[19571 Studies 0% Biological Methylation. Part XVI. 8. Studies on Biolugiml Methylation. Part XVI.* Natwal s'ulph-onium Compds. The Alkyl Methyl Sulphides evokved from the Urine of Dogs by Boiling AIhli. By DEREK LEAVERand FREDERICK CHALLENGER. The urine of dogs when boiled with alkali evolves a sulphide which previous workers have incorrectly assumed to be diethyl sulphide arising by fission of diethylmethylsulphonhm hydroxide Chromatographic examin- ation of the sdphidimine RR1S-+N*SO;C6H,Me-# and the methylsulphon-ium hydroxides SRRWe)OH prepared from the natural sulphide has shown it to be a mixture of methyl n-propyl sulphide and probably n-butyl methyl sulphide the former preponderating. The methyl k-propyl sulphide was also identified by the m.p. and mixed m. p. of its stdphidimine and mercuri- chloride. ABEL stated that when the urine of dogs was warmed with alkali diethyl sulphide was evolved but no proof of identity was given. The m. p. of the mercurichloride of the sulphide was given as 146" and also as 14%". Diethyl sulphide mercurichloride has m. p. 119O. Nevertheless later workers%s74 accepted Abel's statement and Neuberg and Grosser claimed to have isolated the precursor of the sulphide as the phosphotungstate and iodobismuthate and to have identified it as diethylmethylsulphonium hydroxide. No experimental details were supplied and no further communication appeared. Diethylmethylsulphonium iodide with sodium hydroxide at 100' gives mainly ethyl methyl sulphide characterbed as mercurichloride.The work of Ingold and Kuriyan ti also suggests that this sulphide is the main product of the alkaline decomposition of this sulphonium ion. A re-investigation of the sulphide from dog's urine was therefore initiated. Dr. Margaret Whitaker passed the evolved vapours through (a) mercuric cyanide and (b) mercuric chloride. Thiols were absent. The m. p.s of the mercurichloride mercuri- bromide and benzylsulphonium picrate approximated to those of corresponding deriv- atives of methyl n-propyl sulphide. Methyl n-propyl sulphide was not however the only component because even after crystallisation the m. p.s of these derivatives were usually slightly lower than those of the authentic compounds. The identity of the sulphide has been established by paper chromatography the sulphidimine and the methylsulphonium hydroxide being examined.The sulphidimines were detected by spraying with acidified potassium iodide and heating at 80' to give brown spots of iodine. The methylsulphonium hydroxides were detected by bromophenol-blue. Chromatography of the sulphidimine of the natural sulphide produced only one spot * Part XV,J. 1956 1163. Abel 2.physiol. Chem. 1894 a,263. ibid. 1931 198 186; 1933 817 177; 1934 m,211; Xlin. WOCh. 1932,11,177. li ~htisto~os, * Wohlgemuth 2.physwl. Chem.,1933 %El 207. Neuberg and Grosser CmtraZbW Physiol. 190b-.6,19 316 Ingold and Kuriyan J. 1933 991. Leaver and Challerizger with the Same RF value as niethyl n-propyl sulphidimine. On repetition of the chrom- atogram cutting out the appropriate area reduction with tin and hydrochloric acid * to the sulphides and reconversion into the mercurichloride gave a material whose m.p. was still lower than that of methyl n-propyl sulphide mercurichloride. The impurity was therefore another sulphide of similar RFvalue. Five sulphidimines possessed approximately this RFvalue namely (1) diethyl (2) methyl isopropyl (3) n-butyl methyl (4) isobutyl methyl and (5)set.-butyl methyl sulphidimine. When specimens of the methylsulphonium hydroxide from the “natural ” sulphide were chromatographed two spots were produced. The stronger had the RFvalue of dimethyl-.n-propylsulphoniumhydroxide ; that of the weaker was slightly higher. Both diethylmethyl- and dimethylisopropyl-sulphoniumhydroxide had lower RF values than the dimethyl-n-propyl compound [eliminating possibilities (1) and (2) for the second sulphide].The three butyldimethylsulphonium hydroxides (n-,iso- and sec.