Radioactive Iodoprotein in Thyroid Lymph and Blood

Transcription

1 622 Biochem. J. (1966) 1, 622 Radioactive Iodoprotein in Thyroid Lymph and Blood By P. M. DANIEL, L. G. PLASKETT* AND. E. PRATT Department of Neuropathology, Institute of Psychiatry, The Maudsley Hospital, London, S.E. 5, and the Department of Biochemistry, University of Edinburgh (Received 24 January 1966) 1. Samples of thyroid and non-thyroid lymph and of thyroid and peripheral venous blood were obtained from primates and cats that had previously been given radioactive iodine. The distribution of the organic radioiodine between the protein and non-protein fractions in these samples was determined. 2. The proportion of the organic radioiodine in the form of iodoprotein was assessed by paper chromatography, acid-ethanol precipitation, hot-butanol washing, column chromatography and separate estimation of iodotyrosines after enzymic hydrolysis. 3. In thyroid lymph the relative proportion of the organic radioiodine in the form of iodoprotein was 75-98%; in blood it was much lower, probably no more than 6-7%. The absolute concentration of iodoprotein radioactivity also was considerably greater in thyroid lymph than in blood. 4. Enzymic hydrolysis of the protein of the thyroid lymph yielded a pattern of iodoamino acids that corresponded closely with that obtained after hydrolysis of protein extracted from the thyroid gland itself. 5. It can be concluded that the iodoprotein in thyroid lymph consists mainly of thyroglobulin or a closely related compound. A series of studies have been made on radioactive iodine compounds leaving the thyroid gland in lymph (Daniel, Excell, Gale & Pratt, 1962; Daniel, Gale, Plaskett & Pratt, 1963a,b; Daniel, Gale & Pratt, 1963c,d). The present paper describes the application of chromatography and other techniques to the separation of iodoprotein from the other iodine-containing constituents of the thyroid lymph, peripheral plasma and thyroid venous plasma in the cat and in the baboon. METHODS Preparation of animals for collection of lymph and plasma samples. Adult cats, Felis domesticus (3-4.5kg.), were fed on proprietary cat meat and milk, but no fish. Adult baboons, Papio papio (8-16kg.), were fed on the pellet diet M.R.C. 41B supplemented with milk, cabbage and bananas or apples, but no fish. The animals were given inorganic carrier-free [1251]iodide or [131I]iodide (The Radiochemical Centre, Amersham, Bucks.) in.9% NaCl by intramuscular injection 1 or more days before samples were collected. For the cats, doses of 2-1/zc were used and for the baboons 5-2,c. It was estimated that the radiation doses to the tissue of the thyroid gland did not exceed 35 rads for any of the cats nor 45 rads for any of the baboons. In some experiments thyroid-stimulating hormone (Thytropar; Armour Pharmaceutical Co. Ltd., Eastbourne, Sussex) (2milliunits/kg. intravenously) was * Present address: Brooke Bond Research Laboratories, Blounts Court, Sonning Common, Reading, Berks. administered 1-2 hr. before collection of samples, and this was repeated every 2hr. as necessary. The technique employed for collection of lymph from the small lymphatic vessels draining directly from the thyroid gland (Daniel et al. 1963c) and from the cervical lymphatic trunk (Daniel, Pratt, Roitt & Torrigiani, 1966c) has been described elsewhere. Radiation dose. Estimation ofthe radiation dose delivered to the thyroid tissue of the experimental animals was done by a method (P. M. Daniel &. E. Pratt, unpublished work) based on external monitoring of the neck with a portable rate-meter (Panax TM64G with MX18/1 external probe) and on radioactive assay of a sample of the thyroid tissue obtained post mortem. Paper chromatography. Paper chromatograms, 38cm. long from origin to solvent front, were developed by the descending method. The solvent system butan-l-ol-2nacetic acid (1:1, v/v) was employed whenever it was necessary to effect a separation of mono- and di-iodotyrosine; butan-l-ol-dioxan-aq. 2N-NH3 (4:1:5, by vol.) was employed for chromatograms on which it was necessary to separate individual iodothyronines or to estimate quantitatively radioactively labelled thyroxine. Other systems used occasionally were butan-l-ol-pentan-l-olaq. 2N-NH3 (1:1:2, by vol.), butan-l-ol-aq. 9N-NH3 (1:1, v/v), 2-methylbutan-2-ol-aq. 2N-NH3 (1:1, v/v) and collidine saturated with water in an atmosphere containing NH3. All chromatograms were run on Whatman 3MM paper. Marker compounds were detected on the developed chromatograms by means of diazotized sulphanilic acid or ninhydrin (Gross, 1954) or by ultraviolet-absorption methods (Plaskett, 1964).

