123
J. Phygiol. (1976), 257, pp. 123-136 With 6 text-figures Printed in Great Britain
THE TRANSCAPILLARY EXCHANGE OF THYROID HORMONES AND THYROXINE-BINDING PROTEINS BETWEEN BLOOD AND TISSUE FLUIDS
By M. W. SIMPSON-MORGANt AND R. L. SUTHERLAND* From the Department of Experimental Pathology, the John Curtin School of Medical Research, the Australian National University, Canberra, Australia
(Received 14 August 1975) SUMMARY
1. A study has been made of the relative importance of protein bound and unbound hormone in the exchange of thyroid hormones between blood and interstitial fluid. 2. When [l25J]thyroxine (or triiodothyronine) and [1311]human serum albumin were injected simultaneously into the circulation of sheep with chronic lymphatic fistulae, the thyroid hormones were removed from the circulation and appeared in all lymph samples at a greater fractional rate than human serum albumin. 3. The steady-state lymph/plasma concentration ratios of the two specific thyroxine binding proteins were similar to each other and to those of albumin and total thyroxine. 4. Gel filtration studies indicated that the two specific thyroxine binding proteins, ovine serum albumin and human serum albumin, were all of similar molecular size. 5. Concentrations of unbound thyroxine in plasma and various samples of lymph from the one animal were similar. 6. Increasing the proportion of thyroid hormone that was unbound resulted in an increased rate of equilibration of labelled hormone between blood plasma and lymph. 7. Perfusion of the popliteal lymph node demonstrated that thyroid hormones were removed from lymph during its passage through the node. The amount removed was related to the proportion of hormone in the unbound state. * Present address: Department de Chimie Biologique, Facult6 de M6decine de Bicetre, 94270 Bicetre, France. t Present address: Department of Animal Husbandry, Veterinary School, University of Queensland, St Lucia, Brisbane 4067, Australia.
124 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND 8. It was concluded that thyroid hormones cross capillary endothelium mainly as the unbound molecule and that such movement is bidirectional. INTRODUCTION
In the plasma of sheep the majority of circulating thyroxine is bound tightly but reversibly by three thyroxine-binding proteins. The thyroid hormone binding properties of these proteins are similar to those of human thyroxine-binding globulin (TBG), thyroxine-binding prealbumin (TBPA) and serum albumin respectively. As a result of protein binding only 006 % of the total plasma thryoxine circulates as the unbound molecule (Sutherland & Simpson-Morgan, 1975). The physiological role of plasma thyroxine-binding proteins in the transport and distribution of thyroid hormones is not yet fully understood. Before entering the tissue fluids, and ultimately the tissue cells, thyroid hormones must first cross the endothelium of blood capillaries. Since binding to specific plasma proteins is likely to impede such movement, it was considered important to examine quantitatively the effect of protein binding on the rate of transfer of thyroid hormones from blood to interstitial fluid. This was attained by measuring the relative rates of appearance of radioactively labelled thyroid hormones and serum albumin in the lymph of chronically cannulated sheep, following the simultaneous i.v. injection of the two labelled compounds. The work presented in this communication is an extension of that previously reported by other workers from this laboratory (Irvine & Simpson-Morgan, 1974). METHODS
Experimental animals. Adult male and female Merino and Merino cross sheep weighing 35-55 kg were housed indoors in individual pens or metabolism cages and received lucerne chaff, oats and water ad libitum. Surgical procedures. Chronic lymphatic fistulae were established in the afferent and efferent vessels of the popliteal lymph node, the afferent testicular lymphatics, the lumbar lymphatic trunk, the intestinal lymphatic trunk, the efferent hepatic lymph duct, the deep cervical lymph duct and the efferent lymphatics from the prefemoral and prescapular lymph nodes by surgical procedures which have been described (Lascelles & Morris, 1961; Hall & Morris, 1962; Morris & McIntosh, 1971). Lymph was collected into plastic bottles or plastic centrifuge tubes tied to a plastic holder sutured to the animal's skin. Powdered heparin was used as an anticoagulant. Lymph was centrifuged for 10 min at 2000 rev/min and the cell free lymph separated, and frozen at - 200 C until required. Samples of afferent hepatic and afferent renal lymph were kindly donated by Dr G. G. MacPherson. Blood was collected through chronic indwelling jugular vein cannulae. Both serum and plasma were prepared. Materials. ['251]L-thyroxine and [125I]L-3, 3',5-triiodothyronine in 50°% propy-
THYROID HORMONE DISTRIBUTION
125
lene glycol, with an initial specific activity of 40-45 mc/mg, containing 4-5 4ug hormone/ml. were supplied at 2-mnonthly intervals by the Radiochemical Centre, Amersham, England. [131I]sodium iodide (Atomic Energy Commission, Lucas Heights, New South Wales, Australia) was used to label purified ovine and human serum albumin by the method of Helmkamp, Goodland, Bale, Spar & Mutschler (1960). Human serum albumin, 10000/ pure on electrophoresis was supplied by Hoechst Australia Ltd, Sydney, Australia. Sheep fraction V (Commonwealth Serum Laboratories, Parkville, Victoria, Australia) was purified as described by Sutherland, Brandon & Simpson-Morgan (1975). L-thyroxine sodium was supplied by B.D.H. Chemicals Ltd, Poole, England. Transcapillary transfer of thyroxine-binding proteins and thyroid hormones in sheep. Lymphatic cannulae were allowed to flow for 24 hr before an experiment by which time the animals had completely recovered from surgery and were eating and drinking normally. A 3 ml. volume of the animal's own plasma was mixed with approximately 15 Arc [125I]thyroxine or triiodothyronine and 10 /sc ['311]human serum albumin. This mixture was shaken and dialysed overnight at 2° C against a slurry of Amerlite IRA-400 resin as described by Nicoloff & Dowling (1968). Irvine & Simpson-Morgan (1974) demonstrated that after this type of dialysis the tracer contains no iodide, 1-5-2-0 0 of the [1251] as triiodothyronine and the remainder as thyroxine. [125I]triiodothyronine was essentially pure following similar treatment. The radioactive dose was weighed accurately into a sterile syringe and infused into the jugular vein. The cannula was rinsed several times with blood and sterile heparinized saline. Dose time and dose weight were recorded and a portion of the injected material was retained for radio-assay. Blood and lymph samples were collected at timed intervals after injection and plasma separated. Lymph flow rates were recorded. Where volumes permitted 3-0 ml. plasma or lymph was counted in a gamma scintillation spectrometer system after protein precipitation with trichloroacetic acid. Results were expressed as the percentage of the dose remaining/l. plasma or lymph at any fixed time after injection. In some experiments a large dose of thyroxine (5 mg) was injected 72-98 hr after the tracer injection. At this time the thyroxine and human serum albumin had attained isotopic equilibrium as judged by parallel isotope disappearance slopes in blood and lymph. Samples of blood and lymph were collected at short time intervals after injection. In experiments where [125I]triiodothyronine replaced labelled thyroxine, radioactively labelled contaminants were removed by anion exchange chromatography (Sutherland & Irvine, 1973). Perfusion of popliteal lymph nodes. An afferent popliteal lymphatic accompanying the recurrent tarsal vein was cannulated in the direction of lymph flow. Sterile Evan's Blue dye (0-2 ml.) was slowly infused up the afferent lymphatic and all efferent lymphatics were identified, tied off, and one cannulated. A 24 hr collection of sterile lymph was made after clearance of the dye. Cells were spun off and the lymph plasma retained. An infusion system was set up whereby the sterile efferent lymph labelled with [1251]thyroxine or [125I]triiodothyronine (0-2 flc/100 ml.) and [131I]ovine serum albumin (0-2 ,uc/O0 ml.), was infused into the afferent lymphatic at the rate of 1 ml./hr. Varying concentrations of unlabelled thyroxine (0-1 0 g/ml.) were added to the infusion mixture. Each infusion was continued for 4 hr with a 1 hr sterile saline wash between treatments. Efferent lymph was collected over half hourly intervals and the results were expressed as the thyroid hormone: albumin activity ratio in efferent lymph/the thyroid hormone: albumin activity ratio in the infusion mixture.