-) possessed higher RFvalues than the dimethyl-n-propyl compound but a mixture of the sec.-butyl compound and excess of the n-propyl compound was not separable on paper [eliminating possibility (6)]. The pattern of the natural mixture could be reproduced almost exactly by a mixture of dimethyl-n-propyl- and .n-butyldimethyl-sulphoniumhydroxides (by adjusting the relative proportions of the components) but not by a mixture containing the s-propyl and isobutyl compounds. This suggested that the second component of the natural sulphide was probably n-butyl methyl sulphide.A sample of the “ aatural sulphidimine ” was recrystallised finally giving almost pure methyl %-propyl sulphidimine. The material from the mother-liquors was chrom-atographed alongside a synthetic mixture of methyl +propyl and It-butyl methyl sulphidimine. The two spots from the natural mixture agreed in RFvalue with those from the synthetic mixture. The remainder of the natural sulphidimine was also chrom-atographed the appropriate zones were cut out and the sulphidimines so separated reduced to sulphides and converted into the mercurichlorides. That from the band of higher RF value had m. p. and mixed m. p. identical with that of methyl n-propyl sulphide mercurichloride. That from the other band had a rather unsharp m. p. which was how- ever identical with that of n-butyl methyl sulphide mercurichloride.These results indicate that the sulphide evolved from dog’s urine is a mixture of methyl n-propyl and n-butyl methyl sulphide the former predominating. A study of the parent compound (or more probably compounds) in the urine was undertaken and a method of isolation devised. This compound is removed from solution by sulphonic acid cation-exchange resins and can be displaced from these by treatment with stronger bases. Its behaviour during isolation is similar to that of a sulphonium compound bearing an acidic group. The investigation is being continued. The isolation of dimethyl-@-propiothetin chloride from the red marine alga Polysiphorcia fastigiata by one of us and (Miss) Simpson 7 was the first authenticated example of the natural occurrence of a sulphonium compound unless we regard sulphoraphen Me*SO*CH:CH*CH,CH,*NCS isolated from radish by Schmid and Karrer,* as a potential sulphonium compound.This propiothetin was also found in other marine algze (Entwmwpha intestidis and Spimgmpha arcta) by Bywood and Challenger.Q*fQ Maw and du Vigneaud l1 showed that it supports the growth of rats on a methionine-choline-free diet l2containing homo- cystine and Dubnoff and Borsook lafound that methionine is formed from the thetin and Ash Challenger and Greenwood J. 1961 1881 Challenger and Simpson J. 1948 1691. Schmid and Karrer HeZu. Chim. Acfu 1948 81 1017 1087 1497. * Bywood and Challenger Biochem. J. 1963 68 xxvi. lo Bywood Thesis Leeds 1963.Maw and du Vigneaud J. Biol. Chem. 1948,174 381. la du Vigneaud Moyer and Chandler ibid. p. 477. l* Dubnoff and Borsook ibid. 1948 176 789. [1957] Studies on Biological Methylation. Part X VI. homocystine in enzyme preparations from rats. Cantoni then isolated "active methionine " 14 (the ionic S-adenosinylmethionine) from preparations of kidney and liver containing methionine and McRorie et aZ.15 showed that cabbage lettuce and other vegetables contain the methylmethioninesulphonium ion. The same ion qcurs in aSpXagUS.l* The occurrence of dimethyfsulphonium compounds in numerous marine and some fresh water alp in bracken and in several species of Eq2cisetum was established in the Leeds lab0ratories.1~ Ericson and Carlson 18 showed by paper chromatography that dimethyl-p-propiothetin and p-alanine were present in all marine algs examined.In P.fastigiata and Ulva Zactuca where the thetin was present in large amount the quantity of fkdanine though relatively much lower was at a maximum. Challenger l9 suggested that there may be a biogenetic relation between the thetin and p-alanine or (possibly) its betaine similar to that established by Woolley20 between quaternary ammonium compounds of the type of thiamine and various mines. EXPERIMENTAL Preparation of tb Mercurichloride of the Sulphide from Dog's Urine.