NOTES
490
VOL. 29
tion from chlorobenzene afforded the red-brown l-chloro-4-(m bromonaphthyl)anthraquinone, m.p. 260-264'. Anal. Calcd. for CZ4Hl2BrC1O2: C, 64.5; H , 2.7; C1, 7.96. Found: C, 63.8; H, 2.8; C1, 8.05. 0 N,BF, Decomposition of 1-Nitro-4-diazonium Fluoroborate Anthraquinone in o-Dichlorobenzene to Yield l-Nitro-4-(o-dichlorophenyl)anthraquinone.-l-Nitro-4-aminoanthraquinone (50 9.) was diazotized with nitrosylsulfuric acid solution and converted to 1-nitro-4-diazonium fluoroborate anthraquinone, 60 g., S9%, decomposition temperature of 171-172", in the manner described. o x The diazonium salt (20 9.) was suspended in o-dichlorobenzene x = c1, so2 (150 ml.), brought slowly to boiling, and further treated as in the above cases. The solution was reduced in volume to 30 nil. o i and filtered. The solid, 12.1 g. (Found: C1, 7.7% or 4370 of arylation), was recrystallized from chlorobenzene to yield 7.3 g., acid yield 1-aryl-4-aminoanthraquinones. Identical 37%, m.p. 241-244'. amines are obtained by causing the chloro compound to Anal. Calcd. for C2oH9C12SO4: C , 60.3; H , 2.25; C1, 17.85; N , 3 . 5 . Found: C,60.8; H , 2 . 3 ; C1, 17.3; ? ; , 3 . 7 . react with ammonia in an autoclave.s The l-aryl-41-Nitro-4 - bromopheny1anthraquinone.-1 -Nitro - 4 -diazonium nitroanthraquinones on heating with aqueous sodium fluoroborate anthraquinone when decomposed in bromobenzene sulfideg afford the corresponding amines in good yield. gave 39% of arylation (nitrogen analysis). The pure red-brown These appear to be the first arylated a-aminoanthra1-nitro-4-bromophenylanthraquinone isolated after recrystallizaquinones prepared and could be valuable intermediates tion from chlorobenzene had m.p. 295-298'. Anal. Calcd. for CzoHloBrNOa: C, 58.8; H, 2.45; K,3.45. for new anthraquinoid dyes. Found: C , 59.3; H , 2.5; N, 3 . 4 . l-Amino-4-(o-dichlorophenyl)anthraquinone.-l -Chloro-4-(0Experimental dichloropheny1)anthraquinone (6 g.), toluene-p-sulfonamide ( 4 g.), copper acetate ( 2 g.), potassium carbonate ( 2 g.), and oDecomposition of 1-Chloro-4-diazonium Fluoroborate Anthradichlorobenzene (100 ml.) were heated slowly to the boiling point quinone in o-Dichlorobenzene to Yield l-Chloro-4-(o-dichloroand kept under reflux for 5 hr. After cooling the mixture was pheny1)anthraquinone .-l-Chloro-4-amino-anthraquinone (50 g .) filtered, and the precipitate washed with boiling alcohol and water, was dissolved in concentrated sulfuric acid (200 ml.) and diazoand dried. The resulting anthraquinonesulphonamide was retized with nitrosylsulfuric acid solution (200 ml., prepared from crystallized from anisole to give 6 . 9 g., 8 5 % , m.p. 268-269". 20 g. of sodium nitrite and concentrated sulfuric acid) in 1 hr. a t The sulfonamide ( 4 9.) was dissolved in concentrated sulfuric 0-10". The resulting solution was stirred until on the addition of It was poured on to acid (40 ml.) and heated for 1 hr. a t 80-90'. several drops to water a clear orange solution was formed. ice to yield a red precipitate of l-amino-4-(o-dichlorophenyl)The product was poured onto ice (1 kg.). The anthraquinone anthraauinone, which was recrystallized from anisole to yield 2.6 diazonium sulfate was precipitated, removed 'by filtration, g., 93%, m.p. 283-284'. washed with a small amount of cold water, and dissolved in 5 1. Anal. Calcd. for C2oH11CLN02: C , 65.9; H , 3.35; C1, of water, insoluble impurities being removed by filtration. 19.25: N , 3 . 8 . Found: C , 66.0; H , 3.45; C1, 19.9.5; S , Fluoroboric acid (40c/;, 50 nil.) was added with stirring and the 3.9. precipitated 1-chloro-4-diazonium fluoroborate anthraquinone Reduction of l-Nitro-4-(o-dichlorophenyl)anthraquinone.