IN view of the considerable interest in our recent paper on the decomposition of organic compounds labelled with tritium1, we now publish a more detailed account of the stability of the important deoxyribonucleoside thymidine, labelled with tritium or carbon-14. Labelled thymidine is one of the most important tracer compounds at present used by research workers throughout the world. The compound is used for labelling the deoxyribonucleic acid (DNA) of proliferating cells and many papers on its use have been published2. A knowledge of the stability of the label in such a compound, under both chemical and biochemical environments, is obviously of great importance. A detailed examination of the decomposition products of tritium labelled thymidine in aqueous solution by self-irradiation is being undertaken in another laboratory3. We have therefore confined ourselves mainly to a study of the gross effects of radiation decomposition of tritiated thymidine labelled generally, in the 6-position (predominantly) and in the 5-methyl group (exclusively), and thymidine labelled in the 2-position with carbon-14. As in our previous paper1, various storage conditions were examined. For comparison, the effect of g-radiation on solutions of tritiated and carbon-14 thymidine and on freeze-dried tritiated thymidine were also investigated.Preparation of labelled thymidine. Thymidine-2-14C, -6-T and -5-methyl-T were prepared by enzymatic coupling of the correspondingly labelled pyrimidine base, namely, thymine-2-14C, -6-T or -5-methyl-T respectively, with 2-deoxy-D-ribose4. Generally labelled tritiated thymidine was prepared by an isotope exchange reaction between thymidine and tritiated water or tritiated acetic acid in the presence of a platinum catalyst5. Such exchange labelling procedures do not give exclusive tritium labelling in the pyrimidine base portion of nucleosides, as has been claimed in earlier papers6. Self-decomposition of labelled thymidine. The various ways by which radio-isotopically labelled compounds can undergo decomposition by self-irradiation are now well described1,7. In the case of thymidine-2-14C we can neglect the decomposition due to primary (internal) decay of carbon-14 to nitrogen over the period of time we have investigated, due to the very long half-life of carbon-14 (5,680 years). The actual decomposition due to this primary (internal) decay is less than 0.01 per cent per annum. Decomposition products of thymidine-2-14C will therefore arise through primary (external), secondary or chemical effects. Decomposition products of tritiated thymidine can arise by all the modes of self-irradiation effects1.Observations with thymidine-2-14C. Thymidine labelled with carbon-14 specifically in the 2-position of the pyrimidine ring at a specific activity of 18.3 me./millimole, had a decomposition rate of only 0.9 per cent per annum. This rate of decomposition, examined over a period of almost three years, was found to be the same whether the compound was stored in aqueous solution or as a freeze-dried solid at +2 C or at -40 C. Identical results were also obtained by storage on paper (cf. ref. 7) under the same temperature conditions. At room temperature stored in aqueous solution the thymidine-2-14C gave 1.2 per cent thymine-2-14C per annum. For comparison we kept a sample of thymidine-2-14C in tap water exposed to air and occasional sunlight for 5 months in the laboratory. At the end of this period analysis showed that the radio -chemical purity was still 92 per cent and 5 per cent thymine-2-14C was found. This remarkable stability of the carbon-14 labelled thymidine contrasts markedly with the stability of the tritiated compound. A comparison of the decomposition of thymidine-2-14C and thymidine-T(G) (G- = generally labelled with tritium in non-labile positions but the labelling is not necessarily uniform) in the freeze-dried state is shown in Table 1. Table 1. DECOMPOSITION OF FREEZB-DRIED THYMiDiNE-2-14C ANDTHYMIDINE-T(G) Specific Activity Wt. of Temp. Radio-activity in thymi- of Age chemical Compound (mc./ ampoule dine storage months puritym.mole) (mc.) (mg) (C) (%) Thymidine-2-14C 18-3 0-1 2-4 -40 33 98Thymidine-2-14C 18-3 0-01 0-24 -40 33 98 Thymidine-2-14C 16-0 0-01 0-15 0 25 100Thymidine-T(G) 98 2 4-9 -40 38 60 Thymidine-T(G) 82 2 5-9 -40 36 50Thymidine-T(G) 82 2 5-9 -40 36 70 