2 Vol. 1 IODOPROTEIN IN THYROID LYMPH 623 Assay of radioactivity. Liquid samples were assayed for radioactivity in a Nal-crystal scintillation counter (Ekco type N with crystal N553B) or in a Geiger- Muller,B-ray counter (2th Century Electronics type M6). Paper chromatograms were scanned for radioactivity by cutting them transversely into 38 segments of length 1 cm. and assaying the radioactivity of each segment separately. Sometimes, to define the shape and position of peaks of radioactivity more precisely, -5cm. segments were cut. Chromatogram segments were assayed for 1311 under an end-window Geiger-Muller tube (General Electric Co. type EHM2/S) or assayed for 125I in a NaI-crystal scintillation counter (Ekco type N ). Preparation of labelled compounds. The ion-exchange resin and Zn(OH)2 precipitation procedures described below were tested by subjecting pure preparations of radioactively labelled compounds to analysis, or by analysing known mixtures of such compounds. These were prepared by a biological labelling procedure. Inorganic [1311]- or [1251]-iodide (1-1lc) was administered to rats previously fed on a low-iodine diet (Boyd & Oliver, 196; Barnaby, Davidson & Plaskett, 1965) for 2 weeks or more to ensure high uptake of radioactivity into the thyroid gland. The rats were killed 24-48hr. later and the thyroid glands homogenized. Labelled thyroxine, tri-iodothyronine, monoiodotyrosine and di-iodotyrosine were prepared by hydrolysing the thyroid-gland homogenates with pancreatin and chromatographing the hydrolysates. Peaks of radioactivity having Rp values corresponding to those of the known iodothyronines and iodotyrosines were eluted from the paper. A small sample from each preparation was re-run, together with the authentic carrier compound, on another chromatogram in a different solvent system from that used in the first development. This provided a check on the identity of the labelled material and a check on its purity. The iodoprotein material used in the trial analyses was the supernatant from a centrifuged homogenate of thyroid gland in tris buffer, ph8-45, stored at -2 until immediately before use. Hydrolysis with pancreatin. Rat thyroid-gland homogenates were subjected to hydrolysis by the method of Tong & Chaikoff (1958) as modified by Plaskett, Barnaby & Lloyd (1963). The technique employed for hydrolysis of the iodoproteins obtained from lymph and plasma was the same, except that the high ratio of pancreatin to protein in the sample that had been used in the hydrolysis of thyroid tissue (33mg. of pancreatin to about 2mg. wet wt. of thyroid tissue) could not always be maintained. To have adhered to this ratio throughout would have rendered some preparations too bulky for subsequent chromatographic analysis. In one or two experiments, however, the use of a lower ratio led to incomplete hydrolysis. It is noteworthy that satisfactorily complete hydrolysis was always obtained when the blood or lymph protein had been heated previously or treated with organic solvents, but not when undenatured protein was used. Analysis of lymph and plasma samples. (a) By paper chromatography without pretreatment. Biological fluids can be analysed by direct application to the chromatography paper if the sample size is sufficiently small. When plasma is applied at the origin as a band approx. lin. wide, 2pl. is the upper limit of the sample size that will permit satisfactory development (Fletcher, 1956). Some 7 rh 6 5 Q '5 e1.2._, C O Fig. 1. Distribution of radioactivity in 1 cm. segments of a chromatogram on Whatman 3MM paper of baboon peripheral venous plasma. The solvent system was butan- 1-ol-dioxan-2N-NH3. The range of +S.D. is shown with each measurement. o, Origin; A, iodide; B, thyroxine; C, unknown. chromatograms were run under these conditions but a difficulty is that in experiments with tracer amounts of iodine isotopes the radioactivity of 2,il. of plasma is often insufficient to permit satisfactory location and quantitative determination of the radioactive zones after development, or on the other hand unduly long counting times may be required. In later work, to overcome this difficulty, very wide strip chromatograms were run. By applying the plasma to the paper in the form of a band 6in. wide, across the strip, sample size could be increased to 5-1j1i. The 1 cm. segments cut from these chromatograms were too long to be counted efficiently by means of an end-window Geiger-Muller tube: they were therefore counted by coiling them inside an annular polythene surround (Ekoo N552) to a scintillation crystal (Ekco N553B). A chromatogram record obtained in this way is illustrated in Fig. 1. Thyroid lymph contained much higher concentrations of radioactivity than either peripheral or thyroid venous plasma. Consequently, accurate analyses could be carried out on quite small samples containing less than lmg. of protein, so that thyroid lymph could be chromatographed on narrow or wide paper strips with almost equal facility. (b) By butanol extraction followed by paper chromatography of the extracts and residues. To concentrate the low levels of radioactivity in plasma samples, these were acidified and extracted with an equal volume of butan-l-ol. The organic phase was concentrated in vacuo for chromatography. Esterification or deiodination of iodoamino acids and iodination of lipids during the procedures were kept to a minimum by rendering the extracts alkaline before evaporation and by adding a trace of thiouracil. Iodoprotein was separated from plasma samples by treating them with hot butan-l-ol according to the method of Block, Werner & Mandl (1958) to extract inorganic iodide and iodoamino acid iodine from the residual protein. After such treatment samples of plasma protein were found still to contain radioactive iodine and the chemical nature of the radioactivity was investigated chromatographically after pancreatin hydrolysis of the protein residue. (c) By passage through resin columns. A modification