126
M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND
Analytical procedures. The concentrations of the specific thyroxine-binding proteins, albumin and total thyroxine in blood plasma and lymph were measured by techniques described previously (Sutherland & Simpson-Morgan, 1975). Radioactivity measurements in samples containing both 125I- and '31I-labelled compounds were made in a two-channel Packard Gamma Scintillation Spectrometer, Model 578. Windows were chosen such that 131J counted independent of 125I uhile a constant proportion of 131I (16 %) counted in the 1251 channel. Suitable volumes were used such that the efficiency of counting did not change between samples. Normal and heat treated (1 hr at 600 C) sheep serum samples, labelled with [125I]thyroxine and [131I]ovine serum albumin, were fractionated on 2 0 x 90 cm Sephadex G-200 columns according to the method of Andrews (1964). The flow rate was maintained at 10*0 ml./hr by a peristaltic pump. RESULTS
The rate of movement of [12-I]thyroxine ard [1311]human serum albumin from blood to lymph. Following simultaneous i.v. injection of labelled thyroxine and human serum albumin, tracer thyroxine was removed from the blood plasma and appeared in lymph at a greater fractional rate than did labelled albumin (Fig. 1). The rates of transfer of labelled thyroxine and albumin from blood to lymph during a given collection period were estimated from the ratios of the activities in the lymph, to the activities in plasma at the mid-point of that collection period. These ratios which are expressed as percentages in Table 1, varied considerably when lymph was collected from different lymphatics in the same animal, or from the same lymphatic in different animals. However, in any lymphatic the net relative rate of movement of thyroxine from blood to lymph was greater than that of human serum albumin during the first, 1 hr collection period. This difference was relatively constant for any particular lymph pool but differed significantly between different tissue fluid pools due possibly to differences in capillary permeability. The lymph:plasma concentration ratiosfor total thyroxine and the thyroxinebinding proteins. Plasma and lymph samples from twenty-four animals, representing ten different tissue fluid pools were analysed for total thyroxine, albumin and thyroxine-binding protein concentrations. The results are summarized in Table 2 where the lymph plasma concentration ratios of those substances are given. In all samples studied three thyroxinebinding proteins were identified and their association constants corresponded to those of thyroxine-binding globulin (TBG), thyroxine-binding protein-2 (TBP-2) and albumin of sheep serum (Sutherland & SimpsonMorgan, 1975) In the cases of afferent testicular and efferent popliteal lymph, which were studied in a number of animals, there were no significant differences between the mean lymph: plasma concentration ratios of total thyroxine, albumin and the two specific thyroxine-binding proteins. This indicated the absence of a preferential transcapillary transfer
127 THYROID HORMONE DISTRIBUTION mechanism for specific thyroxine-binding proteins. The data from the other eight tissue fluid pools were consistent with such an hypothesis. Gel filtration studies on sheep thyroxine-binding proteins. Since the measured lymph:plasma concentration ratios for the three binding pro100
50 0I 0
10
100 50
°0 10
12 Time (hr)
Fig. 1. The disappearance of [1251]thyroxine (A) and [13Ijhuman serum albumin (+) from plasma and their subsequent appearance in afferent testicular lymph (upper diagram) and intestinal lymph (lower diagram), of two different animals, following simultaneous intravenous injection of the two isotopes. [1251]thyroxine radioactivity in testicular (U) and intestinal lymph ([]). [1311]human serum albumin radioactivity in testicular (*)and intestinal lymph (0).
teins indicated that these proteins were of similar molecular size attempts were made to confirm this using gel filtration. The protein and radioactivity elution profiles for typical separations of normal and heat treated sheep serum are shown in Fig. 2. In both cases the optical density and 5
PHY 257
128 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND radioactivity peaks for albumin coincided. The [125I]thyroxine peak occurred one tube earlier than albumin in the case of normal serum and one tube later for heat treated serum. In normal sheep serum about 55 % of endogenous thyroxine is bound to TBG. After heat treatment (1 hr at 60° C) binding to TBG is completely destroyed and 70 % of tracer thyroxine TABLE 1. The relative rates of movement of [125I]thyroxine and [131I]humpn serum albumin from blood to lymph during the first j hr after injection of the labelled dose Lymph:plasma activity ratios (%) Lymph sample Lumbar trunk Afferent testicular Efferent prescapular Efferent popliteal Intestinal trunk Efferent prefemoral
8-18
albumin 6*61
Relative rate of transfer.t rIl'yroxine Human serum albumin 1-25
4
7.92 + 6.32*
6-33 ± 5*74
1-31+ 0*13
1
1*38
0-416
3*32
4
2*73 ± 1*92
0*768 + 0*389
3-42 ± 1*85
1
4-81
0-91
5-29
1
8-35
1-56
5.35
No. of animals 1
[1311]
[121] Thyroxine
Hluman serum
Mean+s.D. t A ratio greater than 1 indicates more rapid transcapillary exchange of thyroxine *
relative to human serum albumin.
TABLE 2. Lymph: plasma concentration ratios for total thyroxine, the two specific thyroxinebinding proteins (TBG and TBP-2) and serum albumin in samples of lymph from several tissue fluid pools of sheep Total No. of TBG Lymph sample samples thyroxine TBP-2 Albumin 1 Afferent hepatic 0*780 0*909 0-979 0*820 1 0-667 0-858 Efferent hepatic 0*858 0-828 Afferent testicular 9 0.757 + 0.187* 0*782 ± 0*133 0-792 + 0*163 0*841 ± 0-141 1 0-732 0-508 Efferent prescapular 0-781 0-774 1 0*557 0*663 Efferent prefemoral 0*683 0*773 1 0*598 Cervical lymph duct 0-775 0*686 0-558 1 0*400 Lumbar lymph trunk 0-662 0-635 0X674 1 0*418 Afferent renal 0-390 0*572 0-549 8 0*503 + 0*188 0*418 ± 0*160 0*482 ± 0-226 0*473 ± 0*167 Efferent popliteal 1 0*250 0*487 Intestinal lymph 0*425 0*415 trunk *
Mean+ s.D.
129 THYROID HORMONE DISTRIBUTION is associated with TBP-2. Thus the similar elation volumes for ovine serum albumin and the two different [125I]thyroxine peaks indicated that the three thyroxine-binding proteins of sheep serum were of similar molecular size. Human serum albumin and ovine serum albumin were shown to have similar elution patterns from Sephadex G-200 columns. 10,000
A
AA~~~~~~4 ~0
A
C
5000
V
A
10
30
20 A&
O0 10.000
20 1e ~ EA
JA
a~.A
1a
40
100
t
7
di
i
0
C
EA 30
I
30 20 Fraction number
40
L\
100
40
Fig. 2. Protein (A) and radioactivity elution profiles for Sephadex G-200 separations of normal (upper diagram) and heat treated (lower diagram) sheep serum labelled with tracer [L25I]thyroxine (@) and [L31I]ovine serum albumin (A\). Heat treated serum was exposed to 600 C for 1 hr.
