A possible variant of thyroxine-binding globulin in Australian Aborigines

Clinicu Chimicu Acra, 116 (1981) 361-367 Elsevier/North-Holland Biomedical Press

CCA

361

1930

A possible variant of thyroxine-binding globulin in Australian Aborigines M. Dick Department

and F. Watson

of Clinical Biochemistry, Queen Elizubeth II Medico1 Centre, Perth, Western Australiu 6009 (Austruliu) (Received

April I I th, 198 I)

Summary We have recently described a major variation from the normal levels of serum thyroxine-binding globulin in many Australian Aborigines [ 11. Subsequently we presented evidence that affected individuals were widely distributed throughout Australia, that “low” values of thyroxine-binding globulin were not caused by environmental or health factors, but were inherited in an autosomal dominant fashion [2]. Refetoff [3] has shown that the cause of genetically determined low thyroxine-binding globulin levels in Caucasians is alteration in synthesis rate without any structural variation of the protein. Since however the “low” thyroxine-binding globulin of Aborigines is vastly more prevalent and genetically distinct from the X-linked type, we investigated the possibility that this may be a structural variant. Evidence suggestive of this includes results from heat inactivation, competitive binding of thyroxine to thyroxine-binding globulin to measure affinity, and use of a radioimmunoassay different from that used in the original work. The “low” thyroxine-binding globulin of Aborigines may be a protein with a structural variation at or near the binding site for thyroxine, resulting in lower affinity for thyroxine and hence “low” results with assay methods which depend upon the thyroxine binding site. Since the Australoid peoples, to whom Australian Aborigines are racially related, are distributed widely throughout the southern hemisphere it is important to establish whether this variant is found outside Australia in order to avoid the likelihood of misdiagnosis of thyroid disease in such subjects.

Introduction Many Australian Aborigines appear to have a “low” thyroxine-binding (TBG) level in serum when the assay employed is one which utilizes

Address

for correspondence:

Dr. M. Dick, as above.

0009-898 I /8 I /OOOO-0000/$02.50

0 I98 1 Elsevier/North-Holland

Biomedical

Press

globulin the TBG

362

binding site for thyroxine (T4). Such subjects have low or low-normal T4 Ievels (about 30-80 nmol/l compared to Caucasian values of 60- 150 nmol/l) and TBG levels of 4- 11 mg/l compared to Caucasian levels of 12-28 mg/l [I]. The prevalence of this “low0 TBG appears to vary considerably throughout Australia, ranging from about 70% of subjects sampled in Arnhem Land (Northern Territory) to a low of about 22% of subjects near Cairns in Queensland. The overall prevalence in Australia is not known with any exactitude but may be about 40% of all Aboriginal subjects. By contrast, the genetic form of low TBG found in Caucasians has a prevalence of about 1 in 9000 and is transmitted in an X-linked fashion with the characteristic feature of lack of male to male transmission. The autosomal transmission of the ‘“low” TBG suggested to us that this might be a new genetically defined form of TBG. The possibility of a structural variant of TBG in Caucasian low TBG has been investigated and refuted by Refetoff and his co-workers [4] using such techniques as heat inactivation and T, affinity studies of the TBG. We have therefore used similar methods to investigate the “low” TBC of Aborigines. In addition it appeared to us that if there were a structurat variation in the Aboriginal TBG this must be at or near the binding site of ‘Tdin order to account for the low levels of T4 and the “low” TBG results obtained by a method which was dependent on the binding site. By contrast a method for TBG assay which is independent of the T, binding site can be predicted to give ‘rnormal” results for TBG if the major protein part of the molecule is unaffected by structural change and the synthesis rate is in fact similar to that in Caucasians. Electrophoresis might also be expected to indicate some difference between Aboriginal and Caucasian TBG, although the ~croheterogeneity of normal TBG on isoelectric focusing f5] is likely to make this difficult to prove. Simple electrophoresis on cellulose acetate membranes with [ ‘“51]thyroxine autoradiographs might however indicate whether the low TBG of Aborigines is the main carrier protein for thyroid hormone transport as it is in Caucasian subjects.

Samples of Aboriginal blood serum were obtained mainly from the Kimberley area of Western Australia. Caucasian samples were obtained from our routine biochemistry service. In both cases the serum used was surplus to the original requirement and was stored at - 20°C as previous work [2] had shown that TBG is very stable at this temperature.

(1) The routine assay and the one used in all previous work was the Immophase method (Corning Medical, Medfield, MA, USA). This, system is a “sandwich” radioimmunoassay employing ‘251-T4and is dependent upon the T4 binding site of TBG. (2) Our alternative method was the CIS-TBGK (Sorin Biomedica, Saluggia, Italy). This method uses ‘251-TBG and the assay is independent of T4 binding.