-1(50 g., sac/,) was removed by filtration, washed with methanol, Nitro-4-(o-dichlorophenyl)anthraquinone ( 2 . 5 9.) was ground to a and dried (decomposition temperature, 180-181 "). paste with sodium sulfide nonahydrate ( 5 g.), suspended in hot The diazonium salt (20 g.) was suspended in o-dichlorobenzene water, and maintained a t 70-80" with stirring for 90 min. The (150 ml.) and heated slowly to boiling while being stirred. l-amino-4-(o-dichlorophenyl)anthraquinone was removed by During the decomposition the solution turned from red to filtration, washed with hot water, alcohol, and ether, and rebrown and a heavy stream of boron trifluoride escaped. rZfter crystallized from anisole to give brilliant red crystals, 2.1 g., 2 hr., the solution was decolorized with activated carbon and 9170, m.p. 281-284" filtered while hot and the o-dichlorobenzene solution was reduced Anal. Calcd. for C20HllC12S02: C1, 19.3; N , 3 . 8 . Found: in volume to 30 ml. (Found: C1, 19.2YGor 61c/, of arylation.). C1, 19.4; X , 3.85 The yellow product which separated on cooling was recrystallized Similarlv. 1-nitro-4-bromo~henvlanthra~uinone ( 2 R . ) was " , .~ from chlorobenzene to yield l-chloro-4-(o-dichlorophenyl)anthra- converted to 1-amino-4-bromophenylanthraquinone (1.8 g., quinone, 10.5 g., 49c/L, m.p. 267-268". go%), m.p. 285-290'. Anal. Calcd. for C20HaC1,02: C, 61.9; H , 2.3; C1, 27.5. Anal. Calcd. for C20H12BrS02: C, 63.5; H , 3.2; N , 3.7 Found: C , 61.85; H , 2.4; C1, 27.4. Found: C, 62.8; H , 3.1; N, 3.8. 1-Chloro-4-(~-chloronaphthyl)anthraquinone .-l-Chloro-4-diazoniuni fluoroborate anthraquinone (15 9.) was suspended in a-chloronaphthalene (100 ml.) and slowly heated to 180-190' and kept a t this temperature for 1 hr. with stirring. DecolorizThe Synthesis of 4',5'-Diiodo-4-aminoing carbon was added, the solution was filtered, reduced in volume fluorescein Iodine-131 to 10 ml., and then boiled in chlorobenzene (100 ml.). On cooling the solution a yellow precipitate was obtained (Found: C1, 15.55; or 48r; of arylation.) which was recrystallized from ANTHONY J. VERBISCAR' chlorobenzene ( 6 g., 3 5 c / , ) , m.p. 274-275'. Anal. Calcd. for C2dH12C1202: C, 71.5; H , 3.0; C1, 17.6. Argonne Cancer Research Hospital, Chicago, I l l i ~ i s Found: C , 71.75; H , 3 . 0 ; C1, 17.9. 1-Chloro-4-(nitrophenyl)anthraquinone .-l-Chloro-4-diazonReceived April 15, 1963 ium fluoroborate anthraquinone (20 g.) was thermally decomposed in nitrobenzene (I50 ml.). Kitrogen analysis of the reaction product showed 447; of arylation had occurred. Two The method developed by Coons and colleagues3 for recrystallizations from chlorobenzene gave the yellow I-chloro-4labeling antibody proteins with fluorescein isocyanate (nitrophenyl)anthraquinone, n1.p. 273-274", 6.1 g., 30C%. has become a frequently used tool in immunology. Anal. Calcd. for C;oH,&ISO,: C, 66.1; H , 2.8; C1, 9 . 8 ; , N , 3.85. Found: C , 68.0; H , 3.0; CI, 9.9; S , 3.8. Since Coons' original paper was published in 1942 l-Chloro-4-(~-bromonaphthyl)anthraquinone.-~iniilarly, de( 1 ) Regia Cheinical C o . . Chicago, Ill. composition of l-c.hloro-4-diazonium fluoroborate anthraquinone (2) Operated by the University of Chicago for the U. S. Atomic Energy in a-bromonaphthalene gave 43% of arylation. Recrystalliza-
0
Ar
QJ$
%$_______
( 8 ) XI. S.Whelen. U S.Patent 2,100,527: Chem. A b n l r . , 32,Y(i0 (1938). (9) \V. H. Beisler and 1,. W.Jones, J . A m . CkPm. Soc., 44, 2304 ( 1 9 2 2 ) .