Unbound thyroxine concentrations in plasma and lymph. Although the total thyroxine concentration varied appreciably between plasma and various samples of lymph from the one animal, the concentration of unbound thyroxine was relatively constant from sample to sample. Hence the proportion of total thyroxine that was unbound was inversely proportional to the total thyroxine concentration and varied from 0056 + 0019 % (mean + S.D.) in plasma to 0 125 + 0066 % in popliteal lymph. The thyroxine binding curves for various samples of lymph from the one animal were essentially parallel to the plasma curve, indicating that the 5-2
130 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND thyroxine-binding proteins of lymph were a simple physiological dilution of those in plasma (Fig. 3). The rate of movement of [125I]triiodothyronine and LulI]human serum albumin from blood to lymph. Triiodothyronine is less firmly bound to sheep plasma proteins than is thyroxine. The concentration of unbound triiodothyronine in normal sheep is approximately 0-2 ng/100 ml. or 0-25 % of the total (R. .L. Sutherland, unpublished). This is 4-5 times greater than the proportion of thyroxine which is unbound. Following administration of labelled triiodothyronine this hormone waa removed from the circulation and equilibrated with lymph much more rapidly than thyroxine in the preceding experiments (Fig. 4). 2500
-
E~~~ A
a'1000 1000
s
-
C
0
10
1
111li
2
5
10 100 1000 2000 Unbound T4 (ng/100 ml.)
Fig. 3. Thyroxine binding curves (log protein-bound concentration v8. log unbound concentration) for diluted samples (1:150) of blood plasma (A), afferent testicular lymph (*) and efferent popliteal lymph (,A) collected from the same animal.
The effect of elevated unbound thyroid hormone concentrations on the rate of movement of labelled hormone from blood to lymph. A ewe with an efferent popliteal lymphatic fistula was injected sequentially with tracer [1251]thyroxine, tracer [125I]triiodothyronine, and a loading dose (tracer + 5 mg) of thyroxine, at 72 hr intervals. The latter treatment was estimated to raise the total plasma thyroxine concentration to approximately 300 ,ug/ 100 ml. and the percentage unbound thyroxine to 0-23 %. The isotopic equilibration curves for the 12 hr following injection of each dose are shown in Fig. 5. The loading dose of thyroxine, when compared with the tracer thyroxine
131 THYROID HORMONE DISTRIBUTION dose, resulted in the more rapid movement of [1251]thyroxine out of the circulation during the first hour. This was accompanied by the more rapid appearance of [126I]thyroxine in popliteal lymph during the first hour following the loading dose. Subsequently [1251]thyroxine was removed from the circulation at the same fractional rate with both treatments. 1oo
0
0A
10
0o
1*0
04 24 Time (hr)
100
10 I
0
10
0o1
0
12
24
Time (hr)
Fig. 4. The disappearance of [125I]triiodothyronine (A) and ['31I]human serum albumin (+) from plasma and their subsequent appearance in afferent testicular lymph (upper diagram) and efferent popliteal lymph (lower diagram) following the simultaneous i.v. injection of the two isotopes into one animal. [125I]triidothyronine radioactivity in testicular (A) and popliteal lymph (>). [131I]human serum albumin radioactivity in testicular (@) and popliteal lymph (0).
132 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND Equilibration between plasma and lymph occurred earlier following the loading dose but following equilibration the lymph:plasma ratios for radioactive thyroxine were the same in both cases. Despite the fact that almost identical proportions of hormone in the circulation were unbound following the loading thyroxine (0.23 %) and tracer triiodothyronine (0.25 %) doses, the equilibration curves were vastly different. Trijodothyronine [125I] was removed from the circulation at a significantly greater fractional rate than both thyroxine treatments over the entire time period studied. Equilibration of triiodothyronine between plasma and popliteal lymph occurred during the first Ihr collection interval.
100[ 10 o
-
I
I
I
12 6 Time (hr) Fig. 5. The disappearance of radioactively labelled thyroid hormones from plasma and their appearance in popliteal lymph following sequential i.v. injections of tracer [125J]thyroxine, tracer [1251]triiodothyronine and loading dose thyroxine, at 72 hr intervals. Plasma radioactivity following tracer thyroxine (+), loading dose thyroxine (A) and tracer triiodothyronine (@). Popliteal lymph radioactivity after tracer thyroxine (>), loading dose thyroxine (A) and tracer triidothyronine (0). 0
The effect of a loading dose (5 my) of thyroxine during isotopic equilibrium on the distribution of tracer thyroxine and human serum albumin between blood and lymph. During the first 30 min post-injection plasma [125I]_ thyroxine concentration fell at over 100 times the preinjection rate but thereafter the disappearance rate appeared to differ little from the preinjection rate (Fig. 6). Thyroxine disappearance rates in testicular and popliteal lymph followed that of plasma. No comparable changes were observed in the rate of removal of albumin from plasma and lymph. The removal of thyroid hormones from lymph during passage through a lymph node. The recovery of [1251]thyroxine during in situ perfusion of the popliteal node was compared with that of [1311]ovine serum albumin at
133 THYROID HORMONE DISTRIBUTION four concentrations of thyroxine in the lymph. These were compared with the loss of tracer triiodothyronine. The results are summarized in Table 3 where it can be seen that significant portions of thyroxine were lost during passage through the node and these increased with increasing concentra13 -*. _ 10----
08
0A
Al~~~~
A
0.5 'V~~A
05
A
A~~~
0 35
I 90
A
105 Time (hr) Fig. 6. The effect of an intravenous loading dose of thyroxine at isotopic equilibrium (96 hr after tracer dose) on the distribution of [125I]thyroxine and [1311]human serum albumin between plasma and afferent testicular lymph. [125I]thyroxine radioactivity in plasma (A) and testicular lymph (A). [131I]human serum albumin radioactivity in plasma (+) and testicular lymph (0).
95
100
TABLE 3. The effect of changes in the proportion of unbound thyroid hormones on their rate of loss from lymph during perfusion of a popliteal lymph node
Thryoid hormone: Thyroid hormone: Ovine serum Ovine serum albumin activity albumin activity: ratio in the ratio in efferent Proportion unbound infusion E/I lymph
Total* hormone concentration
Unbound* hormone concentration
(/sg/100 ml.)
(ng/I00 ml.)
(%)
(I)
(E)
(%)
1-8 11-8 101-8
2-10 20-98 328-92 5,970 35 0-196
0-117 0-178 0-323 0-596 0-489
0.858±0.001
0-851+0-003 0*967 + 0-004 0-845 + 0-024 0*777+ 0047
99-2
94-6 1-022 + 0-003 0-980 ± 0*013 86-2 0-956 + 0-014 81-3 1,001*8 0 795±0*023 1-110+0-008 0-04 71*6 * Concentrations of hormone in the infusion administered at 1 ml./hr. The mean efferent lymph flow rate was 7 9 ml./hr. Samples containing exogenous thyroxine would be diluted almost sixfold in the node. t Mean + S.D. values for triplicate analysis on 1 ml. infusion. t Mean + S.D. values for half hourly collection of efferent lymph during the 4 hr infusion period.