3 624 P. M. DANIEL, L. G. PLASKETT AND. E. PRATT 1966 of the ion-exchange resin column technique of Galton & Pitt-Rivers (1959), was utilized for the group separation of iodide, iodoamino acids and iodoproteins. Blood plasma or lymph (2-3ml.) was adjusted to ph with 1% (v/v) acetic acid and passed under gravity through a column (5cm.x4mm.) of Dowex 1 (X2; 2-4 mesh) previously equilibrated with -5M-sodium acetate buffer, ph4-5. The sample was washed into the column with water (3 ml.) and eluted with successive portions (2 ml.) of -2Msodium acetate buffer, ph3-6, 5% (v/v) acetic acid and 3M-NaBr until radioactivity in the effluent in each case fell to background level. The total radioactivity in each set of elutions, and in the extruded resin, was measured. The first radioactive peak in the eluted fractions contained a high concentration of protein. The eluates constituting this 'iodoprotein' fraction were combined and concentrated by freeze-drying. The protein was hydrolysed by pancreatin and the radioactivity in the hydrolysate investigated by paper chromatography. For lymph samples a smaller-scale Dowex resin column (3 mm. x 18 mm.) was also used, contained in a short length of polythene tubing from which material could be eluted by using a 5ml. syringe with Luer nozzle. (d) By protein precipitation: (i) Zn(OH)2, (ii) ethanol and (iii) phosphotungstic acid were used. (i) To a sample of plasma (-2ml.) in a tapered 1ml. centrifuge tube were added aq. 1% (v/v) ZnSO4,7H2 (-2ml.) and -5N-NaOH (-2ml.). After centrifuging and washing three times with water (-5ml.), the supernatant and washings, containing the inorganic iodide, were combined and assayed for radioactivity. The precipitate, representing the organically bound iodine, was redissolved in -1ml. of aq. 5% (v/v) acetic acid and water (-3ml.), and, when necessary, assayed for radioactivity separately. (ii) Ethanol (1.6ml.) was added, usually to the redissolved precipitate from the Zn(OH)2 precipitation, or, if inorganic iodide was absent, to a fresh sample of plasma or lymph (-2ml.) to which 1% (v/v) acetic acid (-25ml.) had been added. After centrifuging and washing three times with ethanol (-5ml.) containing 1% (v/v) of 5% (v/v) acetic acid, the combined supernatant and washings were assayed for radioactivity. The precipitate was redissolved in N-NaOH (1 ml.), the tube washed with N-NaOH (1 ml.) and the radioactivity assayed. In this procedure iodoproteins such as thyroglobulin are separated from the commonly occurring iodoamino acids and remain in the precipitate. (iii) Phosphotungstic acid was used to precipitate the protein of thyroid lymph or of thyroid tissue extracts in preparation for its enzymic hydrolysis. To the sample was added 1 or more vol. of aq. 2% (w/v) phosphotungstic acid. After centrifuging and washing twice with water, the residue was hydrolysed with pancreatin and the hydrolysate was investigated by paper chromatography. (e) By iodotyrosine release on hydrolysis. An estimate of the proportion of the radioactive iodine present as iodoprotein was obtained by hydrolysing with pancreatin a crude protein fraction that was prepared either by ionexchange resin separation, hot-butanol washing, or phosphotungstic acid precipitation. The resulting hydrolysate was run on paper either in butan-1-ol-2n-acetic acid or in butan-l-ol-dioxan-2 N-NHs (as in Figs. 2 or 3). The proportion of the radioactivity of the original sample in each iodotyrosine peak was determined. To obtain an 5 S P 4 _P - p, D., 2 Ca3 o 1 Ca O A B C _ Fig. 2. Distribution of radioactivity in 1 cm. segments of a chromatogram on Whatman 3MM paper of a pancreatin hydrolysate of radioactive protein left after extraction of cat peripheral plasma with hot butanol by the procedure of Block et al. (1958). The solvent system was butan-l-oldioxan-2n-nh3., Origin; A, iodotyrosines; B, iodide; C, thyroxine. estimate of the iodoprotein radioactivity from which the iodotyrosines were derived, the values for the two peaks were added together and multiplied by a factor, 1-118, which was calculated from the mean iodoamino acid composition found for iodoprotein extracted from thyroid tissue, so as to allow for the proportion of the iodoprotein radioactivity present in the form of iodothyronines, rather than of iodotyrosines. RESULTS Trials of analytical methods These were carried out in two different ways: by checking the degree of recovery of radioactivity from purified labelled compounds and from the results obtained by repetitive application of these separation procedures to unknown biological samples. Purified labelled compounds The zinc hydroxide precipitation and the ethanol precipitation procedures were tested on -2 ml. samples of non-radioactive cat blood plasma to which had been added either preparations of biologically labelled thyroxine or thyroglobulin or a sample of [31lI]iodide solution, containing in each case sufficient radioactivity to give approx. 