Commission. (3) A . H. Coons, H. J. Creech, R. N. Jones, and E. Berliner, J . Immunol.. 46, 150 (1942).
NOTES
FEBRUARY, 1964 several comprehensive review^^-^ and over 300 papers have appeared. It seemed desirable to extend the usefulness of the method by making a similar reagent which was adaptable to quantitative studies. The present work had been undertaken in order to prepare a fluorescein derivative that would couple with an antibody protein in the normal manner and be radioactive as well as fluorescent. Xitro- and aminofluoresceins I and I1 were synthesized according to the standard KO attempt was made to determine the structure of these isomers until BoreklOa carried out an infrared spectral study of the nitrofluoresceins and structurally related compounds. Borek concluded that the structure of nitrofluorescein I was 5-nitrofluorescein and that the I1 isomer was 4-nitrofluorescein. The ultraviolet spectra in this report (Table I) do not clearly differentiate between these designations. However, the n ~ m e n c l a t u r eused ~ ~ here is preferred over that used previously. lo TABLE I ULTRAVIOLET ABSORPTION SPECTRA" Compound
Fluorescein"." 5-Nitrofluorescein ( I ) 4-Nitrofluorescein (11) 5-Aminofluorescein 4-Aminofluorescein 5-Nitrofluorescein diacetate 4-Kitrofluorescein diacetate 4 ', '5-J~iiodo-4-aminofluorescein
A, m p
E,,,
223 222 225 222 222 285 219 218 205 229 290
60,300 68,209 58,300 60,000 68,200 20,600 70,000 65,500 42,800 47,000 22,300
4 ',5 '-Diiodo-4-nitrofluorescein diacetate 4',5'-Diiodofl~orescein~
204 67,500 204 52,000 230 40,300 a Taken in ethyl alcohol from 200-400 mp on a Beckman Model DK-1 recording spectrophotometer. * Eastmen Kodak Previously reported by C. Hanna and W. T. Smith, grade. Proc. Iowa dcad. Sci., 58, 251 (1951).