134 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND tions of thyroxine. Triiodothyronine loss at tracer concentrations was greater than that at the highest concentration of thyroxine. Because of the large amounts of thyroxine removed at high thyroxine concentrations and since the labelled material removed from the lymph was not recovered during the period of saline perfusion between treatments, these results are unlikely to be explained by tissue uptake in the node. They more likely indicate transfer of hormone into the blood stream. DISCUSSION
The experiments described supply convincing evidence that labelled thyroid hormones are removed from the circulation and appear in interstitial fluid at a great fractional rate than labelled human serum albumin. Such a result could be explained by the preferential transcapillary movement of thyroid hormones bound to proteins other than albumin or by the greater relative rate of movement of unbound hormone. Since human serum albumin, sheep albumin, TBG and TBP-2 all had similar elution profiles on gel filtration and because the measured lymph/plasma concentration ratios for the three sheep proteins did not differ significantly from one another, the former hypothesis was discarded. In experiments where the proportion of unbound thyroxine was elevated, equilibration of labelled hormone between plasma and lymph occurred more rapidly, supporting the thesis that thyroxine exchanges between these pools as unbound thyroxine. Following equilibration in the extracellular fluid, the proportion of labelled thyroxine in plasma and lymph relative to each other was no different from that with tracer doses. This would be expected from consideration of the thyroxine binding properties of plasma and lymph as plotted in Fig. 3. However, the proportion of total labelled thyroxine present in these pools was reduced. Since the fractional rate of elimination of thyroxine from the body was not increased, lower concentrations in the extracellular fluid most likely indicate higher concentrations within cells. Thyroxine is known to be bound intracellularly (Oppenheimer, Surks & Schwartz, 1969) but the relationship between intracellular bound thyroxine and unbound thyroxine must be different from that for plasma and lymph and would seem to favour cellular uptake when thyroxine concentrations are elevated. Experiments in which thyroxine loading doses were administered at isotopic equilibrium support these arguments. Triiodothyronine equilibrated more rapidly between plasma and lymph than did thyroxine due most likely to the higher proportion of triiodothyronine which is unbound. The equilibration curves for labelled triiodothyronine were, however, quite different from those for thyroxine, even
135 THYROID HORMONE DISTRIBUTION when a loading dose of thyroxine was used. This may represent a higher rate of dissociation of triiodothyronine from its plasma binding proteins, different fractional rates of destruction, or different steady-state distribution between intracellular and extracellular pools. The concentration of unbound thyroxine was similar in plasma and any sample of lymph from the same animal. Such a situation would arise whether thyroxine and its binding proteins moved independently, or if thyroxine crossed capillary membrane bound only to proteins. In the former case an unmeasurable transcapillary concentration gradient would be expected for a substance with molecules as small as those of thyroxine even if there was considerable transcapillary transport of that substance (Landis & Pappenheimer, 1963). If lymph is produced as a simple filtrate of plasma with the proteins being reduced in concentration, then thyroxine would dissociate from its binding proteins in the interstitial fluid so that the concentration of unbound thyroxine would be very similar to that in the blood (Oppenheimer & Surks, 1964). In either case the concentration of total thyroxine in blood or interstitial fluid would be determined by the concentration of the thyroxine-binding proteins. Such a transcapillary gradient for total thyroxine has previously been suggested as providing evidence that thyroxine leaves the circulation bound to its binding proteins (Ismail, El-Ridi, Badran, Khalifa, Abel-Hay & Talaat, 1967) but by itself such an observation is insufficient for that purpose. The expeiiments reported in this paper support earlier claims that unbound thyioxine is the freely diffusible form of thyroxine (Ingbar & Freinkel, 1960; Robbins & Rall, 1967) and that tissue uptake of thyroxine is related to the proportion of thyroxine in the unbound state (Hillier, 1969, 1971). These results do not support the claim that thyroxine is distributed in the body bound to its binding proteins (Oppenheimer et al. 1969). It should be pointed out that because of movement of thyroxine in both directions across the capillaries, and possible cellular uptake of thyroxine, the relative rates of movement of thyroxine and protein from plasma to lymph will tend to underestimate the true rate of transcapillary movement of thyroxine relative to that of its binding proteins. Even though most thyroxine seemed to cross capillaries in most tissues as unbound thyroxine, in all tissues a proportion of the hormone which crossed the capillaries must have done so bound to protein. Those differences which were observed in the transcapillary exchange of both thyroxine and protein between various tissues can be explained in terms of known differences in capillary permeability of these tissues (Yoffey & Courtice, 1970). Certainly no evidence was found to suggest that plasma thyroxine-binding proteins had a functional role in the facilitated transport of thyroid hormones into any tissue.
136
M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND
REFERENCES AmDREws, P. (1964). Estimation of the molecular weight of proteins by Sephadex gel filtration, Bsochem. J. 91, 222-233. HAm , J. G. & Mourns, B. (1962). The output of cells in lymph from the popliteal node of sheep. Q. Jl. exp. Phy8iol. 47, 360369. HEi-MxIAMP, R. W., GOODLAND, R. L., BALE, W. F., SPAR, I. L. & MuTscErzR, L. E. (1960). High specific activity iodination of y-globulin with iodine-131 monochloride. Cancer Res. 20, 1495-1500. HTTILmi, A. P. (1969). The release of thyroxine from serum protein in the vessels of the liver. J. Phyaiol. 203, 419-434.. HTTuTTEr, A. P. (1971). The mechanism of thyroxine transfer from plasma to tissue binding sites. J. Physiol. 217, 635-639. INGBAR, S. H. & FREINKEL, N. (1960). Regulation ofthe peripheral metabolism of the thyroid hormones. Recent Prog. Hor-m. Bem. 16, 353-403. IRvINE, C. H. G. & SIMPSON-MORGAN, M. W. (1974). Relative rates of transcapillary movement of free thyroxine, protein-bound thyroxine, thyroxine-binding proteins, and albumin. J. dlin. Invest. 54, 156-164. ISMATL, A. A., EL-RIm, M. S., BADRAN, I., KHATLIA, K., ABDELEHAY, A. R. & TAT, M. (1967). Extravascular circulation of thyroid hormones. Am. J. Physiol. 213, 1391-1396. LANDis, E. M. & PAPPENHEMR, J. R. (1963). Exchange of substances through the capillary walls. In Handbook of Physiology, section 2, vol. 2, ed. HAMILTON, W. F., pp. 961-1034. Washington: American Physiological Society. LAscEIa s, A. K. & MouRns, B. (1961). Surgical techniques for the collection of lymph from unanaesthetized sheep. Q. Jl exp. Physiol. 46, 199-205. MoRms, B. & McINTosH, G. H. (1971). Techniques for the collection of lymph with special reference to the testis and ovary. Acta endocr., Copenh. suppl. 158, 145-168. NICOLOFF, J. T. & DowIaWG, J. T. (1968). Studies of peripheral thyroxine distribution in thyrotoxicosis and hypothyroidism. J. dlin. Invest. 47, 2000-2015. OPPENHEIMER, J. H. & Suoxs, M. I. (1964). Determination of free thyroxine in human serum: A theoretical and experimental analysis. J. dlin. Endoer. Metab. 24, 785-793. OPPENHEIMR, J. H., SUTUKS, M. I. & SCHWARTZ, H. I. (1969). The metabolic significance of exchangeable cellular thyroxine. Recent Prog. Horm. Res. 25, 381-422. ROBBINS, J. & RALL, J. E. (1967). The iodine-containing hormones. In Hormones in Blood, 2nd edn., ed. GRAY, C. H. & BACHARACH, A. L.,.pp. 383-490. New York: Academic Press. SuTHERIAwD, R. L., BRANDON, M. R. & SmaPsON-MORGAN, M. W. (1975). Effect of ionic strength and ionic composition of assay buffers on the interaction of thyroxine with plasma proteins. J. Endocr. 66, 319-327. SUHErLAND, R. L. & IRvINE, C. H. G. (1973). Total plasma thyroxine concentrations in horses, pigs, cattle and sheep: Anion exchange resin chromatography and ceric-arsenite colorimetry. Am. J. vet. Res. 34, 1261-1266. SUTHERLAND, R. L. & SDIMPSON-MORGAN, M. W. (1975). The thyroxine-binding properties of serum proteins. A competitive binding technique employing Sephadex G-25. J. Endocr. 65, 319-332. Yorimy, J. M. & COuTIcE, F. C. (1970). L-ymphatic8, Lymph and the, Lymphomyekoid Complex. London: Academic Press.