363

Heat inactivation of TBG This was carried out as described by Refetoff et al. [4]. Serum was heated at 60°C or 56°C and the remaining TBG was assayed by the Corning-Immophase method. Affinity studies of T, binding to TBG Competitive binding of T4 to TBG was performed by the method of Murphy and Jachan f6] as described by Refetoff et al. 141. The serum was diluted 1 : 40 with barbital buffer pH 8.6: dilution overcomes any effect of T4 binding to albumin while the barbital overcomes the binding of T4 to prealbumin (TBPA). The slope of the curve is then proportional to the T4 association constant. Electrophoresis of TBG This was performed in the conventional manner using a tris-glycine buffer pH 8.6. i25I-T, was added to the serum samples in order to perform autoradio~aphy after which the strips were stained for protein. ‘Thyroxine-binding prealbumin (TBPA) The method employed was immunodiffusion werke AG, Marburg, FRG).

using M.-Partigen plates. (Behring-

Results

Fig. 1 shows the results of heat inactivation studies. At 60°C Caucasian TBG and Aboriginal “normal” TBG both show a half-life of about 7 min, which is in good agreement with the findings of Refetoff et al, [4]. Aboriginal “low” TBG at this temperature shows a half-life of about 3 min. Because the time interval is rather short for the Aboriginal “low” TBG we repeated the experiments at 56’C where the time intervals are greater. The results at 56°C are similar to those at 60°C with the half-life of Aboriginal “low” TBG about half that of Caucasian TBG. TEG

WC

mg/L

11

1-b 5

10

15

20 TIME

i I MINUTES

10

1

Fig. 1. Heat inactivation of TBG as described by Refetoff et al. [4]. Serum was heated at 60°C or 56’C 0, Caucasian normal and the remaining activity assayed by the Corning method.0 Aboriginal Aboriginal normal TBG; A -A, Caucasian low TBG;a -A, TBG;. -0, “low” TBG.

364

Evidence for a difference at the binding site for T4 in Aboriginal subjects with “low” TBG is shown in Fig. 2. Here the slope of the curve is less steep than in a Caucasian subject with normal TBG, and in an Aboriginal with high TBG level, indicating a lesser affinity for T4 in the *‘low” TBG subjects. A high rather than “normal” Aboriginal TBG was chosen merely to give good separation of the curves in the diagram. The lack of parallelism in the Aboriginal “low” TBG is in contrast to the findings of Refetoff et al. 141in the low TBG of Caucasians. Table I shows the results of a typical experiment using the two different RIA methods for TBG on the same subjects. The same set of standards (Corning) was used in both assays to minimise any difference between methods in this respect. Here the Aboriginal ‘“low” TBG results found by the Corning method (T4 site dependent) become “normal” when assayed by the CIS method which does not depend upon T4 binding. Some 15 other “low” TBG sera have been assayed in similar manner: all give “normal” results. Fig. 3 shows an autoradio~aph using rzsI-T, of the binding proteins separated by electrophoresis on cellulose acetate. In Aboriginals with “normaf” TBG as in Caucasians, the major binding protein for T4 is TBG. In the Aboriginal with “low” TBG the major carrier of T4 however appears to be TBPA, with TBG showing a noticeably less dense band on the autoradiograph. Table II shows the results of TBPA assays on Aboriginal subjects with normal TBG, Aboriginal subjects with “low” TBG, and compares them with our reference range For Caucasians.

i

.

t

I

,

I

2

5

ia

20

50

TOTAL

T4 PER TUBE ( pmol b

2. Competitive binding of thyroxine to thyraxine-binding globulin. The slopes of the curves WC Aboriginal with high level of TBE: proportional to the thyroxine association constants. NI-m, l , Aboriginal with “low” level of TBG. A -------A% Caucasian with normal level of TBG: O-

Fig.

365 TABLE

I

COMPARISON OF TBG VALUES IN SERUM BETWEEN CAUCASIANS, ABORIGINES WITH NORMAL TBG, AND ABORIGINES WITH “LOW’ TBG BY TWO DIFFERENT RADIOIMMUNOASSAY METHODS (The Coming immophase method CIS-TBGK method uses 12’I-TBG

uses ‘251-T4 and is dependent on the T4 binding and is independent of the T4 binding site.)

site of TBG.

The

Serum TBG levels (mg/l)

I

Caucasian

2 3 4

I

Aboriginal normal TBG

2

I

Aboriginal “low” TBG

Caucasian

2 3 4 reference

range:

Corning immophase

CIS-TBGK

21 18 23 22

18 17 22 IX

21 20

21 18

7 7 4 4

15 16 16 21

12- 28 mg/l.

with Fig. 3. Autoradiograph of T4 binding proteins using ‘*sI-T, in (from left to right) an Aboriginal normal TBG, an Aboriginal with “low” TBG and a Caucasian. Upper bands, thyroxine-binding lower bands, TBG. Electrophoresis was performed on cellulose prealbumin; centre bands, albumin; acetate strips using a tris-glycine buffer pH 8.6.

