Measurement of soluble transferrin receptor in serum of healthy adults

¹ R&D Systems, Inc., 614 McKinley Place NE, Minneapolis, MN 55413.

² State University of New York Health Science Center at Brooklyn, Brooklyn, NY.

³ University of Cincinnati Medical Center, Cincinnati, OH.

⁴ VA Medical Center, Minneapolis, MN.

⁵ University of Colorado Health Sciences Center, Denver, CO.
^a Author for correspondence. Fax 612/379-6580; e-mail tomd{at}rndsystems.com.

	Abstract

Top Abstract Introduction Materials and Methods Results and Discussion References

 
The concentration of soluble transferrin receptor (sTfR) in serum is reported to be useful in the diagnosis of iron deficiency, especially for patients with concurrent chronic disease, where routine tests of iron status are compromised by the inflammatory condition. A new diagnostic assay for sTfR is calibrated against natural plasma sTfR, thus minimizing calibration discrepancies that result from differences between the analyte and the cellular transferrin receptor used in other assays. Use of the new assay to measure sTfR concentrations in 225 healthy, hematologically normal adults provided a reference interval against which pathological samples could be compared. There was no difference in the reference intervals for men and women and no correlation of [sTfR] with the age of the subject. Black subjects had significantly higher concentrations than nonblacks, and people living at high altitude had higher concentrations than those living closer to sea level. These differences were additive.

	Introduction

Top Abstract Introduction Materials and Methods Results and Discussion References

 
The uptake of iron by cells is mediated by a transferrin receptor (TfR)¹ (1) expressed on their external surface (reviewed in ref. 2). TfR binds diferric transferrin, and the receptor–transferrin complex is internalized into an endosome, where the iron is transferred to the cytosol. After recycling to the cell surface, the apotransferrin dissociates, and the receptor is free to repeat the process. Cells deficient in iron upregulate expression of TfR to compete more successfully for available iron (2)(3)(4).

TfR is a disulfide-linked dimer of two identical 85-kDa subunits (1)(2)(5)(6). Each subunit has a 61-amino acid N-terminal cytoplasmic domain, a transmembrane region, and a large extracellular domain. TfR is shed from cells by proteolytic cleavage at Arg₁₀₀-Leu₁₀₁ (7)(8), just external to the plasma membrane and just after the two interchain disulfide bonds. The product circulates in the blood as soluble TfR (sTfR), a 74-kDa monomer (9) bound to transferrin. The amount of circulating sTfR is proportional to the total amount of cell-associated TfR (10).

Because TfR expression is upregulated when a cell needs more iron and because sTfR is proportional to total TfR, concentrations of sTfR, [sTfR], are increased in plasma or serum of an iron-deficient subject. Some have found the concentration of serum sTfR useful in the diagnosis of iron deficiency (4)(11)(12)(13), especially in patients with chronic inflammatory, infectious, or malignant disease (14)(15)(16)(17), where the usual tests of iron status may be misleading. There has, however, been no comprehensive evaluation of reference interval data for effective use of sTfR as a diagnostic tool. In iron deficiency, [sTfR] increases to about twice the normal concentration (16)(17), but the distribution of values is broad enough that the 95% ranges may overlap. This makes it important to have a reliable and precise assay system, as well as a well-established reference interval and an understanding of the factors that influence it.

We describe here a new assay for the measurement of sTfR as an aid in the diagnosis of iron deficiency, and we report the reference interval of [sTfR] in healthy adults. We investigated the effects of sex, race, age, and altitude of residence—factors known to affect other hematological variables.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results and Discussion References

 
Preparation of monoclonal antibodies.
Monoclonal antibodies were from hybridoma mouse cells prepared with standard protocols for immunization (18), fusion (19), selection of hybrid cells (20), and screening for positive clones. Immunization was with human placental TfR purified according to Turkewitz et al. (21).

Among several screening assays, the most important was for antibodies that could capture sTfR from a pool of normal human serum. We evaluated 14 antibodies, choosing 1 for purification of sTfR from plasma by immunoaffinity chromatography and 2 for use in the sTfR assay.

Preparation of the master calibrator.
To assure uniformity of future lots of ELISA kits, we prepared a large batch of highly purified and well-characterized plasma sTfR for lyophilization in many small aliquots for use as a master calibrator. The method was adapted from that of Shih et al. (7). A fraction precipitated by ammonium sulfate saturated between 40% and 60% was dissolved in a minimum volume of phosphate-buffered saline (PBS; 137 mmol/L NaCl, 8.1 mmol/L Na₂HPO₄, 1.5 mmol/L KH₂PO₄, 2.7 mmol/L KCl, and 0.2 g/L NaN₃, pH 7.4) and then dialyzed against the same buffer. The retentate was chromatographed on an affinity column of immobilized anti-TfR monoclonal antibody (different from those used in the assay kit). sTfR plus sTfR-bound transferrin were retained on the column. After the column was washed with PBS, transferrin was eluted with 0.5 mol/L NaSCN in PBS. After a further wash with 10 column volumes of PBS, sTfR was eluted with ethylene glycol, 250 mL/L in 0.1 mol/L triethylamine (pH 11.5). The eluate was dialyzed against 20 mmol/L Tris (pH 8.0), filtered through a 0.2-µm pore-size filter, and chromatographed on a Mono-Q HPLC column (Pharmacia). This elution was made with a gradient from 0 to 0.5 mol/L NaCl in 20 mmol/L Tris (pH 8.0). Fractions from the column were then run on nonreducing 5–15% polyacrylamide gels. Fractions with pure sTfR were pooled and dialyzed against PBS without azide.