The unlabeled title compound was obtained quite readily by treating a solution of 4-aminofluorescein in 1 N hydrochloric acid with two equivalents of iodine monochloride. The immediate orange precipitate of 4',5'-diiodo-4-aminofluorescein gave a reasonably acceptable elementary analysis and was used in subsequent iodide-131 exchange reactions. The radioactive 4',5'-diiodo-4-aminofluorescein could then be converted to the corresponding isocyanate as needed and coupled with an antibody protein without isolation of the isocyanate. A less satisfactory method was originally used in (4) A . H. Coons, "International Review of Cytology," Vol. 5 , G. H. Bourne, E d . , hcademic Press, Inc., New York, N . Y . 1955, pp. 1-23. ( 5 ) E. H. Beutner, Bacterid. Rev., 16, 49 (1961). (6) .4. H. Coons, J . Immunol., 87, 499 (1961). H. Kaplan, J . E z p t l . M e d . , 61, 1 (1950). (7) A . H. Coons and IM. (8) R. T. Bogert and R. G. Wright, J . A m . Chem. Soc., 17, 1310 (1905). (9) It was found here t h a t 4-nitrofluorescein diacetate exists in two crystalline forms. When Crystallized from benzene the product had m.p. 1941 9 6 0 and from ethanol, m.p. 215-217O. Either form was readily converted to the other. T h e infrared spectra of these t w o forms were only slightly (in potassium bromide) different in the fingerprint region. (10) (a) F. Borek, J . OTQ. Chem., 16, 1292 (1961); (b) R . M. McKinney, J. T . Spillane. and G . W . Pearle, i b i d . . 17, 3386 (1962). (11) Nomenclature from "The Ring Index." RRI5935, Reinhold Publishing Co.. New York, N. Y., 1961.
491
?
NO,
NO2
4',5 '-Diiodo-4-nitrofluorescein
PNitrofl uorescein
p
I
O
H
" 2
" 2
PAminofluorescein
4',5 '-Diiodo-4-aminofluorescein
the preparation of this compound. In this series of reactions 4-nitrofluorescein was iodinated with iodine monochloride in acetic a ~ i d . ' ~ . The ' ~ crude 4',5'diioao-4-nitrofluorescein was purified through the easily crystallized diacetate. Deacetylation followed by catalytic hydrogenation resulted in a 4',5'-diiodo4-aminofluorescein which was low in iodine. Under identical reduction conditions the iodines of 4',5'diiodofluorescein14were not affected. The preference toward 4',5' substitution in fluorescein has been demonstrated effectively by previous w o r k e r ~ . ~ ~ In' - l ~order to show that iodination with iodine monochloride in acidic media occurs in the 4',5'positions, fluorescein was treated with this reagent. Thin layer chromatography showed the main product to be identical with 4',5'-diiodofluorescein obtained from a commercial source,14 but several minor spots indicated impurities.13 The exchange reactions of sodium iodide-131 with 4',5'-diiodo-4-aminofluorescein were accomplished in high radioactive yield by mixing the materials in acetic acid overnight (Table 11). Warming was necessary to effect solution. The mechanism of the exchange reaction with iodide131 is uncertain. It has been reportedlg that iodine atoms and not iodide ions exchange into thyroxine and that only the iodides adjacent to the phenolic function exchange. It is possible that under our exchange conditions iodide ion was oxidized to iodine to catalyze the exchange with organic iodide adjacent to the (12) G . Boyack, G. E. Moore, and D. F. Clausen [Nucleonics, 3, 62 (1948) I treated fluorescein with iodine monochloride generated in situ from
dichloramine T and potassium iodide-131 in acetic acid. No structural evidence was given for the positions of the iodines. (13) A. Roe, R. L. Hayes, and H. D. Rruner IJ. Am. Chem. Soe.. 7 3 , 4483 (1951)l repeated the work of Boyak, e t al..'* b u t found t h a t some chlorination as well as tri- and tetrahalogenation had occurred. (14) The author would like to thank C. Kuimjiam of Distillation Products Industries for his comments on the iodination reactions and for a generous sample of 4'.5'-diiodofluorescein. (15) R. B. Sandin and R. L. Orvis, J . OTQ.C h e m . , 13, 1235 11958). (16) UT.R. Orndorf and A . J. Hemmer, J . A m . Chem. Soc., 49, 1272 (1927). (17) M . A . Phillips, J. Chem. Soc., 724 (1932). (18) R. B. Sandin. A . Gillies, and 5. C. Lynn, J . A m . Chem. S o c . , 61, 2919 (1939). (19) G. I. Gleason, J . Biol. Chem., 113, 837 (1955).