2 counts/sec. Iodide. In the absence of added carrier 8-5% of inorganic [131I]iodide was carried down in the zinc hydroxide precipitate. If carrier inorganic iodide was added at a concentration of O- I,g./ml., only 4-7% was carried down in the washed precipitate. This source of error was not serious under the conditions usually encountered with the plasma and lymph samples. Iodoprotein. The results of subjecting the [131I]- iodoprotein (the rat thyroglobulin preparation) to zinc hydroxide precipitation followed by either one

4 Vol. 1 IODOPROTEIN IN THYROID LYMPH 625 Table 1. Recovery of purified radioactively labelled iodoprotein, thyroxine and inorganic iodide in various fractions after successive precipitations, once with zinc hydroxide and once or twice with acid ethanol The radioactive material, containing in each case approx. 2 counts/sec., was added to normal cat blood plasma (-2 ml.). The proteins were precipitated with Zn(OH)2, and the precipitate was redissolved in aq. 12.5% (v/v) acetic acid (-4ml.) and reprecipitated with ethanol (1-6ml.). In some experiments the ethanol precipitation was repeated. The radioactivity was assayed in the two supernatants and in the final precipitate. In some experiments a Zn(OH)2 and two ethanol precipitations were carried out and the final residue, after measurement of the radioactivity present, was redissolved and subjected to a fresh series of analyses. The mean percentage recovery in each fraction is given + S.E.M., with the number of experiments in parenthesis. Radioactivity recovered (%) Sample... Thyroglobulin Analytical method Supernatant from Zn(OH) (9) (11) precipitation (inorganic iodide) Supernatant from first (9) ethanol precipitation (non-protein organic iodine) Residue from first (9) ethanol precipitation (iodoprotein) Combined supernatants (11) from ethanol precipitation repeated a second time (non-protein organic iodine) Residue from a second (11) ethanol precipitation (iodoprotein) Thyro- 'Iodoprotein' Thyroxine globulin residue from preceding column Thyroxine 1-2 (2) (1) (1) 2-8 (2) (1) 96- (2) (1) 'lodoprotein' residue from preceding column 2-9 (2) 62-7 (2) (2) (1) (1) or two ethanol precipitations of the redissolved zinc hydroxide precipitate are shown in Table 1. In general, over 85% of the radioactivity appeared in the final residue from the ethanol precipitation and was thus assayed as iodoprotein. It was important to discover whether the other 15% of the radioactivity lost in the earlier fractions consisted entirely of non-iodoprotein components that were impurities in the original thyroid homogenate or whether some of it represented lost iodoprotein. To investigate this, the ethanol precipitate was redissolved and put through the complete analytical procedure again. The result is shown in column 3 of Table 1, which shows that 96% of the radioactivity behaved as iodoprotein, indicating that only 4% of the true iodoprotein radioactivity was lost in the non-iodoprotein fractions. Thyroxine. Table 1 shows results obtained by the analysis of chromatographically purified [131I]- thyroxine obtained from rat thyroid gland. Most of the radioactivity appeared in the supernatant from the ethanol precipitation, but a variable and significant proportion appeared in the residue, thus making a spurious contribution to the 'iodoprotein' fraction. This effect was greatly diminished by redissolving the residue and subjecting it to a second ethanol precipitation (Table 1, column 5). After this treatment only 8-3% of the total radioactivity appeared in the iodoprotein fraction, but when this latter fraction was subjected to complete reanalysis 62-7% of the radioactivity was found in the non-protein organic iodine fraction (Table 1, column 6). It seems likely that the radioactivity carried down with the protein precipitate did not actually represent thyroxine, for the artifact tended to be larger when the thyroxine solution had been stored for several weeks at 4 and in addition the radioactivity found in the iodoprotein fraction was always less when the ethanol precipitation was repeated. Biological samples The extent to which the protein that was precipitated by ethanol also carried down thyroxine was investigated by repeated fractionation of a mixture of blood plasma and of radioactively-labelled freshly-prepared partly-purified thyroxine from which inorganic iodide had been removed. Let the thyroxine radioactivity be T and the radioactivity