-The urine was made strongly alkaline with sodium hydroxide and boiled; the volatile products were aspirated through (a) dilute sulphuric acid to absorb volatile bases (b) 4% mercuric cyanide and (c) 3% mercuric chloride.No precipitate was obtained in the mercuric cyanide; indicating the absence of an alkmethiol. The sulphide was precipitated as its mercurichloride m. p. (crude) -166' (sintering from 146') but melting was not complete below about 169'. This behaviour was charwteristic varying only slightly with different samples of urine. The yield of mercuri-chloride per 1. was about 0.02-0.04 g. Recrystallisation from benzene containing mercuric chloride raised the m. p. to 168-160" and the mixed m. p. with authentic methyl #-propyl sulphide mercurichloride (m. p. 166") was 169-182' pound Hg 65.1. Cdc. for (CH8-SC,H,),,6HgC1 Hg 664%)]. Conversion of the Crude MercuvichZoride into Other Derivatives.-The sulphide was regenerated from its mercurichloride by warm dilute sodium hydroxide and separately aspirated into (a) saturated mercuric bromide and (b) 10% v/v alcoholic benzyl bromide.In (a) a mercuri-bromide was obtained which on recrystallisation from benzene had m. p. 113' and mixed m. p. 11&116" with authentic methyl 12-propyl sulphide mercuribromide (m. p. 116"). The alcoholic solution from (b) was left for 2 days diluted with ether and extracted with water from which a precipitak was obtained with sodium picrate. This when recrystallisd from alcohol had m. p. 93-94' and mixed m. p. 94-96' with benzylmethyl-rr-propylsulphonium picrate rn. p. 96-96O (Challenger and Rawlings s1 give m. p. 9b-95.6'). For other data see Table 1. Chromatographic Methods.-The chromatograms were developed at about 20" in glass tanksby the descending front method.(1) Chromatography of sulphidimines method (a). The sulphirfimines were applied to the paper (Whatman No 1) in acetone. The best solvent systems were cycZuhexano1-water and butan-l-ol-isopropyl ether-water (6 2 :4 by vol.). The papers were irrigated with the top (organic) layer in both cases. With butanol4sopropyl ether development was complete in 16 hr. but the chromatograms developed with cyclohexanol required 3-6 days. The RFvalues for a number of sulphidhhes are given in Table 2. Although producing less effective separations of the lower sulphinimines cycZohexano1 gave more compact spotsand greater separations of the higher sulphidimines than butanol-isopropyl l4 Cantoni J.Amer. Chcm. Soc. 1962 74 2942; Confdrences et Rapports 3 CongrZfs Intern. Bio-chimie 1966 p. 233; Challenger Quad.Rev. 1966 9 274 279. lS McRorie Sutherland Lewis Burton Glazener and Shive J. Amy. Chcm. SOC.,1964 76 116. la Challenger and (Miss) Hayward Chem. and Ind. 1964 729. l7 Challenger Leaver and (Mrs.) Whitaker Biockem. J. 1963 66,ii; Leaver Thesis Leeds 1963. l8 Erimn and Carlson Arkiu &mi 1964,6 611. l* Challenger Confbnces et Rapports 3 CongrQ Intern. Biochimie 1966 p. 239. ao Woolley Nature 1963,171 323. *l Challenger and Rawlings,J. 1937 868. Leaver and Challenger ether and it was the preferred solvent. The chromatogramswere sprayed with 1% potassium iodide in 0.2N-hydroehloric acid and heated at 80°. Brown spots were produced by the sulphidimines presumably owing to hydrolysis to a sulphoxide which oxidises iodide to iodine After a few hours the whole paper became brown.. TABLE1. Cmpatism of the m. p.s of cmaspding hrivativw of ths mde sulphide fromdog's wine of methyl n-pql?yl sulphide and of didhyl sulphide. Natural Mixed Derivative sulphide MeProS m. p. EtrS Mercurichloride ....................................... 