492
TABLE I1 IODIDE-131 EXCHANGE REACTIONS Reaction 1 2 4 ‘,5’-Diiodo-4-aminofluorescein 60 mg. 61 nig. Acetic acid solvent 2 . 0 ml. 2 . 0 ml. Sodium iodide-131 aqueous” 0 40ml. 0 40ml. Initial activityb 4 36 mc. 6 02 mc. Reaction conditions 8 . 5 hr. a t 95-102” 19 hr. a t 78-94’ Decay timec 46 hr. 29 hr. Activity of productd 37 pc./mg. 88 pc./mg. Per cent exchange 60 7 0 99% a Carrier free. Activity a t start of exchange reaction. Time elapsed from start of exchange to time of counting. Measured activity a t time of counting.
phenolic function. However, it also is known20-22 that halogenated quinones readily undergo nucleophilic displacement of their halogens. The organic iodide in our fluorescein has the character of an atom ortho to a quinoid as well as phenolic function, and the possibility of nucleophilic displacement by iodide-131 ion should not be dismissed. Experimenta123
4’,5’-Diiodo-4-nitrofluorescein.-A mixture of 1.00 g. of 4nitrofluorescein, 894 mg. of iodine monochloride, and 25 nil. of glacial acetic acid was stirred for 5 hr. The original suspension of 4-nitrofluorescein gradually dissolved and a new orange solid precipitated. After standing overnight, 200 ml. of water was added and the precipitate became red. It was filtered, redissolved in cold 1 K sodium hydroxide, reprecipitated with acetic acid, collected, washed with water, and dried. The crude red 4’,5’-diiodo,4-nitrofluorescein,1.56 g., had n1.p. 228-240’ dec. This crude product was purified somewhat through its diacetate. Acetylation.-The crude 4’,5‘-diiodo-4-nitrofluorescein was refluxed with sodium acetate in acetic anhydride for an hour. Working up the product with recrystallQation from acetic anhydride gave 4’,5’-diiodo-4-nitrofluoresceindiacetate as fine crystals with a slight yellow tinge, m.p. 296-298”. Anal. Calcd. for CzaH13N0&: C, 40.41; H , 1.84; I , 35.59. Found: C, 40.43; H , 1.96; I , 35.55. Deacetylation.--A suspension of the above diacetate in a saturated solution of sodium hydroxide in 907, ethanol was stirred and warmed gently until all dissolved. Dilution with water and acidification with glacial acetic acid gave a purified 4’,5‘diiodo-4-nitrofluorescein monohydrate, m.p. 262-264’ dec. On drying under vacuum a t 110” this solid lost weight equivalent to an equimolar portion of water. On exposure to air the weight again gradually increased. Anal. Calcd. for CzoHeNOJ2.H20: C, 37.11; H , 1.71; I , 39.24. Found: C, 37.0; H , 1.38; I, 39.55. 4’,5’-Diiodo-4-aminofluorescein.A. Iodination of 4-Aminofluorescein in Hydrochloric Acid.-A warm solution of 470 mg. of iodine monochloride in 10 ml. of 1 -4- hydrochloric acid was added to a warm stirred solution of 501 mg. of 4-aminofluorescein in 60 ml. of 0.7 S hydrochloric acid. .4solid precipitated immediately. The mixture was heated nearly to boiling and cooled, and the solid was collected, washed well with water, and dried. The bright orange 4’,5’-diiodo-4-aminofluorescein,860 mg. (99%), m.p. 213-218” dec., gave a reasonable elemental analysis and was used in the exchange reactions. Anal. Calcd. for C2OHllX06I2: C, 40.09; H , 1.85; I , 42.36. Found: C,40.78; H, 2.03; I,42.60. B. Hydrogenation of 4’,5’-Diiodo-4-nitrofluorescein.-The hydrogenation of 286 nig. of purified 4’,5’-diiodo-4-nitrofluores( 2 0 ) Ng. Ph. I3uu-Hoi, R. Royer. a n d M. Hubert-Habert, Rec. Iraa. chim.8 75, 188 (1954). N. Makarova, Zh. Obshch. K h i m . , SO, 1380, (21) A. Ya. Rerlin and .i. 1582 (1960); Chem. Abstr., 65, 4 9 9 , 1500 (1961). (22) d. \V. Hancock, C. E. hIorrel1, and D. R h u m , Tetrahedron Lefters, 987 (1962). ( 2 3 ) Jlelting points were taken on a calibrated Fiacher-Johns hot-stage
block.
