARTMENT OF
2157
CHEMISTRY O F THE UNIVERSITY OF WISCONSIN]
Structural Studies on the Antibiotic Netropsin BY EUGENEE.
VAN
TAMELEN, DWAINM. WHITE,IRVING C. KOGONAND A. D. G. POWELL RECEIVED OCTOBER24, 1955
The antibiotic netropsin is a diacidic base possessing the molecular formula ClsHzsNloOs and is convertible by hydrolysis to one mole each of ammonia, guanidinoacetic acid and the monobasic netropsinine ( C I ~ H ~ I N B O ~ ) .
Several years ago Finlay, Hochstein, Sobin and normally effected, or, less likely, by cyclization of Murphy' described the isolation and characteriza- the guanidinoacetic acid during the isolation or tion of a new antibiotic, Netropsin,2 obtained from purification process. the culture filtrates of the Actinomycete StreptoThe C16-basereferred to above was characterized myces netropsis. Since the publication of the early in this Laboratory as the monopicrate, m.p. 216', work, we have had the opportunity to reinspect the and the monohydrochloride, m.p. 200' dec., as well chemical nature of the antibiotic: the present as by a monobenzoyl derivative, m.p. 271', and a Electrometric communication describes its further characteriza- diacetyl derivative, m.p. 157-158'. tion as well as the results of certain preliminary titration of the free base (PHb 6.8) evidenced only one basic center, thereby making the equivalence degradation studies. Mainly on the basis of the analysis of the sulfate, of the molecular and empirical formulas for netropsithe approximate molecular formula C ~ ~ H ~ Bwas N ~ Snine O ~ virtually a certainty. Since the decomposition affording guanidinoaceassigned to netropsin by the earlier workers. While securing additional analytical data, we observed tic acid and the &base would in all probability that the sulfate was hygroscopic and seemingly constitute the starting point for subsequent strucfirmly solvated, and, depending on the extent of tural studies, the reaction was studied in some dedrying, the nitrogen analyses, for example, varied tail. It was determined that neutralization of ne-, between 24 and 26.5%. We found, however, that tropsin sulfate a t room temperature liberated one the limiting values obtained on exhaustively dried mole of ammonia (observed value 0.96 mole) and portions of the sulfate were fairly constant and indi- that hydrazine, hydroxylamine and nitrogen were not formed in this operation. We were not successcated the formula C I S H ~ ~ N I O O ~ .(or H ~somewhat SO~ better, ClaH2eNlaOs.'/zHz0.H2S04).This assign- ful in isolating any degradation products other than ment was confirmed by analyses on the dipic- those indicated, all of which were formed in good rate and the dihelianthate of netropsin as well as yield. by the additivity of the molecular formulas of the Drastic basic hydrolysis of netropsin sulfate, degradation products obtained on mild basic hy- carried out by refluxing in excess aqueous sodium drolysis (vide infra). Netropsin sulfate gives a neg- hydroxide for eighty hours, yielded 4.03 moles of ligible quantity of acetic acid in the Kuhn-Roth ammonia, on the basis of the C18H26N1003 formula. determination for C-methyl groups. The assumption that guanidinoacetic acid should Finlay, et al., described the degradation of ne- yield two moles of ammonia under similar hydrolytropsin sulfate brought about by the neutraliza- sis conditions was confirmed by the observed value tion with an equivalent amount of alkali, resulting of 2.03 moles, and i t was further determined that in the formation of two bases: one, a water-soluble the &base affords exactly one mole under such compound possessing the empirical formula C3H6- circumstances. Thus the number of moles of amN80; the second, a water-insoluble substance, the monia surrendered by the degradation products, analysis of which indicated the composition c16- taken together with the single mole liberated during HzaNaOa. Repetition of the basic decomposition the mild, basic fragmentation of netropsin sulfate, invariably gave, in our hands, a product C3H7Ns02, corresponds well with the total value liberated by along with the Clb-base, the analysis of which we netropsin sulfate itself. were able to confirm. The smaller fragment, obDuring the course of this investigation, several tained either by direct crystallization or through an observations of incidental interest were made. The ion-exchange procedure, has now been identified as addition of an equivalent amount of sodium methguanidinoacetic acid. We can account for the iso- oxide in methanol to netropsin sulfate followed lation by the earlier workers of the product C3H6- by brief heating served only to convert the antibiotic salt to the &base; furthermore, the latter was NH stable to the same reagent on refluxing in methanol /I for short periods of time. In attempting to subNHz-C-NH CHiCOOH stantiate the report that netropsin sulfate gives a N30-probably glycocyamidine-only by assuming positive Ehrlich test, we noted that no color dea milder degradation of netropsin sulfate than we veloped a t room temperature, although warming resulted in a blue-violet, rather than the usual pink, I coloration. Since netropsin is so readily cleaved by base, the positive Sakaguchi test afforded by the O// \"/'\" antibiotic' may well be due to guanidinoacetic acid (1) A. C. Finlay, F. A. Hochstein, B. A. Sobin and F. X. Murphy, which is liberated in the medium used for the test, THISJOURNAL, 13, 341 (1951). viz., excess aqueous alkali. Because of these com(2) Netropsin is the trademark of Char. Piker and Co., Ine.. for plications, the deductions which can be made on the antibiotic produced by Strrptomwsr nclro#rir.