146-169O 1W0 150-162° * 119O Mdbromide ....................................... 113 116 116-116 - Sulphidimine (from chloramine-'lr) ............... 91-93 104-106 -I46 Dialkylbenzylsulphonium picrate ............... 93-94 96-90 94-9s 113 Diakylbenzylsulphonium styphnate ............ 8-6 77 80-86 129 The derivatives of the natural sulphide had undergone only the minimum of recrystallisation.* Ther natural sulphide mercurichloride m. p. 1S8-160° was used for this mixed m. p. detem-TABLE2. RF value Rp value r Sulphidimine In BuQH-W,O In cyclohexanol5 Sulphidimine rIn Bu'LOH-pr(,O A In cyclohexanol\ Mes ............ 0.73 0.76 MeBun ...... -0*83-0*84 MeEt ......... 0.83 0-81 MeBd ...... .__ 0*83-0*84 Et ............ 0.90 0.83 MeBu' ...... -0*83-0*84 Mew ......... 0.90 0.83 pp,............ 0-96 0.87 Mew ......... -0.83 Bun ......... 0.96 0.91 I' Method (b). In reversed phase " chromatography of sulphidimines Whatman No. 1paper was rendered water-repellent by treatment with a 0.76% w/v solution of Perspex (methyl methacrylate polymer) in chloroform. After application of the sulphidimines in acetone solution the papers were suspended in the tank for 12 hr.without solvent in the trough so allowing them to come to equilibrium with both phases of the solvent system which were placed in separate vessels at the bottom of the tank. The trough was then filled with the aqueous phase which was allowed to move down the paper for a further 12 hr. The most satisfactory solvent system was water-ethyl acetate. The positions of the sulphidimines were revealed as in method (a). The order of RFvalues was the reverse of that obtained by (a) thus afiording a better separation of the higher sulphidimines. (2) Chromatography of bases. Whatman No. 1 paper was washed by irrigation with 2N-hydrochloric acid (1 ml. per sq. in.) followed by distilled water until free from acid.(The use of unwashed paper resulted in poor and irregular background colours.) The solvent system used was n-butanol-acetic acid-water (6:1 :4 by vol.) and development was complete in about 17 hr. The papers were dried and sprayed with a 0.1% solution of bromophenol-blue in alcohol containing acetic acid (0.2-0.6%). The bases gave blue spots on a yellow ground. Preparation of the 'I Natural Sutphidimine."-The natural sulphide " mercurichloride (0.3 g.) was decomposed by sodium hydroxide solution and the sulphide aspirated through saturated aqueous chloramine-T. The sulphidimine was extracted with several small volumes of chloroform and the extracts were dried (CaCI,) and evaporated leaving the sulphidhine as a viscous residue. Paper Chromatografihy of the Natuvat Sulphidimine " by Method (a).-A trace of the sulphidimine (before recrystallisation) was chromatographed in acetone by method (a) with cyclohexanol as solvent alongside a mixture of authentic dimethyl and methyl n-propyl sulphidimine which are readily separated.The" natural sulphidimine " produced only one spot of which the RFvalue was the same as that of the methyl propyl sulphidimine. NO dimethyl sulphidimine was present in the natural compound. The rest of the sulphidimine was crystdised three times (m. p. 93-94O) dissolved in acetone (0.6 ml.) and applied as a band to the base line.. After development with cyclohexanol and drying narrow strips were cut from the edges and treated with a&.