Thymidine-T(G) 82 1 2-95 0 36 70Thymidine-T(G) 261 2 1-83 0 39 65 Observations with tritiated thymidine. The decomposition of thymidine generally labelled with tritium, in aqueous solution, has been previously studied by Murray and Peter sen8. They found that after 10 months' storage in aqueous solution at +4 C, thymidine-T(G) at a specific activity of 0.55 curies per millimole gave rise to only two decomposition products which were identified as thymine-T(Gr) and 2-deoxy-D-ribose-T(G). We have confirmed these results by storing thymidine-T(G) in aqueous solution for 39 months at 40 C. Analysis showed that the radiochemical purity of the thymidine-T(G), having a specific activity of 0.26 curies per millimole, at the end of this period was 35 per cent as thymidine-T(G) with 55 per cent of the remaining activity as thymine-T(G) and approximately 10 per cent 2-deoxy-o-ribose-T(G). These results were different from those we obtained on storing thymidine-6-T or thymidine-5-methyl-T in aqueous solution. These specifically labelled thymidines gave relatively small amounts of thymine-T (less than 5 per cent per annum) and mainly a radiation decomposition product(s) of RF 0.0 on paper chromatography of the specifically labelled thymidine samples in ethyl acetate saturated with phosphate buffer at pH. 6.0 (ref. 9) as solvent.The radiation decomposition product (s) exhibited an ultra-violet light absorption spectrum in aqueous solution at pH. 7.0 similar to thymidine, having a maximum at 267268 mand a minimum at 242 m. This compares with the maximum at 267 mand minimum at 235 mfor thymidine under the same pH. conditions. The spectrum was unchanged in acid solution, but in alkaline solution the spectrum was entirely different from that of thymidine20; both the maximum and minimum disappeared and only an inflexion at 275 mwas observed. However, re-acidification of the solution caused the re-appearance of the maximum at 267 and the minimum at 242 mdemonstrating that the decomposition product (s) is not decomposed in alkaline solution. The decomposition product(s) proved quite stable on keeping in aqueous solution at room temperature and remained unchanged after 7 months storage under these conditions. We did not investigate this product (s) further. The decomposition of thymidine-methyl-T and thymidine-6-T at a number of different specific activities, in aqueous solution, are shown in Tables 2 and 5 respectively. We have repeatedly made the observation that thymidine-T appears to undergo more rapid decomposition in the frozen state (0 C and below) than at + 2 C and above1. Thymine as well as its nucleoside and nucleotide are reported10 to be much more sensitive to decomposition under ultra-violet irradiation in the frozen state than in solution. A similar phenomenon seems to apply to decomposition of thymidine-T by self-irradiation. Insufficient decomposition occurred in the case of the thymidine-2-14C to enable similar conclusions to be drawn. The decomposition of freeze-dried thymidine-6-T at a specific activity of 9-5 curies per millimole is shown in Table 3. We were unable to detect any thymine-T formed as a decomposition product in this case, only the radiation decomposition product(s) having EF 0-0 on paper chromatography in the ethyl acetate-phosphate solvent described here was detected. Effect of -g-irradiation on labelled thymidine. The decomposition of thymidine labelled with carbon-14 or tritium in aqueous solution on y-irradiation (2.5 megarads) was as to be expected. Virtually all the thymidine was destroyed during the 30-min irradiation and paper chromatography of the products in ethyl acetate-phosphate as here, gave mainly the radiation decomposition product (s) of EF 0.0. This product (s) exhibited the same ultra-violet light absorption characteristics as the product (s) isolated from the self-decomposition of thymidine-T or carbon-14. A good correlation between the products formed from self-decomposition and those formed by the radiolysis of inactive thymidine in aqueous solution by X-rays at 0 C have previously been reported3'11. We have found that appreciable quantities of thymidine-T survive the Y-irradiation if a free radical scavenger, in this case benzyl alcohol, is present. Some results are recorded in Table 4.Chemical stability of the label. Chemically, thymidine is a relatively