J. Phygiol. (1976), 257, pp. 123-136 With 6 text-figures Printed in Great Britain
THE TRANSCAPILLARY EXCHANGE OF THYROID HORMONES AND THYROXINE-BINDING PROTEINS BETWEEN BLOOD AND TISSUE FLUIDS
By M. W. SIMPSON-MORGANt AND R. L. SUTHERLAND* From the Department of Experimental Pathology, the John Curtin School of Medical Research, the Australian National University, Canberra, Australia
(Received 14 August 1975) SUMMARY
1. A study has been made of the relative importance of protein bound and unbound hormone in the exchange of thyroid hormones between blood and interstitial fluid. 2. When [l25J]thyroxine (or triiodothyronine) and [1311]human serum albumin were injected simultaneously into the circulation of sheep with chronic lymphatic fistulae, the thyroid hormones were removed from the circulation and appeared in all lymph samples at a greater fractional rate than human serum albumin. 3. The steady-state lymph/plasma concentration ratios of the two specific thyroxine binding proteins were similar to each other and to those of albumin and total thyroxine. 4. Gel filtration studies indicated that the two specific thyroxine binding proteins, ovine serum albumin and human serum albumin, were all of similar molecular size. 5. Concentrations of unbound thyroxine in plasma and various samples of lymph from the one animal were similar. 6. Increasing the proportion of thyroid hormone that was unbound resulted in an increased rate of equilibration of labelled hormone between blood plasma and lymph. 7. Perfusion of the popliteal lymph node demonstrated that thyroid hormones were removed from lymph during its passage through the node. The amount removed was related to the proportion of hormone in the unbound state. * Present address: Department de Chimie Biologique, Facult6 de M6decine de Bicetre, 94270 Bicetre, France. t Present address: Department of Animal Husbandry, Veterinary School, University of Queensland, St Lucia, Brisbane 4067, Australia.
124 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND 8. It was concluded that thyroid hormones cross capillary endothelium mainly as the unbound molecule and that such movement is bidirectional. INTRODUCTION
In the plasma of sheep the majority of circulating thyroxine is bound tightly but reversibly by three thyroxine-binding proteins. The thyroid hormone binding properties of these proteins are similar to those of human thyroxine-binding globulin (TBG), thyroxine-binding prealbumin (TBPA) and serum albumin respectively. As a result of protein binding only 006 % of the total plasma thryoxine circulates as the unbound molecule (Sutherland & Simpson-Morgan, 1975). The physiological role of plasma thyroxine-binding proteins in the transport and distribution of thyroid hormones is not yet fully understood. Before entering the tissue fluids, and ultimately the tissue cells, thyroid hormones must first cross the endothelium of blood capillaries. Since binding to specific plasma proteins is likely to impede such movement, it was considered important to examine quantitatively the effect of protein binding on the rate of transfer of thyroid hormones from blood to interstitial fluid. This was attained by measuring the relative rates of appearance of radioactively labelled thyroid hormones and serum albumin in the lymph of chronically cannulated sheep, following the simultaneous i.v. injection of the two labelled compounds. The work presented in this communication is an extension of that previously reported by other workers from this laboratory (Irvine & Simpson-Morgan, 1974). METHODS
Experimental animals. Adult male and female Merino and Merino cross sheep weighing 35-55 kg were housed indoors in individual pens or metabolism cages and received lucerne chaff, oats and water ad libitum. Surgical procedures. Chronic lymphatic fistulae were established in the afferent and efferent vessels of the popliteal lymph node, the afferent testicular lymphatics, the lumbar lymphatic trunk, the intestinal lymphatic trunk, the efferent hepatic lymph duct, the deep cervical lymph duct and the efferent lymphatics from the prefemoral and prescapular lymph nodes by surgical procedures which have been described (Lascelles & Morris, 1961; Hall & Morris, 1962; Morris & McIntosh, 1971). Lymph was collected into plastic bottles or plastic centrifuge tubes tied to a plastic holder sutured to the animal's skin. Powdered heparin was used as an anticoagulant. Lymph was centrifuged for 10 min at 2000 rev/min and the cell free lymph separated, and frozen at - 200 C until required. Samples of afferent hepatic and afferent renal lymph were kindly donated by Dr G. G. MacPherson. Blood was collected through chronic indwelling jugular vein cannulae. Both serum and plasma were prepared. Materials. ['251]L-thyroxine and [125I]L-3, 3',5-triiodothyronine in 50°% propy-
THYROID HORMONE DISTRIBUTION
125
lene glycol, with an initial specific activity of 40-45 mc/mg, containing 4-5 4ug hormone/ml. were supplied at 2-mnonthly intervals by the Radiochemical Centre, Amersham, England. [131I]sodium iodide (Atomic Energy Commission, Lucas Heights, New South Wales, Australia) was used to label purified ovine and human serum albumin by the method of Helmkamp, Goodland, Bale, Spar & Mutschler (1960). Human serum albumin, 10000/ pure on electrophoresis was supplied by Hoechst Australia Ltd, Sydney, Australia. Sheep fraction V (Commonwealth Serum Laboratories, Parkville, Victoria, Australia) was purified as described by Sutherland, Brandon & Simpson-Morgan (1975). L-thyroxine sodium was supplied by B.D.H. Chemicals Ltd, Poole, England. Transcapillary transfer of thyroxine-binding proteins and thyroid hormones in sheep. Lymphatic cannulae were allowed to flow for 24 hr before an experiment by which time the animals had completely recovered from surgery and were eating and drinking normally. A 3 ml. volume of the animal's own plasma was mixed with approximately 15 Arc [125I]thyroxine or triiodothyronine and 10 /sc ['311]human serum albumin. This mixture was shaken and dialysed overnight at 2° C against a slurry of Amerlite IRA-400 resin as described by Nicoloff & Dowling (1968). Irvine & Simpson-Morgan (1974) demonstrated that after this type of dialysis the tracer contains no iodide, 1-5-2-0 0 of the [1251] as triiodothyronine and the remainder as thyroxine. [125I]triiodothyronine was essentially pure following similar treatment. The radioactive dose was weighed accurately into a sterile syringe and infused into the jugular vein. The cannula was rinsed several times with blood and sterile heparinized saline. Dose time and dose weight were recorded and a portion of the injected material was retained for radio-assay. Blood and lymph samples were collected at timed intervals after injection and plasma separated. Lymph flow rates were recorded. Where volumes permitted 3-0 ml. plasma or lymph was counted in a gamma scintillation spectrometer system after protein precipitation with trichloroacetic acid. Results were expressed as the percentage of the dose remaining/l. plasma or lymph at any fixed time after injection. In some experiments a large dose of thyroxine (5 mg) was injected 72-98 hr after the tracer injection. At this time the thyroxine and human serum albumin had attained isotopic equilibrium as judged by parallel isotope disappearance slopes in blood and lymph. Samples of blood and lymph were collected at short time intervals after injection. In experiments where [125I]triiodothyronine replaced labelled thyroxine, radioactively labelled contaminants were removed by anion exchange chromatography (Sutherland & Irvine, 1973). Perfusion of popliteal lymph nodes. An afferent popliteal lymphatic accompanying the recurrent tarsal vein was cannulated in the direction of lymph flow. Sterile Evan's Blue dye (0-2 ml.) was slowly infused up the afferent lymphatic and all efferent lymphatics were identified, tied off, and one cannulated. A 24 hr collection of sterile lymph was made after clearance of the dye. Cells were spun off and the lymph plasma retained. An infusion system was set up whereby the sterile efferent lymph labelled with [1251]thyroxine or [125I]triiodothyronine (0-2 flc/100 ml.) and [131I]ovine serum albumin (0-2 ,uc/O0 ml.), was infused into the afferent lymphatic at the rate of 1 ml./hr. Varying concentrations of unlabelled thyroxine (0-1 0 g/ml.) were added to the infusion mixture. Each infusion was continued for 4 hr with a 1 hr sterile saline wash between treatments. Efferent lymph was collected over half hourly intervals and the results were expressed as the thyroid hormone: albumin activity ratio in efferent lymph/the thyroid hormone: albumin activity ratio in the infusion mixture.