5 626 P. M. DANIEL, L. G. PLASKETT AND. E. PRATT 1966 from paper chromatography were in fairly good agreement. Blood plasma contained 13-31% of the organic radioactive iodine in the iodoprotein fraction, non-thyroid lymph contained -11%, cervical lymph contained 9-61% and thyroid of impurities wholly bound to the protein be U. A constant fraction, a, of the thyroxine will be carried down with the ethanol-precipitated protein in each stage and, provided that no breakdown products of thyroxine (partly bound to protein) are present, the radioactivity in the supernatant after one fractionation will be (1- a)t and that in the precipitate will be U+ at. If this precipitate is redissolved in aqueous solution and the ethanol precipitation is repeated, the radioactivity in the supernatant will be a(l - a)t and that in the new precipitate will be U + oc2t. An assessment of the size of a was obtained by comparing the ratio of the radioactivity in the supernatant from the second ethanol precipitation to that in the supernatant from the first precipitation, i.e. a(l ō)t/ (1- c)t. The fraction a was found to be about -42 and an appropriate correction, equal to *42 of the iodoamino acid iodine radioactivity, was always subtracted from the value found for the iodoprotein radioactivity and added to that found for the iodoamino acid radioactivity. Nature of the radioactive iodine of peripheral pla&ma, thyroid venous plama, thyroid lymph and peripheral lymph Both in cats and baboons the concentration of radioactivity in the thyroid venous blood was higher than it was in the peripheral blood and that in the thyroid lymph was still higher. Compared with the blood the thyroid lymph contained a relatively larger proportion of the radioiodine (usually more than 9%) in organically bound form. lodoprotein content. Some iodoprotein was detectable in all samples of blood, and of thyroid and cervical lymph and in most samples of peripheral lymph by acid-ethanol precipitation, by paper chromatography and, when used, by the hot-butanol washing technique of Block et al. (1958) or by ion-exchange resin chromatography. The results from acid-ethanol precipitation and lymph contained 83-98%. The hot-butanol washing method, when used, gave similar results, but ion-exchange resin chromatography and the method depending on iodotyrosine separation gave lower values for the iodoprotein in blood (Table 2). The samples from one baboon were each subjected to analysis by two different methods and in addition a direct estimate of thyroglobulin radioactivity was obtained (P. M. Daniel,. E. Pratt, I. M. Roitt & G. Torrigiani, unpublished work). The acidethanol precipitation method and paper chromatography gave similar results (Table 3), but a much smaller amount of radioactivity (about 4% in blood and 7% in cervical lymph) is carried down in each case in a thyroglobulin-anti-thyroglobulin immune precipitate made according to the procedure of Daniel et al. (1966c). In spite of this the rather small difference between thyroid venous and peripheral blood plasma on the one hand and the larger one between cervical lymph and peripheral lymph on the other are roughly comparable for all three methods. Nevertheless, the technical difficulties of measuring low concentrations of iodoprotein in blood plasma are such that the values given by the paper chromatography or acidethanol precipitation methods should only be accepted as the upper limits to the actual values. All the techniques for the estimation of iodoprotein [with the exception of the Block et al. (1958) method, which was not employed for lymph] showed that the radioactivity in thyroid lymph was predominantly (7-98%) in the form of iodoprotein. For example, a sample of thyroid lymph from a cat was applied directly to Whatman 3MM paper, a chromatogram was run in butan-1-oldioxan-2 N-ammonia and the radioactivity estimated in each 1cm. segment. About 93% of the Table 2. Iodoprotein radioactivity in thyroid lymph or in peripheral blood estimated by hydrolysis of a crude iodoprotein fraction and separation from it of the iodotyro8ine8 by paper chromatography For comparison values found for the same samples by other methods are given. is shown as the percentage of the total organic radioiodine. Sample Cat thyroid Method lymph Iodotyrosines in protein hydrolysate 74-6 Paper chromatography 8-2 Hot-butanol washing Ion-exchange resin chromatography 83-4 The iodoprotein radioactivity Radioactivity (% of total organic radioiodine) Baboon thyroid lymph Cat peripheral blood plasma Baboon peripheral blood plasma

6 Vol. 1 IODOPROTEIN IN THYROID LYMPH 627 Table 3. lodoprotein in thyroid (cervical trunk) and non-thyroid lymph and in thyroid venous and peripheral venous blood plasma (baboon) The radioactivity in iodoprotein in each sample was estimated both by paper chromatography and by acidethanol precipitation. For comparison the radioactivity carried down from these samples in a thyroglobulinanti-thyroglobulin immune precipitate is also given (P. M. Daniel,. E. Pratt, I. M. Roitt & G. Torrigiani, unpublished work). The concentration of organically bound radioiodine in each of the samples was 6-75 counts/min./g. The iodoprotein radioactivity found is given in counts/min./g. of sample. Radioactivity (counts/min./g.) Venous blood plasma Lymph Non-thyroid Thyroid Non-thyroid Cervical Paper chromatography Acid-ethanol precipitation Immune precipitation radioactivity remained in the three segments adjacent to the origin. The only other peak of radioactivity found was in the same position as that