€ified potassium iodide to reveal the position ofthe sulphidimine.The area containing the sulphidimine on the unsprayed portion of the paper was then cut out and the sulphidimine removed with acetone It was [19571 Studies on Biological Methylation. Part XVI. reconverted into the mercurichloride by tin in boiling 2~-hydr&onc acid the sulphide so produced being aspirated into3% mercuric chloride solution.6 The resulting mercurichloride X (m. p. 16Q0,sintering at 160') was recrystallised twice from alcohol-benzene containing a little mercuric chloride but its m. p. could not be raised above 169-160" with sintering from 160". The components of the natural sulphide had not been separated by chromatography of the sulphidimine. Preparation of the Methylsulphonium Hydroxide from the Natural Sulphide.-The mercuri-chloride of the natural sulphide was warmed with sodium hydroxide solution and the sulphide aspirated through two tubes containing alcohol.To the resulting solution an excess of methyl iodide was added and after 2 days the solution was evaporated. The syrupy residue was dissolved in water treated with silver oxide and evaporated to 1 ml. in a vacuumdesiccator giving an aqueous solution of the methylsulphonium hydroxide. The preparations A and B were carried out with two specimens of mercurichloride; (A) was obtained from the mercuri- chloride sample X,which after mcrystaWtion had been recombined with its mother-liquors ; (B)was obtained froman untreated sample (0.1 g.) of the natural sulphide mercurichloride. Chromatographyof Spthetic Sulphonium Hydroxides.-The Rp values of a number of trim1 sulphonium hydroxides were SMe,)OH 0.30 ; SMe,Et)OH 0.40; SMe,Pm)OH 0.44; SMe,Pr?oH 0.37; SMe,Bun)OH 0.61; SMe,Bui)OH 0.48 ; SMe,Bua)OH 0.46 ; SMePrn,)OH 0.66.Each of the butyldimethylsulphonium hydroxides was mixed with about five parts of dimethyl-n-propylsulphoniumhydroxide and the three mixtures were chromatographed on the same paper. That containing the sec.-butyl compound was not separated. In accordance with the lip values the mixture (N) containing the n-butyl compound separated more clearly than that (I)containing the isobutyl compound. Chromatography of the Methylsulphonium Hydroxide of the NaturaZ Sulphide.-hparation A (see p. 43) when chromatographed separated into two distinct zones,a strong one and a weaker one of higher RFvalue.The same result was obtained with B and the stronger spot in each case had the Rp value of dimethyl-n-propylsulphoniumhydroxide run as a control. A second chromatogram was prepared in which B was run alongside mixtures N and I (see above). The relative proportions of the two components in the synthetic mixtures were not quite the same as those in the natural mixture (estimated from the intensities of the coloured spots) and consequently neither N nor I gave patterns exactly the same as preparation B although N resembled it more closely than I. It was possible by altering slightly the proportions of the ingredients to make the pattern given by N almost identical with that of the natural mixture but this was not possible with I which always remained chromato- graphically distinct.PreParat&m and Chromatopaphy of Methyl n-Prq5yl Sulphidimine from the Natural Su@hidimine.-The mercurichloride of the natural sulphide (0.42 g.) was converted into the sulphidimine and purified by precipitation three times from acetone with a mixture of light petroleum (b. p. 40-60") and ether and then by recrystallisation threetimes from alcohol-light petroleum (b. p. S0-8Oo). The m. p. became constant at 103-104° and was 104.6" when mixed with authentic methyl n-propyl sulphidimhe (m. p. 104-106'). The mother-liquors from all six crystallisations were combined and evaporated (see p. 40) and the residue crystallised from acetone-ether to remove any toluene+-sulphonamide.The sulphidimine (R) s<! obtained was dissolved in acetone (0.6 ml.). A sample was chromatographed by method (b) alongside a mixture of methyl n-propyl (2 parts) and n-butyl methyl sulphidimine (1 part). The twospots produced by the natural sulphidimine had the same RBvalues as those produced by the synthetic mixture. The remaining natural sulphidimine (R) was applied in acetone