Chen-Fee Lim

Transport of thyroxine into cultured hepatocytes: effects of mild non-thyroidal illness and calorie restriction in obese subjects

OBJECTIVE Inhibitors of cellular T4 transport leading to diminished plasma T3 production have been identified as 3‐carboxy‐4‐methyl‐5‐propyl‐2‐furanpropanoic acid (CMPF) and indoxyl sulphate in uraemia and bilirubin and non‐esterified fatty acids (NEFA) in critically ill patients with hyperbilirubinaemia. We question whether other factors are responsible for the altered thyroid hormone parameters observed in mild illness and during calorie restriction.

EFFECT OF ORAL FRUSEMIDE ON DIAGNOSTIC INDICES OF THYROID FUNCTION

We studied the acute effect of standard therapeutic doses of oral frusemide on indices of thyroid function in 34 hospital in‐patients with congestive cardiac failure. A transient decrease in total T4, elevation in the T3 resin uptake and consequent increase in the free T4 index (FT4I) were seen 2–5 h after ingestion of frusemide at a chronic morning dosage of 80,120 or 250 mg. The FT4I pre‐ vs post‐frusemide values after 250 mg of drug were 109 ± 12 vs 129 ± 18 (P < 0.05) after 120mg 92 ± 14vs 119 ± 12 (P < 001), and after 80 mg 102 ± 6 vs 112 ± 4 (P < 0.01) (mean ± SEM). Similar increases in apparent free T4 measured by an analogue tracer assay (free T4 RIA sol, Henning, Berlin) were seen after frusemide. In a time course study, the major change in the T3 uptake 120 min after frusemide ingestion correlated with the change in serum frusemide concentration. When frusemide was added to serum in vitro its influence was greatest in methods that involved least dilution of serum. In two of the patients difficulty in clinical assessment of thyroid status was compounded by the effect of oral frusemide on FT41. We conclude that oral frusemide may influence biochemical assessments of thyroid function in patients with congestive cardiac failure. It is necessary to consider the time interval between ingestion of high doses of oral frusemide and blood sampling in evaluating such results.

A naturally occurring furan fatty acid enhances drug inhibition of thyroxine binding in serum

We studied the thyroxine (T4)-displacing effects of a naturally occurring, highly albumin-bound furanoid acid that accumulates in serum in renal failure to concentrations in excess of 0.2 mmol/L. This substance, 3-carboxy-4-methyl-5-propyl-2-furanpropanoic acid (CMPF), has been shown to displace acidic drugs from albumin binding. The effects of CMPF on ligand binding were assessed in the following systems: (1) T4 binding to T4-binding globulin (TBG) and transthyretin (TTR), (2) T4 binding in undiluted serum, (3) T4-displacing potency of fenclofenac, furosemide, diflunisal, and aspirin in undiluted serum, (4) serum binding of [14C]-drug preparations, and (5) serum binding of [14C]-oleic acid. CMPF had a minor direct effect on T4 binding to TBG comparable in relative affinity to that of aspirin, ie, almost 7 orders of magnitude less than T4 itself. CMPF alone at a concentration of 0.3 mmol/L, which produced only a 10% to 14% increase in free T4 augmented the T4-displacing effects of high therapeutic concentrations of the various drugs in undiluted serum as follows: furosemide by 180%, fenclofenac by 160%, diflunisal by 130%, and aspirin by 40%. In the presence of fenclofenac, increments of CMPF from 0.075 to 0.3 mmol/L progressively augmented the T4-displacing effect of this drug, associated with a progressive increase in its calculated free concentration. CMPF also inhibited the binding of [14C]-oleic acid, suggesting that in some situations CMPF could also indirectly influence thyroid hormone binding by increasing the unbound concentration of nonesterified fatty acids (NEFA), as previously described.(ABSTRACT TRUNCATED AT 250 WORDS)

Non‐isotopic spectrophotometric determination of the unbound fraction of drugs in serum*

A spectrophotometric method to measure the free fraction of highly‐bound drugs in serum has been established for a range of non‐steroidal anti‐inflammatory drugs (NSAIDs) and for frusemide. Spectrophotometry is used to measure fractional transit of drug from a large volume of dialysate to a small volume of serum during dialysis to equilibrium. The method, which depends on the principle that drug transit from dialysate to serum is proportional to serum binding, requires neither isotopic drug preparations nor specific drug assays, is independent of extraction efficiency from the dialysate and requires no measurements from the serum compartment. Estimates of percent unbound fraction (%UF) for aspirin (6.0 ± 0.9%), phenylbutazone (0.9 ± 0.2%), and frusemide (1.8 ± 0.2%) were comparable with those obtained with 14C drug preparations. Values for %UF were determined for eleven additional NSAIDs. The method was valid for a four‐fold change in serum: dialysate ratio. Kinetics of frusemide binding to serum were comparable using [14C]frusemide and the test method. This technique may have general application in establishing the %UF for substances that are extensively bound to serum proteins and for identifying sera that show abnormal binding.

