111
I\
The 5-amino derivative 111 was obtained by nitrosation of tropolone and then oxirnation and reduct i o n 2 Very recently the dibenzoyl derivative IV of l-amino-7-imino-l,3,5-cycloheptatriene was prepared from 1,3-diazaazulene by reaction with benzoyl ~ h l o r i d e . ~Attempted hydrolysis of IV yielded 2-phenyl-l,3-diazaazulene with either acidic or basic catalysts. The aniinoirnines I were obtained in 3 0 4 0 % yield by reaction of the tetrafluorocycloheptadienes with primary amines and with ammonia. I (R = H ) : (hemihydrate), 30y0 yield of yellow crystals from ether, m.p. 112-113’. Anal. Calcd. for CiHsN2.1/2H20: C, 65.09; H , 7.02: N, 21.69. Found: C, 65.01; H, 6.93; N, 31.54. I (R = H ) : CFICOOH salt. A n d Calcd. for C7H8N2CF3COOH:N, 11.96, Found: X , 11.92. I (R = CHs): 757, yield of yellow crystals from methanol, m.p. 66.5-67’. ilnul. Calcd. for C9HlzN2:C, 72.93; H , 8.16; N, 18.91. Found: C, 73.05; H,8.26; N, 18.92. 1 (R = p-c1c&): 70% yield of red crystals, n1.p. 1G8-1iO0. Anal. Calcd. for C1gH14NZC12: C, 66.87; H , 4.13; N, S.21. Found: C, 66.71 H, 4.22; N, 8.13. The infrared and ultraviolet spectral data are interpreted as being consistent with the proposed structures. Furthermore, nuclear magnetic rrsonance studies indicated that the nitroien atonisof the dimethylaminoimine I (R = CH3) are identical. Thus, either the hydrogen atom is shared equally between the two nitrogen atoms or there is rapid exchange of hydrogen between the atoms.4 The adjacency of the nitrogen atoms was further established by formation of stable chelates with Ni++, Co++, and C u + + ions, and by conversion of the aniinoirnines to tropolone. Although stable to strong aqueous acids or bases, the dimethylaminoiniine I (I< = CHJ reacted with amines in dioxane coritaining a catalytic amount of acetic acid to yield new aminoimines derived by displacement of the methylamino group.
Ring substitution has been effected by various electrophilic reagents. In contrast to tropolone, which gives a mixture of mono-, di- and tribromo derivatives, reaction of I (R = p-CICeH4) with one mole of bromine gave a single isomer in 90% yield, red crystals, m.p. 181-182’. It was established
.
--.
‘NI11~
-___-
(2) ’r. Nozoe, &Sdta, I. 5. [to, kI. hlalsiii illid ’I, hIatsiida, PrL‘C. J ~ f i a nA c u d . , 29. 565 (1953). ( 3 ) I . hfurata. BILU.C h e ~ nSnc. Jiifmrt, 32, 841 [ l ! I S l ) . (4) We are indebted t o Drs. W. D. Phillips and D . B. Chesuut for this determination and interpretation.
iridirectly that substitution had occurred a t the 4position by condensation of 3,4-dibromo-6,6,7,7tetrafluorocycloheptene with fi-chloroaniline to obtain a monobronioaIiiiiioiriiiiie identical with the product obtained from the bromination reaction. The l-amino-7-thioxo-l,3,5-~ycloheptatrienes I1 were obtained by action of hydrogen sulfide on the aniinoimines I. The N-substituted derivatives are stable compounds, soluble in dilute acids and insoluble in strong bases. Kuclear magnetic resonance studies of the nitrogen atom indicate that there is little, if any, contribution to the structure by the iminothiol tautomer.4 Chelate derivatives were formed with Cu++,Ni++ and Co++ions. I1 (R = H), 0 0 % yield of yellow crystals from methanol, m.p. 137-138.5’. A n d . Calcd. for CiHiNS: C, 61.27; €1, 5.15; N, 10.21; S, 23.37; mol. wt., 137.2. F o u n d : C, G1.20; H, 5.15; K , 9.94; S,23.40; m9!. w t , , 122. ‘The corresponditig N-inethyi xiid N-p-tolyl derivatives have been synthesized and the observed analytical data are in accord with the proposed structures. These and other reactions of the aniinoiiiiiries I and the amiriothioxocycloheptatrienes I1 will be fully reported shortly. CONTRIBUTION S o . 591 moni \V,R. BRASEN CETTRALRESEARCH DEPARTMEST H. E, HOLMQUIST EXPERIMEXTAL STATION R. E. BEKSON E. I. DU PONTDE KEXOCRS AND ConwAss n‘ILMINGTON, DELSWARB