126
M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND
Analytical procedures. The concentrations of the specific thyroxine-binding proteins, albumin and total thyroxine in blood plasma and lymph were measured by techniques described previously (Sutherland & Simpson-Morgan, 1975). Radioactivity measurements in samples containing both 125I- and '31I-labelled compounds were made in a two-channel Packard Gamma Scintillation Spectrometer, Model 578. Windows were chosen such that 131J counted independent of 125I uhile a constant proportion of 131I (16 %) counted in the 1251 channel. Suitable volumes were used such that the efficiency of counting did not change between samples. Normal and heat treated (1 hr at 600 C) sheep serum samples, labelled with [125I]thyroxine and [131I]ovine serum albumin, were fractionated on 2 0 x 90 cm Sephadex G-200 columns according to the method of Andrews (1964). The flow rate was maintained at 10*0 ml./hr by a peristaltic pump. RESULTS
The rate of movement of [12-I]thyroxine ard [1311]human serum albumin from blood to lymph. Following simultaneous i.v. injection of labelled thyroxine and human serum albumin, tracer thyroxine was removed from the blood plasma and appeared in lymph at a greater fractional rate than did labelled albumin (Fig. 1). The rates of transfer of labelled thyroxine and albumin from blood to lymph during a given collection period were estimated from the ratios of the activities in the lymph, to the activities in plasma at the mid-point of that collection period. These ratios which are expressed as percentages in Table 1, varied considerably when lymph was collected from different lymphatics in the same animal, or from the same lymphatic in different animals. However, in any lymphatic the net relative rate of movement of thyroxine from blood to lymph was greater than that of human serum albumin during the first, 1 hr collection period. This difference was relatively constant for any particular lymph pool but differed significantly between different tissue fluid pools due possibly to differences in capillary permeability. The lymph:plasma concentration ratiosfor total thyroxine and the thyroxinebinding proteins. Plasma and lymph samples from twenty-four animals, representing ten different tissue fluid pools were analysed for total thyroxine, albumin and thyroxine-binding protein concentrations. The results are summarized in Table 2 where the lymph plasma concentration ratios of those substances are given. In all samples studied three thyroxinebinding proteins were identified and their association constants corresponded to those of thyroxine-binding globulin (TBG), thyroxine-binding protein-2 (TBP-2) and albumin of sheep serum (Sutherland & SimpsonMorgan, 1975) In the cases of afferent testicular and efferent popliteal lymph, which were studied in a number of animals, there were no significant differences between the mean lymph: plasma concentration ratios of total thyroxine, albumin and the two specific thyroxine-binding proteins. This indicated the absence of a preferential transcapillary transfer
127 THYROID HORMONE DISTRIBUTION mechanism for specific thyroxine-binding proteins. The data from the other eight tissue fluid pools were consistent with such an hypothesis. Gel filtration studies on sheep thyroxine-binding proteins. Since the measured lymph:plasma concentration ratios for the three binding pro100
50 0I 0
10
100 50
°0 10
12 Time (hr)
Fig. 1. The disappearance of [1251]thyroxine (A) and [13Ijhuman serum albumin (+) from plasma and their subsequent appearance in afferent testicular lymph (upper diagram) and intestinal lymph (lower diagram), of two different animals, following simultaneous intravenous injection of the two isotopes. [1251]thyroxine radioactivity in testicular (U) and intestinal lymph ([]). [1311]human serum albumin radioactivity in testicular (*)and intestinal lymph (0).
teins indicated that these proteins were of similar molecular size attempts were made to confirm this using gel filtration. The protein and radioactivity elution profiles for typical separations of normal and heat treated sheep serum are shown in Fig. 2. In both cases the optical density and 5
PHY 257
128 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND radioactivity peaks for albumin coincided. The [125I]thyroxine peak occurred one tube earlier than albumin in the case of normal serum and one tube later for heat treated serum. In normal sheep serum about 55 % of endogenous thyroxine is bound to TBG. After heat treatment (1 hr at 60° C) binding to TBG is completely destroyed and 70 % of tracer thyroxine TABLE 1. The relative rates of movement of [125I]thyroxine and [131I]humpn serum albumin from blood to lymph during the first j hr after injection of the labelled dose Lymph:plasma activity ratios (%) Lymph sample Lumbar trunk Afferent testicular Efferent prescapular Efferent popliteal Intestinal trunk Efferent prefemoral
8-18
albumin 6*61
Relative rate of transfer.t rIl'yroxine Human serum albumin 1-25
4
7.92 + 6.32*
6-33 ± 5*74
1-31+ 0*13
1
1*38
0-416
3*32
4
2*73 ± 1*92
0*768 + 0*389
3-42 ± 1*85
1
4-81
0-91
5-29
1
8-35
1-56
5.35
No. of animals 1
[1311]
[121] Thyroxine
Hluman serum
Mean+s.D. t A ratio greater than 1 indicates more rapid transcapillary exchange of thyroxine *
relative to human serum albumin.
TABLE 2. Lymph: plasma concentration ratios for total thyroxine, the two specific thyroxinebinding proteins (TBG and TBP-2) and serum albumin in samples of lymph from several tissue fluid pools of sheep Total No. of TBG Lymph sample samples thyroxine TBP-2 Albumin 1 Afferent hepatic 0*780 0*909 0-979 0*820 1 0-667 0-858 Efferent hepatic 0*858 0-828 Afferent testicular 9 0.757 + 0.187* 0*782 ± 0*133 0-792 + 0*163 0*841 ± 0-141 1 0-732 0-508 Efferent prescapular 0-781 0-774 1 0*557 0*663 Efferent prefemoral 0*683 0*773 1 0*598 Cervical lymph duct 0-775 0*686 0-558 1 0*400 Lumbar lymph trunk 0-662 0-635 0X674 1 0*418 Afferent renal 0-390 0*572 0-549 8 0*503 + 0*188 0*418 ± 0*160 0*482 ± 0-226 0*473 ± 0*167 Efferent popliteal 1 0*250 0*487 Intestinal lymph 0*425 0*415 trunk *
Mean+ s.D.
129 THYROID HORMONE DISTRIBUTION is associated with TBP-2. Thus the similar elation volumes for ovine serum albumin and the two different [125I]thyroxine peaks indicated that the three thyroxine-binding proteins of sheep serum were of similar molecular size. Human serum albumin and ovine serum albumin were shown to have similar elution patterns from Sephadex G-200 columns. 10,000
A
AA~~~~~~4 ~0
A
C
5000
V
A
10
30
20 A&
O0 10.000
20 1e ~ EA
JA
a~.A
1a
40
100
t
7
di
i
0
C
EA 30
I
30 20 Fraction number
40
L\
100
40
Fig. 2. Protein (A) and radioactivity elution profiles for Sephadex G-200 separations of normal (upper diagram) and heat treated (lower diagram) sheep serum labelled with tracer [L25I]thyroxine (@) and [L31I]ovine serum albumin (A\). Heat treated serum was exposed to 600 C for 1 hr.