stable substance, and it can be heated in neutral or alkaline solution up to 120 C for at least 1 h without decomposition5. This can be used as a method for sterilizing aqueous solutions of tritiated or carbon-14 labelled thymidine. On keeping for long periods in aqueous solution, hydrolytic cleavage of the N-glycoside linkage occurs at a rate which increases with increasing temperature of storage. Thymidine is hydrolysed by mineral acids giving thymine and 2-deoxy-D-ribose17. The carbon-14 label at carbon-2 is lost only by complete disruption of the thymidine molecule. In contrast, tritium atoms in organic molecules are sometimes lost by exchange with the hydrogen atoms of the solvent in which the compound may be dissolved14. This exchange, of course, occurs without changing the chemical form of the compound. However, the tritium atoms in the 6-position and the 5-methyl group of thymidine are firmly bound to the molecule and are not exchanged for hydrogen when the compound is heated in acid or alkaline solution. Only in the presence of a hydrogen transfer catalyst, for example, platinum metal, is the tritium label likely to be exchanged, and then only at elevated temperatures, to any appreciable extent. Biological stability of the label. Provided the thymidine is not metabolized, the carbon-14 label at C-2 remains intact. Chemically stable tritium atoms may become labile under biological conditions14'21. There is at present, however, no evidence to suggest that the tritium label at C-6 or in the 5-methyl group becomes labile and hence lost by exchange in biological systems. In the presence of thymine oxidase the tritium, label at carbon-6 is lost12, due to the formation of methyl barbituric acid, and in systems where demethyIation of the 5-position occurs the tritium label in this position is also obviously lost13. Some thought is therefore necessary when selecting the correct position of isotopic labelling for specific investigations.Table 2. DECOMPOSITION OF THYMIDINE 5 METHYL-T IN AQUEOUS SOLUTION Specific activity Activity in Volume of Wt.of thymi- Temp. of AgeRadio-chemical (mc./ ampoule solution dine storage months puritym.mole) (mc.) (ml.) (mg) (C) (%) 1060 1 1-4 0-23 0 2 1001520 1 2 0-16 0 3 95 3200 2 0-53 0-15 -40 14 854700 2 0-5 0-1 -40 8 92 Table 3. DECOMPOSITION OF FREEZE-DRIED THYMIDINE-6-TWt. of Radio- Sample No.* Age (months) thymidine (nig) Storage state Temperature chemical (C) purity (%)1 2 0-053 Vacuum sealed -40 85 2 2 0-053 Vacuum sealed -80 933 3 0-053 Vacuum sealed room (20) 95 4 3 0-053 Vacuum sealed room (20) 925 3 0-053 Air sealed -80 80 6 4 0-053 Vacuum sealed 0 987 4 0-053 Air sealed 0 92 8 4 0-053 Air sealed 0 859 4 0-053 Vacuum sealed 0 80 10 6 0-053 Vacuum sealed 0 9011 6 0-053 Vacuum sealed 0 65 12 6 0-053 Vacuum sealed -40 7513 6 0-053 Air sealed -40 60 14 6 0-053 Vacuum sealed room (20) 9015 6 0-053 Air sealed room (20) 90 16 6 0-053 Vacuum sealed -80 7017 6 0-053 Air sealed -80 70 * Each sample contained 2 me.Table 4. EFFECT OF Y RADIATION ON THYMIDINE 6-T AND 2-14C (Dose: 2.5. 5-5 x 106 rads for 30 min at room temperature)* Wt. of Vol. of temperature Benzyl Radio-Thymidine-6-T thymidine Sample No. (g) Activity (mec) solution (ml.) alcohol (% by vol.) chemical purity (%) 1 (control) 92 1 1 1 100o 92 1 1 1 70 3 192 1 1 nil<5 4 292 1 1 nil<5 5 392 1 1 1 856 492 1 1 nil<5 7 592 1 1 1 708 692 1 1 nil<5 9 792 1 1 1 8010 892 1 1 nil<5 11 992 1 1 1 7012 (freeze-dried 53 2 98 solid)13 (freeze-dried 53 2 92 solid)Thymidine-2-14C Sample No.1 1,323 0-1 0-83 nil<1 2 132 0.01 0.42 nil<1 * Carried out by the irradiation service U. K Atomic Energy Autherity, Harwell Berks.Not irradiated. Table 5. DECOMPOSITION OF THYMIDINE -6-T IN AQUEOUS SOLUTIONSpecific Activity Radio- Sample activity in Wt.of Vol. of Age Temp. chemical.No. (c./ ampoule thymidine solution (months) (C) purity m.mole) (mc.) (g) (ml.) (%) 1 2-5 2 200 0-29 12 -40 722 2-5 2 200 0-29 7 -40 90 3 2-5 2 200 0-29 8 0 904 1-14 0-5 90 0-25 10 room 100 (20)5 14-1 2 34 0-5 1-5 -40 90 6 14-1 2 34 0-5 12 -40 557 14-1 2 34 0-5 3-5 room 100 (20)8 14-1 1 17 0-5 5 room 97 (20)9 3-6 0-2 134 0-2 10 -40 80 10 1-4 1 170 0-5 10 -40 9011 12-1 0-5 10 0-5 10 0 65 12 4-8 0-2 10 0-5 10 0 70Conclusions. Thymidine-2-14C is stable for a number of years when stored in aqueous solution below + 2 C. The rate of decomposition by