RECEIVED JANUARY 9, 1960 QUANTITATIVE ASPECTS OF NUCLEATION I N CALCIUM PHOSPHATE PRECIPITATION’
Sir: After many many years of a single-minded view, prompted primarily by the theories of Robison,2 a new and different approach to the problem of biologically induced calcification has developed. Currently, it is almost universally recognized that collagen fibers possess the unique property of inducing crystals of calcium phosphate to f ~ r m . ~ , ~ This change in viewpoint resulted primarily from a clarification of the solubility properties of bone mineral it~elf.5~6It is now clear that normal serum is highly supersaturated with respect to bone mineral if solid phase is firesent as it always is in the living animal. Bone mineral prototypes (hydroxyapatites) and dead bone powders evince widely varying dissolution products (Ca X P inorganic) which are strongly dependent upon experimental conditions, but rarely if ever are products greater than 10 (Ca X P i in (mg.%)2)encountered. Living bone in culture, however, will support higher products7**approaching the physiological (ca. 20). (1) This paper is based on work performed under contract with the United States Atomic Energy Commission a t T h e University of Rochester Atomic Energy Project, Rochester, N e w York. ( 2 ) R. Robison, “The Significance cf Phosphoric Esters in Metabolism,” S e w York University Press, S e w York, N. Y., 1932. (3) W. l>.Neurnan and h i . Neuman, “Tlie Chemical Dynamics of Bonr Mineral,” T h e University of Chicago Press, Chicago, Ill., 1958. (4) 31. J. Glimcher, Rev. B l o d c r r . Phys., 31. 31,9 (1959). ( 5 ) 8 . s. Strates. W. P . N e u m a n a d G. J. Lerinskas, J . Piiyr. ( / i e v z . , 61, 27‘3 ( I X I ? ) . (li) G . 1. 1,evinskus an,! \i.I ? . Nriiruan. i h i d . , 5 9 , 101 ( l W i , j j . ( 7 ) R. E, C. Nordiii, J . J3 . C h e u ! . , 227. 551 (1957). ( 8 ) C. A. Bassett and B. . C. Nurdin, A d a Orlkopard. Scand. 28, 211 ( l Y 6 Y ) .
COMMUNICATIONS TO THE Eurruw
Feb. 20, 1960
63
TABLE I INDKCTION OF CRYSTAL FORMATION BY COLLAGEN Determination of the point of precipitathn was accomplished by techniques published earl$r (Fig. 1, ref. 5 ) . Conditions were p = 0.16; t = 3'7 ; PH 7.4; tinie of equilibration, 3 days. ---Point so. expts.
Mean
50
of precipitationStandard Standard deviation deviation of mean
Control solutions 7 50 4.1 Collagen-seeded 11 20 4.7 a Expressed as Ca X P, in (mg. % j z .
1.6 1.4
4';
(397
-
j
r- i
4
t-
REGIOS OF mTASTA6I LITY
1
I
L-
I
CB x
pi
PRECIPITAYIUN
'
KSp
CaHPO+
REGION OF LTDERSAT'Y I N ABSENCE OF
i
CATALYZED INDUCTION
SERUM LEVELS Presumably, this phenomenon results from localSOLUBILIZING ACTION ized acid production by the living bone cells themOF CELL SECRETION selve~.~.~ In the absence of preformed solid phase, it is the I Ksp of CaHP04.2H20 which seems to govern the B O W hlINERAL stability of the aqueous calcium: phosphate ~ y s t e r n . ~ Spontaneous precipitation never has been described sutmnary of prcserit inforn~ationcmiccrning Fig l.-!i in solutions having products, Ca X P i , less than diisolution and formxtion o f hone niineral given in terms of about 35 (mg. % ) 5 although varying degrees of metastability have been observed (depending the serum prcduct, Ca X P,, in (mg. yo)*. upon conditions) with solutions having products of physiological concentrations of calcium and as high as 50. These general findings are sum- phosphate. These data are summarized in Table marized in Fig. 1. 