as a band to the base line of a large chromatogram on Perspex-coated paper and developed by method (b). Narrow strips were cut from the edges and after treatment with potassium iodide were used as guides to the positions of the separated sulphidimines on the main unsprayed chromatogram. The bands were cut out and boiled separately with aqueous sodium hydrogen sulphite and volatile products aspirated through water to absorb sulphur dioxide and 3% mercuric chloride solution.The mercurichloride from the band of higher RF value melted at 164-166O alone and at 166" when mixed with authentic methyl n-propyl sulphide mercurichloride (m. p. 166"). That from the band of lower Rp value melted at 110-114° sintering at 90°. This behaviour was identical with that of n-butyl methyl sulphide mercurichloxide prepared from pure n-butyl Leave? and ChlleBger methyl sulphidimine by boiling with sodium hydrogen sulphite and aspiration of the sulphide into mercuric chloride. Alkaline Decomposition of Sulphonium Iodides.-( 1) Diethylmethylsulphoniurniodide. The iodide (1 g.) was boiled with 2~-sodium hydroxide and the vapours were aspirated through aqueous 3% mercuric chloride for 11 hr.Precipitation of mercurichloride (m. p. 117-119" sintering from 100") then ceased. This complex was recrystallised four times from ethanol as glistening white plates having finally m. p. 126-127" (sintering at 122') and mixed m. p. 120-127" with the mercurichloride of ethyl methyl sulphide (m. p. 127"). The m. p. of the uncrystallised product was near to that of diethyl sulphide mercurichlonde (119") but the substance was clearly a mixture. Challenger and Simpson 7 state (without further details) that the m. p. of the mercurichloride of the sulphide evolved from diethylmethylsulphonium iodide and 2~-sodium hydroxide is 127". -Ingold and Kuriyan find that formation of alcohol represents 46% of the reactions in the thermal decomposition of the hydroxide from which it appears that the sulphide mixture contains at least 66% of ethyl methyl sulphide.(2) Ethyldimethylsulphonium iodide. The method was as in (l),and iodide (1 g.) and 2~-sodium hydroxide were used (time 10 hr.). The mercurichloride sintered slightly below 126" and melted at 127". Recrystallisation from ethanol gave m. p. 127-129" (sintering 126") unchanged on crystallisation. Ethyl methyl sulphide mercurichloride melts at 127". Ingold and Kuriyan rj find that the sulphonium hydroxide gives 73% of alcohol on thermal fission and from the above results it appears that methanol predominates. (3) Dimethyl-n~,rqpylsulphoniumiodide. Reaction was as in (1) but with (a) 2N-and (b) 6N-sodium hydroxide.In expt. (a) the unrecrystallised mercurichloride had m. p. 164-166" with slight setering from l67-16Oo; in (b) the m. p. was 166" fairly sharp with slight sintering from 169". Methyl n-propyl sulphide mercurichloride melts at 166" and dimethyl sulphide mercurichloride at 167-168'. Ingold and Kuriyan 6 found 92% of alcohols onthermal fission of this sulphonium ion. Some of the unrecrystallised mercurichloride obtained in decomposition (a) was decomposed by sodium hydroxide and the sulphides were converted into the sulphidimine. This was a mixture since on paper chromatography [method (a)]with the sulphidimines of dimethyl and methyl n-propyl sulphides as controls the presence of both these sulphideswas detected. (4) Methyldi-n~rqpylsulphoniumiodide. Decomposition of this iodide (1 g.) with 2N-sodium hydroxide for 11 hr.gave mercurichloride m. p. 110-140" unrecrysm. Crystallis-ation from ethanol gave a product having finally m. p. 163-164' (sintering at 160"). Methyl n-propyl sulphide mercurichloride melts at 166". Ingold and Kuriyan 6 found 82% alcohol formation from the hydrochloride. Preparation of Reference and Intermediate Compounds.