p'-"
215s
E. E.
VAN
TAMELEN, D. M. WHITE,I. C. KOGONAND A. D. G. POWELL
1'01.
is
tion containing the hydrochloride was evaporated to a volume of several milliliters and applied to a water-washed column prepared from Dowex-1 in the basic phase. The free base could be eluted with water, and the solution thus obtained was evaporated to dryness a t room temperature under reduced pressure, thereby yielding a colorless, crys talline product. Guanidinoacetic acid also could be obtained by direct crystallization. The aqueous filtrate obtained, after removal of the Cls-base, by decomposition of 9.8 g. of netropsin sulfate, was reduced to a volume of 20 ml., filtered and allowed to stand in the refrigerator for 48 hours. The colorless, crystalline precipitate which had deposited n a s filtered off and washed successively with a small amount of water, ethanol and ether. After drying in a vacuum desiccator, the product weighed 1.08 g. (50% yield). Anal. Calcd. for C ~ H ~ N J OC,~ : 30.79; H, 6.03; K , 35.88. Found: C, 31.28; H , 6.06; N , 35.35. TABLE I The infrared spectra, measured in Kujol mull, of t h e Compound Solvent Lmsx m r log 6 above base and of authentic guanidinoacetic acid were inNetropsin sulfate O . l L V H C l 236,297 4 . 3 0 , 4 32 distinguishable. The guanidinoacetic acid isolated as described above mas Clj-base O.1NHCI 283 4.33 converted t o the picrate, m.p. 201' dec., and the hydroEthanol 243,304 4 1 4 , 4 . 3 3 chloride, m.p. 189-190' dec. The melting points of the Monobenzamide of corresponding salts of authentic guanidinoacetic acid melt, CIs-base Ethanol 234,303 4 . 3 7 , 4 . 4 8 respectively, a t 201-202' dec. and 191" dec.8 The appromixed melting points showed no depression. To Diacetate of C1,-base Ethanol 240,301 4 , 1 4 , 4 23 priate complete the identification, the guanidinoacetic acid obfrom netropsin was converted to hydantoinimine hyAcknowledgment.-Financial support was pro- tained drochloride,a m.p. 209-210" dec. E o depression in melting vided by the National Microbiological Institute, point was observed on admixture with authentic hydantoinDepartment of Health, Education and Welfare imine hydrochloride, m.p. 211-212" dec. Again, the in(Grant No. G-4070). The authors are indebted to frared spectra (mulls) of the two salts were identical. Determinations of Ammonia Released During Basic HyN r . A. C. Finlay of Chas. Pfizer and Co., who pro- drolyses. Neutralication of Netropsin Sulfate.-One gram vided the netropsin sulfate used in this investiga- of well-dried netropsin sulfate was treated with 20 ml. of tion. 0.103 N sodium hydroxide in a closed system. During a 24-hour period, a slow stream of nitrogen was passed through Experimental the reaction mixture, the ammonia entrained being trapped Salts of Netropsin.-For the analytical determinations, in 5% boric acid solution and titrated with standard hydronetropsin sulfate was repeatedly recrystallized from hot chloric acid (brom cresol green-methyl red indicator). -4 water, from which solvent, on cooling, it was deposited as total of 0.96 mole of ammonia was detected. well-formed, colorless needles. The individual samples Since no water-insoluble gas was liberated during the were dried in vacuo a t 80-125" to constant weight. A large neutralization of netropsin sulfate, free nitrogen was not a number of analyses were obtained, and the values on samples degradation product. Furthermore, color tests with salidried a t higher temperatures (100' and up) fell between cylaldehyde4 demonstrated the absence of hydrazine or hynarrow limits. The results cited below are the averages of tlroxylamine among the hydrolysis products. the more satisfactory runs. Drastic Basic Hydrolysis of Netropsin.