1. Until now the idea that collagen can function as The case for collagen now seems complete. It a crystal nucleator in vivo has suffered from a serious can specifically cause crystal formation under quantitative defect. Crystal induction by col- physiological conditions. Whether it does this in lagen has been demonstrated in several labora- viao and how it accomplishes this remain questions has been shown to be yet to be answered by further research. t o r i e ~ . ~ s ~The ~ , ~process l exquisitely specific4J0 and morphologically the DEPARTMENT OF RADIATION BIOLOGY OF MEDICINE A N D DENTISTRY early phases of crystal induction are very similar SCHOOL HERBERT FLEISCH OF ROCHESTER both in vivo and in v i t r ~ . ~Nonetheless, ~~'~ until THEUNIVERSITY ~VILLIAMF. XEUMAN YORK 20, SEW the present, these examples of crystal induction ROCHESTER RECEIVED JANUARY 11, 1960 have all involved solutions in the region of metastability (>K,, CaHP04.2ELO) involving products of Ca X Pi from 35 to 50 (mg.%)2.3t5J1J3Yet to PROTON RESONANCE SPECTRUM AND STRUCTURE be functional in the living animal, collagen must OF THE AZULINEUM I O N be shown to induce crystal formation a t physiologi- Sir: cal products (ca. 20), a region of undersaturation The structure of the azulinium ion is of considerwith respect to solid formation, able interest in concection with the electron charge Quite recently,I4 crystal induction has been distribution in azulene itself. Protonation of demonstrated with demineralized, fresh dentin azulene in acid solutions may be expected to occur a t products, Ca X P i , of 20, well below the critical a t the 1 and 3 positions, which, according to K,, CaHP04.2H20. This represented the first molecular orbital calculations,l have the greatest occasion, to our knowledge, of crystal nucleation excess a-electron charge density. It is also known by a non-living system under physiologically mean- that the protons a t the 1 and 3 positions of azulene ingful conditions. The tissue residue employed, are readily exchanged in D,S04 solutions.2 We however, undoubtedly contained substances other have now been able to confirm the structure of the than collagen. Hence, the case for collage11 as azulinium ion on the basis of a high-resolution the nucleator, though greatly strengthened in a proton resonance spectrum. A previous attempt3 quantitative sense, was still not proven. In experi- to determine the structure from the proton resoments just completed, a collagen'j prepared by the nance in concentrated sulfuric acid was unsuccessful method of Einbinder and SchubertlG also has due to inadequate resolution. proved able to cause crystal formation in solutions Figure l a shows the proton resonance spectrum (9) G. R. Martin, H. E. lirschein and W.I'. Keumun, THIS JOUR- of a 7 mole % solution of pure azulene in CCL. N A L , 80, 6201 (1958). A scale in c.!s., referred to the resonance signal of (10) M. I. Glirncher, A. 1. Hodge and IF. 0. Schrnitt, Proc. K a f l . CH2C12 (dotted line), is shown a t the bottom of the A c a d . Sci. U .S., 43, 860 (1957). figure. The assignment and analysis of the azu(11) B. N . Uachra, A. E. Sobel and J. W.StanFord, A r c h . Biochem. B i o p h y s . , 84, 79 (1959). lene proton spectrum has been described pre(12) S. Vitton Jackson and J. T. Randall, " R o n c Structure and viously.4 Shown 011 the sanic' scale ill Fig. l b is hIetaboli5ni," Ciba Fotinii;ition Symposium, ed. G . E. IV. W ~ ~ l ~ t c n l i o l i n e the proton resoiiance spectruiii of a '7 iiiole yo and C. U. O'Connor, London, 1956.
I
( 1 3 ) hl. Lanini and W. 17. Neuman. A r c h . Z ' u f h d , 66, 20.1 (!!l38), (11) C . C. Svlonions and W. F. Neutiiiiri, in preparatic,n. (1:) Upon analysis, this prcyaration was e5bentially C;r free, certainly less than 0.1 p . p , m .
(16)
I. Einfinder and
M. Schubert,
J. B i d . Chem., 188, 335 (1951).
(1) R . Pat iser, J . ' ~ ~ P I I PZ h. y s . , 25. I I l ? (ill3i). ( 2 ) J3. IIcilbrotiner, private cotntii~itiicati~~n. ( 3 ) H. hI. Prey, J . C h e m Phys., 25, GUI) (196fj). (4) W. G. Schneider, H. J. Eerristein and J. A . Poplc, NAL, 80, 3497 (1958).
1'111sJ U I , K -