of an added non-radioactive thyroxine marker. A proportion of the radioactivity in peripheral blood was also retained near the origin after paper chromatography. It should be noted that not only was this iodoprotein radioactivity a larger proportion of the total radioactivity in the thyroid lymph than it was in peripheral blood plasma but because of the much greater concentration of radioactivity in thyroid lymph the concentration of iodoprotein in it was often several hundred times greater than the concentration in peripheral plasma. Nature of the peptide-linked iodoamino acids in the iodoprotein fraction of plama and lymph Hydrolysis of the iodoprotein fractions from plasma and lymph, followed by chromatography of the hydrolysates, served to identify the iodoamino acids that formerly had been peptide-linked. The finding of known iodoamino acids that were only demonstrable after hydrolysis established the true iodoprotein character of these fractions. In all experiments on the hydrolysis of the plasma iodoprotein fraction peripheral plasma was used. After separation of protein from low-molecularweight iodine compounds by the butanol extraction method of Block et al. (1958), the protein was subjected to hydrolysis with pancreatin. The results of chromatographing such a hydrolysate showed that some iodotyrosines were present (Fig. 2), but there was more radioactivity in the thyroxine than in the iodotyrosines. The total radioactivity in the iodotyrosine fraction represented approx. 6-5% of the total radioactivity in the original plasma sample. A sample of cat peripheral serum was passed through a Dowex resin column and the fraction o Ca Fig. 3. Distribution of radioactivity in 1 cm. segments of a chromatogram on Whatman 3MM paper of crude iodo. protein, subjected to hydrolysis by pancreatin. The solvent system was butan-1-ol-2n-acetic acid. The crude iodoprotein was obtained from cat peripheral plasma, after administration of thyroid-stimulating hormone and thyroid massage, by separation on Dowex 1 column (see the text). O, Origin; A, monoiodotyrosine; B, di-iodotyrosine; S, solvent front. that failed to be adsorbed on the column contained about 7% of the organically combined radioactivity of the original sample. This fraction, from which inorganic iodide and iodoamino acids had been removed, was subjected to hydrolysis with pancreatin, followed by butanol extraction of the hydrolysate. When this extract was run on a paper chromatogram the radioactivity appeared in several peaks not coinciding exactly with known marker compounds (Fig. 3). The bulk of protein to be digested in this sample was very large so that an unusually low ratio of pancreatin to protein in the sample had to be used. Therefore it is likely that the poor separation was due to incomplete

7 628 P. M. DANIEL, L. G. PLASKETT AND. E. PRATT 1966 digestion of the protein. Nevertheless, after pancreatin digestion a proportion of the radioactivity that had previously remained at the origin now 6, migrated in the solvent system. 64 ( The hydrolysis products of the iodoprotein in AI ~~~~~D thyroid lymph were more readily identified because 2 2 the radioactivity of iodoprotein/mg. of protein was very much greater in this thyroid lymph than in blood. Figs. 4 and 5 show radioactivity scans of chromatograms of the hydrolysed protein of thyroid lymph from a cat and a monkey respectively, compared with similar scans from hydrolysates of thyroid tissue protein from the same animals. The percentage of the total radioactivity found in each iodoamino acid is shown in Table 4. The composition of the lymph iodoprotein did not differ significantly from that of thyroid-gland iodoprotein. Fig. 4. Distribution of radioactivity in 1cm. segments of DISCUSSION chromatograms on Whatman 3MM paper of (a) hydrolysed cat-thyroid lymph protein compared with (b) hydrolysed cat-thyroid tissue protein. The solvent system was butan- 1-ol-2 N-acetic acid. o, Origin; A, iodide; B, monoiodotyrosine; a, di-iodotyrosine; D, iodothyronines; 5, solvent front..6 o 1 m -4) - O.p._'. Ca.S. 4 - (a) B c D 1- VVA El Fig. 5. Distribution of radioactivity in 1 cm. segments of chromatograms on Whatman 3MM paper of (a) hydrolysed baboon-thyroid lymph protein compared with (b) hydrolysed baboon-thyroid tissue protein. The solvent system was butan-1-ol-2n-acetic acid. o, Origin; A, iodide; B, monoiodotyrosine; C, di-iodotyrosine; D, iodothyronines; S, solvent front. An unexpected finding from our studies was that high concentrations of iodoprotein are present in thyroid lymph (Daniel et al. 1963a,b). Early theories that thyroxine might leave the thyroid gland in a peptide-linked condition (e.g. Harington, 1933) were put aside by most workers when it was shown that the main component of the serum iodine was free thyroxine (Taurog & Chaikoff, 1947; Laidlaw, 1949). The evidence of the present work is therefore contrary to most current opinion and it is important to assess the weight of whatever previous evidence exists that conflicts with the present findings. We have been unable to trace any previous studies on the composition of pure thyroid lymph. In one instance cervical lymph of the dog, part of which must have drained from thyroid tissue, was found