-Alkyl methyl sulphides. The methyl n- and iso-propyl sulphide and n-butyl methyl sulphide were prepared from the alkanethiols by addition to sodium ethoxide (1 equiv,) in ethanol. Dimethyl sulphate (0.6 mol.) was added and the mixture boiled under reflux for 30 min. and poured into water. The separated sulphide was fractionated with a 12 in. column. Ingold Jessop Kuriyan and Mandour used methyl iodide.iso-and sec.-Butyl methyl sulphide were prepared from methanethiol sodium ethoxide (1 equiv.) in ethanol and the alkyl bromide (1 equiv.) by the same procedure. Vogel and Cowan prepared isobutyl methyl sulphide from isobutanethiol and dimethyl sulphate. sec.-Butyl methyl sulphide boils at 112-113" (Found C 67.6; H 11.6; S 30.3. C,H,,S requires C 67.6 ; H 11.6 ; S 30.8%). The yields of these sulphides were usually 66--66%. Sulphidimines. The sulphidimines were prepared from the sulphide and saturated chloramine-T and recrystallised to constant m. p. from alcohol-light petroleum (b. p. 00-80"). The following new compounds were prepared Methyl n-propyl sulphidimine m. p. 104-106° (Found C 60.9 H 6.6; N 6.3. C,,H,,O,NS requires C 60.9; H 6.6 ; N 6.4%) ; methyt isOpvspVZ sulphidimine m.p. 116-117° (Found C 604; H 6.6; N 6.1%) ; is&tyZ methyZ sulphidimine m. p. 122-123" (Found C 63-1; H 7.1; N 6.2. C,,H,,,O,NS requires C 62.7; H 7.0; N 6.1%); see.-butyl methyl sulphidimine m. p. 79-80" (Found C 62.6; H 6.8; N 6.1%). AZkylsulphonium iodides (see Table 3). These were usually prepared by mixing the sulphide with methyl iodide in equivalent proportions and leaving them in a closed vessel. When the Ingold Jessop Kuriyan and Mandour J. 1933 633. Vogel and Cowan J. 1943 21. [1957J Studies on Biological Methylutim. Part X VI. deposited syrup frequently discoloured by iodine did not increase in amount it ww separated, washed with dry ether and recrystallised from dry alcohol-ether.The iodides were deliques- cent and were analysed by titration. They were converted into the sulphonium hydroxides in aqueous solution by aslight excess of silver oxide filtered and used for chromatography. No. R 1 Me-2 _. 4 Me 6 -6 Me 7 -8 Me 9 Et-10 11 -12 Me 13 -14 -16 Me 16 -17 -18 Me 19 Me 20 Me No. / C 1 -6 48.2 7 8 46.0 9 60.1 10 47.9 11 44.6 12 -16 17 18 19 20 Methyl n-propyl sulphide (3 g.) and n-propyl iodide (6.7 g.) were refluxed for 6 hr. a syrup separating. Nitromethane was then added. After some weeks addition of ether gave a brown syrup. Dissolution in alcohol left white insoluble crystals m. p. 216' (decomp.). Trimethykulphonium iodide decomposes at 216O.With dry ether the alcoholic extract gave a syrup which with aqueous potassium mercuri-iodide gave a yellow paste. After five recrystallisations from acetoneether this had m. p. 106'. Methyl di-n-propylsulphonium mercuri-iodide prepared by use of methyl iodide melted at 106". The benzyl- and 4-nitrobenzyl-sulphoniumsalts were prepared from the sulphide and the bromide in alcohol or in the case of the diethyl-4-nitrobenzyl compound without a solvent. Unchanged halide was removed by addition of water and ether and the aqueous solution precipitated with sodium picrate or styphnate. Methyl-n-lbrqpylacetottretin Picrolmate. When methyl rt-propyl sulphide (3 g.) and bromo- acetic acid (4.6 g.) were mixed the temperature rose and a syrup separated ; formation of this ceased after 2 days.It was removed and washed with dry ether; dissolution in ethanol and precipitation with dry ether did not effect crystallisation. It was therefore dissolved in water and passed through a column containing Zeo-Karb 225 resin (40-70 mesh; nominally 4.6% cross-linked; H+form). After being washed with water the column was developed with glycine (0*2M)until the amino-acid was detected in the effluent by ninhydrin. The effluent was evaporated the residue extracted with ethanol and the insoluble glycine removed. Evapor-ation of the ethanol left the syrupy thetin