-Using either esAnal. Calcd. for C ~ ~ H ~ ~ N ~ O O ~C,. H40.90; ~ S O ~H: , cess aqueous 0.5 N barium hydroxide or sodium hydroxide, netropsin sulfate was hydrolyzed by refluxing until no more 5.30; S, 26.51; S , 6.06. Calcd. for C I ~ " ~ N I O O ~ J / ~ H ~ O. H2S04: C, 40.22; H , 5.44; N, 26.07; S , 5.96. Found: ammonia vr-as released (60-80 hours of refluxing was required for completion). During the hydrolysis the ammonia was C, 40.70; H, 5.44; N,25.87; S,6.20. swept through with nitrogen and determined as described Setropsin sulfate gave, in the Kuhn-Roth determination, ahove. The highest value for moles of ammonia released 0.1% C-methyl. was 4.03. Netropsin picrate was prepared by adding a n excess of Drastic Basic Hydrolysis of Guanidinoacetic Acid.-Using aqueous picric acid t o a hot solution of purified netropsin sulfate in water. Since the picrate decomposed during the conditions described for netropsin, guanidinoacetic acid was hydrolyzed completely in 60-75 hours, affording 2.03 attempted recrystallizations, it was merely dried t o constant moles of ammonia. weight (56" in vacuo) and analyzed. The salt decomposes Drastic Basic Hydrolysis of the C15-Base.--Under similar at 232O, after browning and sintering a t about 225'. circumstances, the Cls-base released exactly one mole of Anal. Calcd. for C18H2eNlo03.2CsHsN307: c , 40.54; ammonia. H , 3.60; N, 25.23. Found: C, 40.72; H , 3.64; N,25.31. Salts and Derivatives of the Cx-Base.--As obtained from Netropsin helianthate, obtained by the addition of a the aqueous basic decomposition of netropsin sulfate, the warm aqueous solution of methyl orange t o a warm aqueous C15-base is a pink microcrystalline solid, crystallizable solution of netropsin sulfate, was purified by repeated crys- from hot water, after which the melting point is 253-255" tallization from aqueous ethanol. The orange plates melted d w . The base also can be obtained by treatment of a t 215' dec. nistropsin sulfate with an equivalent amount of sodium Anal. Calcd. for C l ~ H 2 6 h ' ~ ~ 0 ~ ~ I ~ ~ 0 ~ 2C,C ~ ~ H ~ ~ N ~inOmethanol. ~S: methoxide After being heated for 5 minutes, 52.17; H , 5.52; 9,21.16. Found: C, 51.87; H , 5.47; ice was added. a f t e r reheating and subsequent coolirlg, S,20.86. the product precipitated as a pale pink powder. Thc analytical data on purified material substantiate the CljHzoIsolation and Identification of Guanidinoacetic Acid.TsOa formulation of Finlay, et al.' Setropsin sulfate (1.OO g.) was decomposed with an equivaThe picrate, m.p. 216" tlec., was securcd through the use of lent amount of alkali according t o the previously published method.' The C15-base, which precipitates out com- aqueous picric acid. pletely on standing overnight, was filtered off. The aqueAnal. Calcd. for C ~ ~ H ~ O K O O ~ . @ ~C,H :41.02; S ~ O ~ 13, : ous filtrate, which contained the guanidinoacetic acid, was 3.32; N, 21.27. Found: C, 41.35; H , 3.97; K,21.00. applied t o a column made up from Dowex-50 ion-exchange resin (acid form). On passing dilute hydrochloric acid (3) Korndorfer, Arch. Phavm., 242, 629 (1905); H. King, J . Chem. Soc., 2375 (1930). through the column, guanidinoacetic acid was eluted as the (4) F. Feigl, "Qualitative Analysis by Spot Tests," Nordemaon hydrochloride; the qualitative detection was accomplished through teetn rising Sakagwhi reagent. The aqtmous snluPublishing Co., l a c . , N e w ITork. N. Y., 1937, p. 149-162.
the basis of these color tests are limited. Finally, the high degree of similarity between the ultraviolet spectra' of netropsin sulfate and the C15-base signifies that the chromophoric system present in the former has been largely preserved during the basic degradation to the latter and that the change involved is therefore probably not a deep-seated one. It is clear from the foregoing that netropsin is composed only of the guanidinoacetic acid and the CI~H~ON moieties, ~ O ~ linked together with the incorporation of a molecule of ammonia, and that the structural problem involves in essence the nature of the C15-base and its attachment to the other two components.