Unbound thyroxine concentrations in plasma and lymph. Although the total thyroxine concentration varied appreciably between plasma and various samples of lymph from the one animal, the concentration of unbound thyroxine was relatively constant from sample to sample. Hence the proportion of total thyroxine that was unbound was inversely proportional to the total thyroxine concentration and varied from 0056 + 0019 % (mean + S.D.) in plasma to 0 125 + 0066 % in popliteal lymph. The thyroxine binding curves for various samples of lymph from the one animal were essentially parallel to the plasma curve, indicating that the 5-2
130 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND thyroxine-binding proteins of lymph were a simple physiological dilution of those in plasma (Fig. 3). The rate of movement of [125I]triiodothyronine and LulI]human serum albumin from blood to lymph. Triiodothyronine is less firmly bound to sheep plasma proteins than is thyroxine. The concentration of unbound triiodothyronine in normal sheep is approximately 0-2 ng/100 ml. or 0-25 % of the total (R. .L. Sutherland, unpublished). This is 4-5 times greater than the proportion of thyroxine which is unbound. Following administration of labelled triiodothyronine this hormone waa removed from the circulation and equilibrated with lymph much more rapidly than thyroxine in the preceding experiments (Fig. 4). 2500
-
E~~~ A
a'1000 1000
s
-
C
0
10
1
111li
2
5
10 100 1000 2000 Unbound T4 (ng/100 ml.)
Fig. 3. Thyroxine binding curves (log protein-bound concentration v8. log unbound concentration) for diluted samples (1:150) of blood plasma (A), afferent testicular lymph (*) and efferent popliteal lymph (,A) collected from the same animal.
The effect of elevated unbound thyroid hormone concentrations on the rate of movement of labelled hormone from blood to lymph. A ewe with an efferent popliteal lymphatic fistula was injected sequentially with tracer [1251]thyroxine, tracer [125I]triiodothyronine, and a loading dose (tracer + 5 mg) of thyroxine, at 72 hr intervals. The latter treatment was estimated to raise the total plasma thyroxine concentration to approximately 300 ,ug/ 100 ml. and the percentage unbound thyroxine to 0-23 %. The isotopic equilibration curves for the 12 hr following injection of each dose are shown in Fig. 5. The loading dose of thyroxine, when compared with the tracer thyroxine
131 THYROID HORMONE DISTRIBUTION dose, resulted in the more rapid movement of [1251]thyroxine out of the circulation during the first hour. This was accompanied by the more rapid appearance of [126I]thyroxine in popliteal lymph during the first hour following the loading dose. Subsequently [1251]thyroxine was removed from the circulation at the same fractional rate with both treatments. 1oo
0
0A
10
0o
1*0
04 24 Time (hr)
100
10 I
0
10
0o1
0
12
24
Time (hr)
Fig. 4. The disappearance of [125I]triiodothyronine (A) and ['31I]human serum albumin (+) from plasma and their subsequent appearance in afferent testicular lymph (upper diagram) and efferent popliteal lymph (lower diagram) following the simultaneous i.v. injection of the two isotopes into one animal. [125I]triidothyronine radioactivity in testicular (A) and popliteal lymph (>). [131I]human serum albumin radioactivity in testicular (@) and popliteal lymph (0).
132 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND Equilibration between plasma and lymph occurred earlier following the loading dose but following equilibration the lymph:plasma ratios for radioactive thyroxine were the same in both cases. Despite the fact that almost identical proportions of hormone in the circulation were unbound following the loading thyroxine (0.23 %) and tracer triiodothyronine (0.25 %) doses, the equilibration curves were vastly different. Trijodothyronine [125I] was removed from the circulation at a significantly greater fractional rate than both thyroxine treatments over the entire time period studied. Equilibration of triiodothyronine between plasma and popliteal lymph occurred during the first Ihr collection interval.
100[ 10 o
-
I
I
I
12 6 Time (hr) Fig. 5. The disappearance of radioactively labelled thyroid hormones from plasma and their appearance in popliteal lymph following sequential i.v. injections of tracer [125J]thyroxine, tracer [1251]triiodothyronine and loading dose thyroxine, at 72 hr intervals. Plasma radioactivity following tracer thyroxine (+), loading dose thyroxine (A) and tracer triiodothyronine (@). Popliteal lymph radioactivity after tracer thyroxine (>), loading dose thyroxine (A) and tracer triidothyronine (0). 0
The effect of a loading dose (5 my) of thyroxine during isotopic equilibrium on the distribution of tracer thyroxine and human serum albumin between blood and lymph. During the first 30 min post-injection plasma [125I]_ thyroxine concentration fell at over 100 times the preinjection rate but thereafter the disappearance rate appeared to differ little from the preinjection rate (Fig. 6). Thyroxine disappearance rates in testicular and popliteal lymph followed that of plasma. No comparable changes were observed in the rate of removal of albumin from plasma and lymph. The removal of thyroid hormones from lymph during passage through a lymph node. The recovery of [1251]thyroxine during in situ perfusion of the popliteal node was compared with that of [1311]ovine serum albumin at
133 THYROID HORMONE DISTRIBUTION four concentrations of thyroxine in the lymph. These were compared with the loss of tracer triiodothyronine. The results are summarized in Table 3 where it can be seen that significant portions of thyroxine were lost during passage through the node and these increased with increasing concentra13 -*. _ 10----
08
0A
Al~~~~
A
0.5 'V~~A
05
A
A~~~
0 35
I 90
A
105 Time (hr) Fig. 6. The effect of an intravenous loading dose of thyroxine at isotopic equilibrium (96 hr after tracer dose) on the distribution of [125I]thyroxine and [1311]human serum albumin between plasma and afferent testicular lymph. [125I]thyroxine radioactivity in plasma (A) and testicular lymph (A). [131I]human serum albumin radioactivity in plasma (+) and testicular lymph (0).
95
100
TABLE 3. The effect of changes in the proportion of unbound thyroid hormones on their rate of loss from lymph during perfusion of a popliteal lymph node
Thryoid hormone: Thyroid hormone: Ovine serum Ovine serum albumin activity albumin activity: ratio in the ratio in efferent Proportion unbound infusion E/I lymph
Total* hormone concentration
Unbound* hormone concentration
(/sg/100 ml.)
(ng/I00 ml.)
(%)
(I)
(E)
(%)
1-8 11-8 101-8
2-10 20-98 328-92 5,970 35 0-196
0-117 0-178 0-323 0-596 0-489
0.858±0.001
0-851+0-003 0*967 + 0-004 0-845 + 0-024 0*777+ 0047
99-2
94-6 1-022 + 0-003 0-980 ± 0*013 86-2 0-956 + 0-014 81-3 1,001*8 0 795±0*023 1-110+0-008 0-04 71*6 * Concentrations of hormone in the infusion administered at 1 ml./hr. The mean efferent lymph flow rate was 7 9 ml./hr. Samples containing exogenous thyroxine would be diluted almost sixfold in the node. t Mean + S.D. values for triplicate analysis on 1 ml. infusion. t Mean + S.D. values for half hourly collection of efferent lymph during the 4 hr infusion period.