366

Discussion

There is no doubt that large numbers of Aborigines have a variation from the normal pattern of thyroxine transport and that this appears to be due to a decreased affinity of TBG for thyroxine. We have shown that this variation does not seem to be caused by known environmental or health factors such as malnutrition, liver dysfunction or alcoholism, but instead seems to be genetically determined 121.The marked increase in susceptibility to heat inactivation suggests that the cause of the decrease in affinity is present in the TBG molecule and is not due to extraneous factors such as an abnormal metabolite or a drug. The evidence therefore suggests that the “low” TBG of Aborigines is in fact a variant form of TBG which may have a structural difference at or near the binding site for T4. The difference in structure could be a change of amino acid sequence, and if this is so and our genetical findings are correct, it would indicate that a structural gene for TBG sequencing is present on an autosome while the gene determining synthesis rate is X-linked. In Aborigines with variant TBG the main carrier of T4 appears to be TBPA, rather than TBG which carries about 70% of T4 under usual conditions. The actual serum levels of TBPA in subjects with the variant TBG do not however differ from normal (Table II) and this is understandable as only about 0.5% of the circulating TBPA is occupied by T4 in Caucasian subjects [3]. In Caucasian subjects with inherited low levels of TBG normal thyroid function is the rule, although there is a su~estion of an ass~iation with hyperthyroidism [7]. Aborigines with variant TBG do not appear to differ in other respects to those with Caucasian normal TBG levels and there does not appear to be any special association with thyroid dysfunction; however, no clinical investigation of this nature has as yet been undertaken. We have observed several cases of genuine hypothyroidism in subjects with the variant TBG; these subjects have T4 levels of less than 30 nmol/l and have the increased levels of serum TSH associated with this condition. We have also observed a few cases of subjects with variant TBG who appear to be hyperthyroid; this diagnosis is considerably more difficult from the laboratory point of view as reference ranges for T4 and T, are not yet known with any accuracy and free thyroxine index values are not relevant. We intend to establish such reference values for Aboriginal subjects with variant TBG. If the variant TBG of Aborigines occurred as a mutation in relatively modem times and is confined to Australia, the special problems in the diagnosis of thyroid TABLE

II

COMPARISON OF SERUM THYROXINE OF ABORIGINES (results in mg/l) Number subjects Aborigines Aborigines

with normal TBG with “low” TBG

Caucasian

reference

range

100-400

15 17 mg/l.

BINDING

of

PR~LBUMIN

LEVELS

IN TWO GROUPS

Mean

SD

Range

267 259

75 66

160-364 112-422

367

dysfunction can fairly readily be overcome; although it must be expected that in the future, Aboriginal and part-Aboriginal subjects will be found elsewhere in the world rather more commonly than they are at present. Another possibility exists however: that the variant TBG of Aborigines is an early form of TBG, in which case it may be found in other peoples of the southern hemisphere who have genetic affinity to the Australian Aborigine. Such peoples might include the inhabitants of New Guinea, the so-called Negritos of Malaya and the Philippines and possibly other groups ranging from Easter Island to Madagascar [S]. In many of these areas thyroid function testing will be either primitive or non-existent and where tests are introduced these will probably be of a simple nature, such as T4 and triiodothyronine uptake. Uncritical acceptance of Caucasian reference ranges with these simple tests will lead to misdiagnosis, if the variant form of TBG is present in the population. It will thus be important to establish whether the variant TBG is present in populations outside of Australia. Acknowledgements

We thank the Commissioner of Public Health for permission to publish this paper and Professor D.H. Curnow for discussion and comment. References 1 Dick M, Watson F. Prevalent low serum th~oxine-binding globulin ievei in Western Australian Aborigines. Med J Aust 1980; 1: 1IS- 118. 2 Watson F, Dick M. Distribution and inheritance of low serum thyroxine-binding globulin levels in Australian Aborigines, Med J Aust 1980; 2: 385-387. 3 Refetoff S. Thyroid hormone transport. In: De Grout LJ, Cahill GF, Ode11 WD, Martini L, Potts JT. Nelson DH, Steinberger E, Winegrad A, eds. Endocrinology, Vol. I. New York: Grune and Stratton: 1979; 347- 356. 4 Refetoff S, Robin NI, Alper CA. Study of four new kindred with inherited thyroxine-binding globulin abnormalities. J Clin Invest 1972; 5 I : 84% 867. 5 Lasne Y, Lasne F, Benzerara 0. ~croheterogeneity and polymo~hism of human serum thyroxinebinding globulin. Biochim Biophys Acta 1980; 624: 333-339. 6 Murphy BEP, Jachan C. The determination of thyroxine by competitive protein-binding analysis employing an anion-exchange resin and radiothyroxine. J Lab Chn Med 1965; 66: 161- 167. 7 Horwitz DL, Refetoff S. Graves’ disease associated with familial deficiency of thyroxine-binding globulin. J Clin Endocrinol Metab 1977; 44: 242-247. 8 Bellwood PS. The peopling of the Pacific. Scientific American 1980; 243: 138- 147.