self-irradiation under these conditions is less than 1 per cent per annum, but there is no evidence that this state of storage is superior to keeping the compound in the freeze-dried state or on paper7,16,18. As no thymine-2-14C could be detected at this storage temperature, the small amount of thymine-2-14C which is formed on keeping the solution of thymidine-2-14C at room temperature is probably produced by slow hydrolysis of the thymidine rather than by radiolytic fission of the 2V-glycoside bond. Table 1 shows that thymidine labelled with carbon-14 is much more stable than the tritiated compound of comparable specific activity, allowing for the difference in the energy of the carbon-14 p-radiation (max. 0.155 MeV) and the tritium b-radiation (max. 0.018 MeV). In the solid state the self-absorbed energy per unit time for the tritiated thymidine will be greater than for the carbon-14 labelled material, due to the low range and penetrating power of the tritium betas. Primary (external) effects would thus be expected to cause more rapid decomposition in the case of the tritiated thymidine, in agreement with the experimental observations. Thymid-ine-6-T at high specific activity (9.5 curies/m.mole) stored as a freeze-dried solid, kept better than we had expected. The rate of decomposition in this condition, even at room temperature, is only 2030 per cent per annum. The freeze-dried compound is less stable when stored at low temperatures, for example, 80 C. We are unable to explain this sensitivity to radiation at low temperatures at the present time.In general, the presence of air in the ampoules did not affect the rate of decomposition. Even more surprising is the stability of freeze-dried thymidine-6-T to g-irradiation15. Only a few per cent decomposition was recorded, as can be seen from Table 4. No thymine-T or carbon-14 labelled thymine was detected in the decomposition of freeze-dried thymidine-T or carbon-14 respectively, indicating that no appreciable breaking of the N-glycoside bond occurs by primary (external) decomposition. Solutions containing thymidine labelled with tritium or carbon-14 cannot be sterilized by g-irradiation without decomposition of the thymidine, even in the presence of a free radical scavenger. However, in the latter circumstances a high proportion of the compound may survive as is seen from Table 4. The y-irradiation of thymidine solutions produces at least six different products and the main product(s) has Up 0.0 in ethyl acetate-phosphate buffer on paper chromatography. This product (s) is similar to the product (s) produced by self-irradiation of labelled thymidine19. Weiss and collaborators19 have shown that pyrimidine ring structures remain intact on irradiation and do not break down into smaller molecules. The ultraviolet light absorption spectrum of our radiation decomposition product (s) would also suggest this although we did not further investigate the product. The decomposition rates for thymidine-5-methyl-T and thymidine-6-T at comparable specific activities are respectively 0.51 per cent per month and 1.2 per cent per month. This rate of decomposition increases when the solutions are stored below 0 C. Thymine-T is formed as a decomposition product even at low temperatures and probably arises by radiolytic fission of the 2V-glycoside bond and not by hydrolytic fission. Analysis of a number of samples of thymidine-6-T kept in aqueous solution by distillation of the solvent showed that less than 5 per cent tritium is 'exchanged' into the aqueous phase on keeping samples for 1 year. This result again supports the view that practically no degradation of the pyrimidine ring occurs forming smaller molecules which may be volatile or 'leak' their tritium into solution by exchange.Decomposition products will vary to some extent depending on the position of the label within the molecule, which further complicates the interpretation and comparison of such data. The analytical methods used for our radiochemical purity determinations were as previously described1. Inactive thymidine was detected on paper ehromatograms with the ultra-violet light from a 'Chromatolite' Engehard Hanovia lamp (ultra-violet-range 250280 m). We thank Dr. John R. Catch and Mr. D. A. Lambie for their interest and the U.K. Atomic Energy Authority for permission to publish these results