134 M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND tions of thyroxine. Triiodothyronine loss at tracer concentrations was greater than that at the highest concentration of thyroxine. Because of the large amounts of thyroxine removed at high thyroxine concentrations and since the labelled material removed from the lymph was not recovered during the period of saline perfusion between treatments, these results are unlikely to be explained by tissue uptake in the node. They more likely indicate transfer of hormone into the blood stream. DISCUSSION
The experiments described supply convincing evidence that labelled thyroid hormones are removed from the circulation and appear in interstitial fluid at a great fractional rate than labelled human serum albumin. Such a result could be explained by the preferential transcapillary movement of thyroid hormones bound to proteins other than albumin or by the greater relative rate of movement of unbound hormone. Since human serum albumin, sheep albumin, TBG and TBP-2 all had similar elution profiles on gel filtration and because the measured lymph/plasma concentration ratios for the three sheep proteins did not differ significantly from one another, the former hypothesis was discarded. In experiments where the proportion of unbound thyroxine was elevated, equilibration of labelled hormone between plasma and lymph occurred more rapidly, supporting the thesis that thyroxine exchanges between these pools as unbound thyroxine. Following equilibration in the extracellular fluid, the proportion of labelled thyroxine in plasma and lymph relative to each other was no different from that with tracer doses. This would be expected from consideration of the thyroxine binding properties of plasma and lymph as plotted in Fig. 3. However, the proportion of total labelled thyroxine present in these pools was reduced. Since the fractional rate of elimination of thyroxine from the body was not increased, lower concentrations in the extracellular fluid most likely indicate higher concentrations within cells. Thyroxine is known to be bound intracellularly (Oppenheimer, Surks & Schwartz, 1969) but the relationship between intracellular bound thyroxine and unbound thyroxine must be different from that for plasma and lymph and would seem to favour cellular uptake when thyroxine concentrations are elevated. Experiments in which thyroxine loading doses were administered at isotopic equilibrium support these arguments. Triiodothyronine equilibrated more rapidly between plasma and lymph than did thyroxine due most likely to the higher proportion of triiodothyronine which is unbound. The equilibration curves for labelled triiodothyronine were, however, quite different from those for thyroxine, even
135 THYROID HORMONE DISTRIBUTION when a loading dose of thyroxine was used. This may represent a higher rate of dissociation of triiodothyronine from its plasma binding proteins, different fractional rates of destruction, or different steady-state distribution between intracellular and extracellular pools. The concentration of unbound thyroxine was similar in plasma and any sample of lymph from the same animal. Such a situation would arise whether thyroxine and its binding proteins moved independently, or if thyroxine crossed capillary membrane bound only to proteins. In the former case an unmeasurable transcapillary concentration gradient would be expected for a substance with molecules as small as those of thyroxine even if there was considerable transcapillary transport of that substance (Landis & Pappenheimer, 1963). If lymph is produced as a simple filtrate of plasma with the proteins being reduced in concentration, then thyroxine would dissociate from its binding proteins in the interstitial fluid so that the concentration of unbound thyroxine would be very similar to that in the blood (Oppenheimer & Surks, 1964). In either case the concentration of total thyroxine in blood or interstitial fluid would be determined by the concentration of the thyroxine-binding proteins. Such a transcapillary gradient for total thyroxine has previously been suggested as providing evidence that thyroxine leaves the circulation bound to its binding proteins (Ismail, El-Ridi, Badran, Khalifa, Abel-Hay & Talaat, 1967) but by itself such an observation is insufficient for that purpose. The expeiiments reported in this paper support earlier claims that unbound thyioxine is the freely diffusible form of thyroxine (Ingbar & Freinkel, 1960; Robbins & Rall, 1967) and that tissue uptake of thyroxine is related to the proportion of thyroxine in the unbound state (Hillier, 1969, 1971). These results do not support the claim that thyroxine is distributed in the body bound to its binding proteins (Oppenheimer et al. 1969). It should be pointed out that because of movement of thyroxine in both directions across the capillaries, and possible cellular uptake of thyroxine, the relative rates of movement of thyroxine and protein from plasma to lymph will tend to underestimate the true rate of transcapillary movement of thyroxine relative to that of its binding proteins. Even though most thyroxine seemed to cross capillaries in most tissues as unbound thyroxine, in all tissues a proportion of the hormone which crossed the capillaries must have done so bound to protein. Those differences which were observed in the transcapillary exchange of both thyroxine and protein between various tissues can be explained in terms of known differences in capillary permeability of these tissues (Yoffey & Courtice, 1970). Certainly no evidence was found to suggest that plasma thyroxine-binding proteins had a functional role in the facilitated transport of thyroid hormones into any tissue.
136
M. W. SIMPSON-MORGAN AND R. L. SUTHERLAND
REFERENCES AmDREws, P. (1964). Estimation of the molecular weight of proteins by Sephadex gel filtration, Bsochem. J. 91, 222-233. HAm , J. G. & Mourns, B. (1962). The output of cells in lymph from the popliteal node of sheep. Q. Jl. exp. Phy8iol. 47, 360369. HEi-MxIAMP, R. W., GOODLAND, R. L., BALE, W. F., SPAR, I. L. & MuTscErzR, L. E. (1960). High specific activity iodination of y-globulin with iodine-131 monochloride. Cancer Res. 20, 1495-1500. HTTILmi, A. P. (1969). The release of thyroxine from serum protein in the vessels of the liver. J. Phyaiol. 203, 419-434.. HTTuTTEr, A. P. (1971). The mechanism of thyroxine transfer from plasma to tissue binding sites. J. Physiol. 217, 635-639. INGBAR, S. H. & FREINKEL, N. (1960). Regulation ofthe peripheral metabolism of the thyroid hormones. Recent Prog. Hor-m. Bem. 16, 353-403. IRvINE, C. H. G. & SIMPSON-MORGAN, M. W. (1974). Relative rates of transcapillary movement of free thyroxine, protein-bound thyroxine, thyroxine-binding proteins, and albumin. J. dlin. Invest. 54, 156-164. ISMATL, A. A., EL-RIm, M. S., BADRAN, I., KHATLIA, K., ABDELEHAY, A. R. & TAT, M. (1967). Extravascular circulation of thyroid hormones. Am. J. Physiol. 213, 1391-1396. LANDis, E. M. & PAPPENHEMR, J. R. (1963). Exchange of substances through the capillary walls. In Handbook of Physiology, section 2, vol. 2, ed. HAMILTON, W. F., pp. 961-1034. Washington: American Physiological Society. LAscEIa s, A. K. & MouRns, B. (1961). Surgical techniques for the collection of lymph from unanaesthetized sheep. Q. Jl exp. Physiol. 46, 199-205. MoRms, B. & McINTosH, G. H. (1971). Techniques for the collection of lymph with special reference to the testis and ovary. Acta endocr., Copenh. suppl. 158, 145-168. NICOLOFF, J. T. & DowIaWG, J. T. (1968). Studies of peripheral thyroxine distribution in thyrotoxicosis and hypothyroidism. J. dlin. Invest. 47, 2000-2015. OPPENHEIMER, J. H. & Suoxs, M. I. (1964). Determination of free thyroxine in human serum: A theoretical and experimental analysis. J. dlin. Endoer. Metab. 24, 785-793. OPPENHEIMR, J. H., SUTUKS, M. I. & SCHWARTZ, H. I. (1969). The metabolic significance of exchangeable cellular thyroxine. Recent Prog. Horm. Res. 25, 381-422. ROBBINS, J. & RALL, J. E. (1967). The iodine-containing hormones. In Hormones in Blood, 2nd edn., ed. GRAY, C. H. & BACHARACH, A. L.,.pp. 383-490. New York: Academic Press. SuTHERIAwD, R. L., BRANDON, M. R. & SmaPsON-MORGAN, M. W. (1975). Effect of ionic strength and ionic composition of assay buffers on the interaction of thyroxine with plasma proteins. J. Endocr. 66, 319-327. SUHErLAND, R. L. & IRvINE, C. H. G. (1973). Total plasma thyroxine concentrations in horses, pigs, cattle and sheep: Anion exchange resin chromatography and ceric-arsenite colorimetry. Am. J. vet. Res. 34, 1261-1266. SUTHERLAND, R. L. & SDIMPSON-MORGAN, M. W. (1975). The thyroxine-binding properties of serum proteins. A competitive binding technique employing Sephadex G-25. J. Endocr. 65, 319-332. Yorimy, J. M. & COuTIcE, F. C. (1970). L-ymphatic8, Lymph and the, Lymphomyekoid Complex. London: Academic Press.