May 20, 1956
REACTIONS OF ETHYLENIMINES : DISSOCIATION
The hydrochloride, prepared by solution of the base in 0.1 N hydrochloric acid and subsequent evaporation in a desiccator, melted a t 200’ dec. Anal. Calcd. for CI,H2oN6O3.HC1: C, 48.84; H, 5.74; X, 22.79; C1, 9.61. Found: C, 48.81; H , 5.72; N, 22.71; C1, 9.60. The monobenzamide was prepared by adding 0.15 ml. of benzoyl chloride to 400 mg. of base suspended in pyridine (4.0 ml.). After standing for two hours, the reaction mixture was worked up by addition of 20 ml. of ether, which resulted in the deposition of a gummy solid. After being washed with ether and triturated with chloroform, the solid was crystallized from ethanol. The amide
2159
(230 mg.) was secured as colorless needles, m.p. 271’. Anal. Calcd. for C22H24N604: C, 60.59; H, 5.53; K , 19.28. Found: C, 60.51; H, 5.50; N, 19.22. The diacetyl derivative of the Cls-base was obtained by acetylation with isopropenylacetate in glacial acetic acid, the reaction mixture being refluxed for one hour. T h e product, isolated by ether precipitation, was recrystallized twice from ethanolic ether, after which it melted a t 157158’. Anal. Calcd. for C ~ ~ H ~ & ~ ~ S ( C H ~CC, O54.70; )?: H, 5.80. Found: C, 54.78; H , 5.81. MADISON, WISCOSSIS
~~
[COYTRIBUTION FROM
THE
METCALFLABORATORIES, BROWNUNIVERSITY]
Reactions of Ethylenimines. VIII. Dissociation Constants BY CHARLES E. O’ROURKE, LEALLYN B. CLAPPAND
JOHN
0. EDWARDS
RECEIVED OCTOBER 7 , 1955 Ethylenimine, 2,2-dimethylethylenimine and 2-ethylethylenitnine were found to be weak bases with PKB values of 5 99, 5.36 and 5.69, respectively.
In spite of the fact that ethylenimines(aziridines) have been known since 188S1no apparent attempt has been made to determine their basic strengths in aqueous solution. The problem was avoided in these laboratories for a long time because i t appeared that hydrolysis would make such a measurement difficult by the usual techniques. But a hint was available in the literature2that the reaction with water was slow and our own recent kinetic work on the hydrolysis of ethylenimines3 in acid made the measurement of base strength appear feasible. The dissociation constants of ethylenimine, 2,2dimethylethylenimine, 2-ethylethylenimine and three related amino alcohols have been determined by use of the glass electrode PH meter. While the dissociation constants measured thus and presented in Table I are not true thermodynamic values, the relative basicities may be determined quite precisely. In this actual case, the values of K B may be close to the thermodynamic values since the values of Ka for the conjugate acids of the imines and amino alcohols were found to be independent of ionic strength in the range from 0.01 to 0.1. The constants obtained for the three amino alcohols compare favorably with the thermodynamic values given by Glasstone and S ~ h r a r n . ~ TABLE I DISSOCIATIOX COWSTANTS OF ETHIZENIMINES ASD AMIXOALCOHOLSAT2j0 Compound
Ethylenimine 2-Ethyleth ylenirnine 2,2-Dirnethylethylenimine
Ethanolamine 2-ilmino-1-butanol 2-Methyl-2-amino-1-propanol Ref. 4.
~ K B
(obsd.)
PKna
RH
6 . 9 9 1. o x 2.0 X 5.36 4 . 3 X 4.56 2 . 8 X 4.45 3.3 x 4.29 5 . 3 X
(lit.)
10-6
j.69
10 -0 10-5 10-5
4.55 4.45 10V5 1 . 2 8
(1) S. Gabriel, Ber.. 21, 1049 (1888). ( 2 ) H. Freundlich and W. Neumann, Z p h y r i k . C h m , 87A, 69 (19141, and later papers of Freundlich. (31 V. B Schatz and 1,. R. Clapp, THISJ O U R N A L , 77, 5: 13 ( . 9 5 5 ) , (4) S. Ciasatone and A, F . MchramI i b i d . , 6 0 , 1213