VOL.29
NOTES
cein was carried out in 10 ml. of ethanol at room temperature for 1 hr. using hydrogen a t 53 p.s.i. and about 2 g. of Raney nickel catalyst. The catalyst was filtered, and the ethanol solution was reduced in volume by gentle heating in a stream of nitrogen. The addition of 100 ml. of water produced a n unfilterable suspension which was extracted into ether. Evaporation gave 156 mg. of a crude bright orange 4’,5’-diiodo-4-aminofluorescein, m.p. 205-210” dec., which was low in iodine. Under identical conditions the iodines of 4’,5’-diiodofluorescein were not affected. 4’,5’-Diiodofluorescein.-A mixture of 3.32 g. of fluorescein and 3.24 g. of iodine monochloride in 25 ml. of glacial acetic acid waa warmed to 115” for 1 min., cooled, and stirred for 8 hr. a t room temperature. The solid was collected and washed successively with 10 ml. of acetic acid and 100 ml. of water. It was then suspended in acetone, collected, and dried giving 3.54 g. of crude orange 4’,5’-diiodofluorescein as a powder, m.p. 22G235’. An Eastman Kodak grade material had m.p. 247-250’ dec. Thin layer chromatography on Merck silica gel G using methanol showed these compounds to be identical except for several minor spots in our product, which may be the chloride or other polyhalogenated material.18 No purification of this product was attempted. Exchange Reactions with Iodide- 131.-The 4‘,5’-diiodo-4aminofluorescein was suspended in acetic acid in a centrifuge tube. Aqueous sodium iodide-131 was added and the mixture was stirred and heated to about 100” using a silicone bath. At this temperature nearly all of the solid dissolved. After the reaction time indicated in Table 11, the mixture was cooled, diluted with water, and centrifuged. The solid was washed three times with water, then dried, and stored in a vacuum desiccator. There was no change in the melting point.
Acknowledgment.-The problem was suggested by Dr. Robert J. Hasterlik to whom the author is grateful for his encouragement. The author also would like to thank 3Ir. Robert Hart for doing the spectral studies, Mr. William Saschek of the University of Chicago Microanalytical Laboratory for the elemental analyses, and RIr. Leon Gortler for synthesizing the 4- and 5nitrofluorescein diacetates. The generous and helpful comments, as well as the general assistance, of Dr. Weldon Brown of the Argonne National Laboratories also are appreciated.
The Reaction of Alkyl Borinates with a-Amino Acids IVANH . SKOOG
Contribution Xo. 869 from the Central Research Laboratories, Minnesota Mining and Manufacturing Company, St. Paul 19, Minnesota Received July $9,1963
The stabilization of boron compounds through complex formation with amines is well-established, By this method the air-sensitive alkyl- or aryl-substituted boron derivatives can be transformed into stable solids. For instance, a diarylborinic acid readily condenses with ethanolamine or 8-hydroxyquinoline to form an internally stabilized compound. Formation of mixed anhydrides of borinic and amino acids also should result in stabilization. A recent British patent2 disclosed that heating trialkylborines with amino acids produces this type of compound, but no examples were given for the triarylborines. (1) (a) J. Douglass. J . Org. Chem.. ‘26, 1312 (1961); (b) R. Letsinger a n d I. Skoog, J . A m . Chem. Soc., 77,2491 (1955). (2) K. Lang. F Schubert. and K Xutzel, British Patent 905,093 (Septem ber5, 1962).