hydroxide. Picric and chloroplatinic acid gave oily Stlcdies (WL Biological Methyktiort. Pad XVI. precipitates. Solid picrolonic acid gave a picroZonatc which on crystallisation from ethanor-.ligroin (b. p. 60-80") had m. p. 14&141° (Found C 46.6; H 4.8. Cl,HI,,0,N4S reqnires C 46.6; H 409%). MethyZ-n+w@y&a-fw~othetin picrolmafe. From methyl n-propyl sulphide (3 g.) and a-bromopropionic acid (6.1 g.)a syrup slowly separated. It did not crystallise so was converted as above into the hydroxide and the picrolonate m. p. 136-138" (Found C 48.0; H 4.9. C,,H,,O,N,S requires C 48.0 ; H 6.0%). DimethyZpyruvothetin bromide. From dimethyl sulphide (2.0 g.) and bromopyruvic acid (6-6g.) in dry ether a white solid quickly separated and after 1-2 hr. was washed with dry ether. In a bath preheated to 120" the bromide had m. p. 143-144" (decomp.) (Found Br 34-7. C,H,O,SBr requires Br 3409%). 3-CurboxyallyldimefhyZs~phoni~m bromide.YBromocrotonic acid (1.66 g. 0.01 mole) in dry ether (10 c.c.) and dimethyl sulphide (0.62 g. 0.01 mole) were mixed. After a week the colourless needles of the sulptronium bromide. were washed with dry ether and recrystallised from ethanol-ether (Found Br 36.1. C,H,,O,SBr requires Br 36.2%). a-HydroxydatnefhyZ y-butyothefin and a-chlwodimethyl y-bufyrothefin picrolonates. Methyl-methioninesulphonium iodideU (5-8 g.) in water (160 c.c.) was shaken with excess of silver chloride and the mixture filtered. Equal volumes (40 c.c.) of N-sodium nitrite and N-hydrochloric acid were slowly added simultaneously with shaking. The mixture was then heated to 60' and left at room temperature for 2 hr. neutralised with silver oxide filtered and passed through a column of Zeo-Karb225 (4.6% cross-linked; H+ form; 6 g.).The column was developed with 0.2~-sodium hydroxide and the effluent collected as fourteen fractions of 10 C.C. each. By chromatography of each fraction on No. 64 Whatman paper with bromo- phenol-blue (see p. 42) two bases were recognised; that of higher RFvalue (a)in fractions 4-7 and the second (b) in fractions e14. Fractions 13 and 14 contained traces of methyl- methioninesulphonium hydroxide and number 14 contained sodium. Glycine (1 g.) was added to fractions67 and the solution refractionated on Zeo-Karb 226 as before. Of the 13 fractions of 10 C.C. which were obtained numbers 2-10 contained glycine and base (a),number 11 contained glycine base (b) and sodium and number (8) only sodium.Fractions 2-9 were combined and evaporated under reduced pressure and the dry residue was extracted with ethanol leaving glycine. The extract was combined with fractions 4-6 from the first fraction- ation. The resulting solution free from base (b) contained halogen. Evaporation in vucuo treatment with picrolonic acid as before and crystallisation from alcohol-Egroin gave a picroEonate m. p. 139-140" (Found C 43.2; H 4.4; N 12.6; C1 8.3. Cl,H,,O,NSC1 requires C 43.0; H 4.3; N 12.6; C1 7.9%). Fractions 8-13 from the first fractionation containing base (b) were halogen-free. Treat-ment as before gave a ficrolonate m. p. 147-148" (Found C 46.2; H 4.7; N 13.3. C,,H,,O,N,S requires C 44-9; H 4.7; N 13.1%). We thank Dr. Margaret I. Whihker for help during the early stages of the work in the preparation of derivatives from the crude natural sulphide also Mrs.P. Harvey who afforded facilities for the collection of urine at the Animals' Clinic Kirkstall Lane Leeds and the late Sir Edward Mellanby who kindly supplied a large quantity from the Nutrition Laboratories of the National Institute for Medical Research Mill Hill,London. D. L. thanks the Department of Scientific and Industrial Research and the University of Leeds for maintenance grants. DEPARTMENTORGANIC OF CHEMISTRY, LEEDS. [Received,July 17th 19ciS.] THEUNIVERSITY Toennies and Kolb J. AM. Chem. SOC.,1945,67,849; Atkinson and Poppelsdorf,J. 1961,1378.