to contain iodoprotein (Dobyns & Hirsch, 1956). We have found no other comparable work on thyroid lymph. All investigators now agree that some time after the administration of radioactive iodine the chief component of the organic iodine fraction in the blood plasma is thyroxine and not iodoprotein. However, few workers have been able to account for 1% of the radioactivity in blood plasma in terms ofknown compounds. For example, Wynn, Fabrikant & Deiss (1959) devised a Dowex resin column procedure for separating the iodine fractions of plasma and obtained a quantitative elution from these columns, but in preparing for the elution procedure they found that only 88-98% of the total radioactivity in plasma could be adsorbed on the resin. The identity of 2-12% of the radioactive compounds therefore remained obscure. Similarly, butanol extraction procedures applied to blood plasma do not account for all of the radioactivity. If iodoprotein were present in biological fluids,

8 Vol. 1 IODOPROTEIN IN THYROID LYMPH 629 Di8tribution of radioactivity between the various iodoamino acid fractions in hydrolysate8 Table 4. of iodoprotein derived from thyroid lymph and from thyroid ti88ue in the cat and the baboon Source of protein Cat thyroid lymph Cat thyroid tissue Baboon thyroid lymph Baboon thyroid tissue Radioactivity (% of total organic radioiodine) Monoiodotyrosine * 42-1 Di-iodotyrosine lodothyronines *6 it would be particularly likely to resist butanol extraction (Robbins, Rall, Becker & Rawson, 1952) and to pass unchanged through columns of ionexchange resin, so that its presence could easily be missed with many of the usual analytical procedures. Usually a few per cent of the organically bound iodine is not accounted for and will contain any iodoprotein that may be present. The findings of the present work do not therefore conflict with those of previous authors. There is, however, need for caution in accepting high values for iodoprotein in blood, for in the butanol extraction method of Block et al. (1958), even though thiouracil is added, there is a danger of the hot acid conditions causing iodination of protein at the expense of the inorganic iodide fraction. In ion-exchange resin columns incomplete adsorption of components that are usually adsorbed can lead to increase in the apparent iodoprotein fraction. In paper chromatography the mass of dried protein at the origin may be penetrated only with difficulty by the solvent, so that some low-molecular-weight compounds within it may fail to migrate and may thus be wrongly taken to be iodoprotein. Confirmation that part of the radioactivity in the protein fraction of blood from animals previously given radioactive iodine is in fact due to iodoproteins is provided by the presence of iodotyrosines in a hydrolysate of the butanol-extracted protein fraction (Fig. 2). By far the greater part of any free iodotyrosines present in the original sample would have been removed by the acid butanol. However, acid-butanol extracts of plasma did not contain detectable levels of radioactivity in the form of iodotyrosines, so that it seems unlikely that any was present in the original plasma. Therefore it may be concluded that the iodotyrosine fraction identified in the hydrolysate was released from peptide linkage during hydrolysis. On the other hand, it is unlikely that more than a part of the thyroxine found in the hydrolysate was derived from the breakdown of iodoprotein, for no known iodoprotein has a ratio of thyroxine to iodotyrosines high enough to account for the proportions in which they have been found in the hydrolysates. It therefore seems likely that much of the thyroxine in the hydrolysate represented free hormone from the original sample that had not been completely removed by the acid-butanol extraction. On the other hand, all the iodotyrosines, and probably part of the thyroxine, will have been released by the hydrolysis of iodoprotein. Two other methods were used for investigating the distribution of radioactive iodine in the blood and lymph samples, the zinc hydroxide precipitation method and the ethanol precipitation method, and these were best applied serially in this order to the same sample so as to avoid interference by inorganic iodide in the ethanol precipitation. By this means a separation can be made into three fractions representing inorganic iodide, thyroxine and iodoproteins in which each fraction is contaminated with other fractions to an extent that is either negligible or for which suitable corrections can usually be made. However, when these methods are used for the study of samples that contain relatively high proportions of inorganic iodide and thyroxine, e.g. blood samples, the contamination of the iodoprotein fraction by these substances is likely to be serious and the values obtained must be regarded as upper limits. That some iodoprotein is present in blood and lymph samples is shown by all the methods used, although nearly always the experimental errors tend to exaggerate the proportion of the radioactivity attributable to iodoprotein, and the values, especially for iodoprotein in the blood and cervical lymph, must be regarded as maximum values. In thyroid lymph, on the other hand, the proportion of the radioactivity present in the iodoprotein fraction is so high as to render any analytical artifact unimportant. The estimation of iodoprotein radioactivity by the method depending on hydrolysis of crude iodoprotein and separation on paper of the iodotyrosines thus released, though indirect and laborious, is likely to give more reliable answers than the other methods. This method can be applied to an iodoprotein fraction already partially purified, and a high degree of specificity will be obtained provided that the original blood or lymph samples do not contain appreciable amounts of iodotyrosines and