Measurement of the unbound fraction of long chain nonesterified fatty acids (NEFA) in serum: Methodological considerations

Long-chain nonesterified fatty acids (NEFA) are extensively bound to albumin; knowledge of their unbound concentrations is important in evaluating the numerous biologic effects attributed to these compounds. We measured the unbound fraction of five long-chain NEFA in serum using the equilibrium partition of 14C-NEFA between heptane and aqueous phases. Commercial 14C-NEFA preparations gave non-linear estimates of unbound fraction with serum dilution, consistent with the presence of polar tracer impurities, but 14C-NEFA purified by alkaline ethanol extraction gave an approximately linear relationship between unbound fraction and serum dilution over a 4096-fold range of dilution, provided that pH of the aqueous phase remained stable. Mean unbound percentages were: myristic acid 0.0066, linolenic acid 0.0019, arachidonic acid 0.0017, oleic acid 0.00078 and palmitic acid 0.00061. These data suggest that some previous studies appear to have overestimated the free fraction of long-chain NEFA at physiological albumin concentrations by at least one order of magnitude.

Effect of loop diuretics and nonsteroidal antiinflammatory drugs on thyrotropin release by rat anterior pituitary cells in vitro

The close inverse-feedback relationship between serum free thyroxine (T4) and thyrotropin (TSH) is altered in some patients receiving therapeutic doses of drugs such as furosemide, fenclofenac, and diphenylhydantoin. We therefore examined the effect of nonsteroidal antiinflammatory drugs (NSAID), diuretics, and diphenylhydantoin on TSH release in rat anterior pituitary cells in primary culture. TSH content of the culture medium was measured at 22 hours at 37 degrees C either with or without thyrotropin-releasing hormone ([TRH] 10 nmol/L) in medium containing 0.5% bovine serum albumin. The mean basal TSH release by pituitary cells was 6.2 +/- 1.2 ng/mL (n = 10) and was not influenced by unlabeled triiodothyronine ([T3] 100 nmol/L) or any of the drugs tested at < or = 400 mumol/L, except ethyacrynic acid. TRH 10 nmol/L increased mean TSH release by 346% +/- 95% (n = 10). T3 1 and 100 nmol/L inhibited TRH-stimulated TSH release by 24% and 31%, respectively (P < .001), whereas TRH-stimulated TSH release was inhibited by 100 mumol/L meclofenamic acid (29%), fenclofenac (28%), furosemide (24%), and diphenylhydantoin (48%) (P < .001 v TRH alone). Meclofenamic acid and furosemide (100 mumol/L) did not significantly alter the inhibitory effect of T3 1 nmol/L on TRH-stimulated TSH release. These in vitro studies suggest that meclofenamic acid, fenclofenac, furosemide, and diphenylhydantoin could influence TSH release by attenuating the TSH response to TRH. This effect may influence T4-TSH relationships when these agents are used in vivo.

Anomalous binding characteristics of human thyroxine binding globulin due to a dilution-dependent separation artefact

Contrary to the accepted view, a recent study using Sephadex column separation suggested that thyroxine binding globulin (TBG) binds T4 and T3 with similar affinity, but with a much larger capacity for T4 than T3. We have evaluated this finding by comparing this separation method with equilibrium dialysis, taking account of the effect of serum dilution with each method. Estimates of free T4 fraction by equilibrium dialysis (with magnesium chloride precipitation) were valid over a wide range of serum dilutions. In contrast, Sephadex column separation gave a major overestimate of free hormone (underestimate of binding) in less diluted serum, indicating that this method cannot be used to establish a value for T4 affinity independent of serum dilution. Such a systematic error will result in a greater underestimate of affinity for the ligand with higher affinity when two ligands are compared at a single serum dilution. By equilibrium dialysis at 37°C, the affinity of T4 for TBG was ∼13‐fold higher than that of T3, while the capacity of TBG for both T4 and T3 was close to the concentration of immunoreactive TBG. The previous report of similar T4 and T3 affinities appears to be due to a dilution‐dependent underestimate of T4 affinity inherent in Sephadex column separation. Direct comparison of binding kinetics of various ligands requires a separation method that is valid over a wide range of binding protein concentrations.