9 63 P. M. DANIEL, L. G. PLASKETT AND. E. PRATT 1966 which indeed we have not been able to detect in significant quantities. The values given for peripheral blood plasma by this method (iodoprotein 6-7% of total organic radioactivity) are lower than those given by any of the other methods used except for the immune precipitation method (Table 3), the results of which may not be strictly comparable because of the possible failure of some of the iodoprotein to react with the antithyroglobulin. Another method has been devised more recently for the separation of iodoprotein by using a Sephadex G-2 column (Daniel et al. 1966c) and our experience suggests that this may prove to be one of the best methods for future work. Although the values given by some methods for the concentration of iodoprotein may be too high, in general our results suggest that at least 6-7% of the organic radioiodine in blood and some 75% ofthat in thyroid lymph is in the form ofiodoprotein. The close similarity of the products of hydrolysis of the iodoprotein in thyroid lymph with the products of hydrolysis of the protein extracted from the thyroid gland itself suggests that the main component of the iodoprotein in thyroid lymph closely resembles thyroglobulin. In the present experiments all the animals had been given radioiodine and the possibility has to be considered that radiation damage to the thyroid tissue may have caused the release of thyroglobulin into the thyroid lymph or increased the amount that was already present in the lymph. On the other hand, appreciable concentrations of thyroglobulin have been detected by radioimmunoassay in the peripheral blood plasma or in the thyroid lymph of man, monkey and rat (G. Torrigiani, I. M. Roitt & D. Doniach, personal communication; Daniel, Pratt, Roitt & Torrigiani, 1966a,b,c), when radioiodine had not previously been given and thus when there was no possibility of radiation damage to the gland. The presence of comparatively large concentrations of radioactivity in the iodoprotein of thyroid lymph suggests that, especially for this compound, the lymphatic pathway is important. A comparison of the relative roles of the lymphatic and venous systems in the outflow of iodine compounds from the thyroid gland (Daniel, Plaskett & Pratt, 1966d) has confirmed that the lymphatic pathway is relatively more important for iodoprotein than for iodoamino acid iodine or for inorganic iodide. We are grateful to Professor R. B. Fisher for reading the manuscript and for helpful suggestions and to Miss M. M. Gale for valuable assistance. Mrs M. Waite gave skilled technical help. Thanks are due to the Nuffield Foundation, the Royal Society and the Fleming Memorial Fund for Medical Research (P. M. D.), and the Research Fund of the Bethlem Royal and Maudsley Hospitals (P. M. D. and. E. P.), for grants in aid of this work. REFERENCES Barnaby, C. F., Davidson, A. M. & Plaskett, L. G. (1965). Biochem. J. 95, 811. Block, R. J., Werner, S. C. & Mandl, R. H. (1958). Arch. Biochem. Biophy8. 73, 9. Boyd, G. S. & Oliver, M. F. (196). J. Endocrin. 21, 25. Daniel, P. M., Excell, B. J., Gale, M. M. & Pratt,. E. (1962). J. Phy8iol. 16, 6P. Daniel, P. M., Gale, M. M., Plaskett, L. G. & Pratt,. E. (1963a). J. Phy8iol. 165, 65P. Daniel, P. M., Gale, M. M., Plaskett, L. G. & Pratt,. E. (1963b). Nature, Lond., 198, 392. Daniel, P. M., Gale, M. M. & Pratt,. E. (1963c). Quart. J. exp. Phy8iol. 48, 138. Daniel, P. M., Gale, M. M. & Pratt,. E. (1963d). J. Phy8iol. 169, 33. Daniel, P. M., Plaskett, L. G. & Pratt,. E. (1966d). J. Phy8iol. (in the Press). Daniel, P. M., Pratt,. E., Roitt, I. M. & Torrigiani, G. (1966a). J. Phy8iol. 183, 15P. Daniel, P. M., Pratt,. E., Roitt, I. M. & Torrigiani, G. (1966b). J. Phy8iol. 183, 33P. Daniel, P. M., Pratt,. E., Roitt, I. M. & Torrigiani, G. (1966c). Immunology (in the Press). Dobyns, B. M. & Hirsch, F. Z. (1956). J. clin. Endocrin. Metab. 16, 153. Fletcher, K. (1956). Ph.D. Thesis: University of London. Galton, V. A. & Pitt-Rivers, R. (1959). Biochem. J. 72, 31. Gross, J. (1954). Brit. med. Bull. 1, 218. Harington, C. R. (1933). The Thyroid Gland, pp London: Oxford University Press. Laidlaw, J. C. (1949). Nature, Lond., 164, 927. Plaskett, L. G. (1964). Chromat. Rev. 6, 91. Plaskett, L. G., Barnaby, C. F. & Lloyd, G. I. (1963). Biochem. J. 87, 473. Robbins, J., Rall, J. E., Becker, D. V. & Rawson, R. W. (1952). J. clin. Endocrin. Metab. 12, 856. Taurog, A. & Chaikoff, I. L. (1947). J. biol. Chem. 171, 439. Tong, W. & Chaikoff, I. L. (1958). J. biol. Chem. 232, 939. Wynn, J., Fabrikant, I. & Deiss, W. P. (1959). Arch. Biochem. Biophys. 84, 16.