The authenticity and purity of sTfR were established by N-terminal sequence analysis, by sodium dodecyl sulfate–polyacrylamide gel electrophoreses (SDS-PAGE), and by analysis of amino acid composition, and all results were compared with published data. The mass of purified protein was calculated from the amino acid analysis. The ratio of mass to ELISA immunoreactivity of the master calibrator sTfR and of partially purified sTfR (mass estimated by quantitative SDS-PAGE) established that the extensive purification did not modify the immunoreactivity of the sTfR.

Preparation of kit calibrators.
To obtain a sufficient yield of sTfR for kit calibrators, we modified the preparation of pure sTfR by elimination of the final Mono-Q column and the chaotropic removal of transferrin from sTfR on the immunoaffinity column. The process resulted in a high yield of material that was 35–45% sTfR. Calibrators were prepared from this material by reference to the immunoreactivity of the master calibrator.

sTfR assay.
The assay operates on the quantitative two-site immunoenzymometric ("sandwich") technique. A monoclonal antibody specific for sTfR is precoated onto the wells of a microplate. Then, 20 µL of calibrator or sample and 100 µL of assay diluent are added to the wells and incubated for 1 h, during which sTfR becomes bound to the immobilized antibody. After any unbound material is washed away, 100 µL of a second monoclonal antibody conjugated to horseradish peroxidase is added and incubated for 1 h, during which the conjugate binds to the captured sTfR. After another washing away of unbound material, the amount of bound conjugate is detected by reaction for 30 min with a specific substrate, which yields a colored product that is proportional to the amount of conjugate (and thus to the amount of sTfR in the sample).

The color reaction is stopped with hydrochloric acid, and the concentration of sTfR in each well is read from a calibration curve of absorbance vs [sTfR].

Sample collection.
Three hundred healthy, adult, paid volunteers were recruited by advertisement at four centers, three at or below an altitude of 300 m (New York, Cincinnati, and Minneapolis) and one at 1600 m (Denver) above sea level. The study was approved by the Institutional Review Board at each center, and informed consent was obtained in writing from all subjects. Pregnant women, recent or frequent blood donors, and persons taking prescribed medication (except hormone contraceptives or hormone replacement therapy) were excluded.

Brief demographic data (age, sex, pre- or postmenopausal status, racial group, and illness in the past 6 months) were collected by interview, and a blood sample of 25 mL was drawn by venipuncture. Of this, 5 mL was drawn into anticoagulant; the remainder was allowed to clot and was then centrifuged to separate the serum. The whole blood and 5 mL of the serum were sent to a reference laboratory (Laboratory Corporation of America, Raritan, NJ) for a complete blood count and determination of iron status (serum ferritin, serum iron, total-iron binding capacity, and transferrin saturation). The remaining serum was stored in 200-µL aliquots at -70 °C until the end of the study, after which one aliquot was thawed and analyzed for sTfR.

Statistical analysis.
A multifactorial analysis of variance was used to assess the effects of sex, pre- or postmenopausal status in women, race (FDA classification (22)), and altitude on sTfR concentration. The effect of age was investigated by regression analysis (MacAnova, School of Statistics, University of Minnesota, St. Paul, MN; Statistica, StatSoft, Tulsa, OK).

	Results and Discussion

Top Abstract Introduction Materials and Methods Results and Discussion References

 
Assay characteristics.
The assay was calibrated with [sTfR] from 3 to 80 nmol/L; values outside this range were not read. The minimum detectable dose, calculated as the concentration represented by 2 SD above the mean of 20 replicates of the zero calibrator, was 0.5 nmol/L. There was no cross-reactivity with human diferric transferrin or apotransferrin or with ferritin from human heart, liver, or spleen. Intraassay precision was assessed by measuring 20 replicates of three samples at one time. The means and (CVs) were 6.9 (4.3%), 18.2 (7.1%), and 40.2 (6.2%) nmol/L. Interassay precision was assessed by measuring three samples in 60 separate assays spread over 3 months, with use of three batches of reagents by multiple operators. The means (and CVs) were 10.9 (6.4%), 25.7 (5.4%), and 67.2 (5.7%) nmol/L.

Linearity was established by serial dilution of five serum samples supplemented with sTfR to concentrations of 52–70 nmol/L (i.e., from 65% to 88% of the highest-concentration calibrator). Dilutions from 1:2 to 1:16 with sample diluent showed a mean recovery of 101% (range 95–111%). Analytical recovery was established by addition of sTfR (16, 32, or 48 nmol/L) to each of five serum samples with endogenous [sTfR] of 10.8–21.9 nmol/L. The mean recovery of added [sTfR], calculated as (total measured [sTfR] - endogenous [sTfR]) x 100/added [sTfR], was 97% (range 95–100%).

Calibration against plasma sTfR was intended to improve the accuracy of the assays results. Assays used in published studies had been calibrated with cellular disulfide-linked dimeric TfR from placenta. This introduces inherent uncertainty about whether the dimer behaves in the assay the same as two monomers would and whether dimer-specific or dimer-masked epitopes are involved in the assay. The difficulty of assigning a value to a calibrator that differs from the analyte is illustrated by the different mean values quoted for normal adults for other assays. The published means vary from 0.25 mg/L (23) to 5.36 mg/L (14). For comparison, the mean reported here, 19.6 nmol/L, corresponds to 1.47 mg/L, or 1.66 mg/L of a single subunit of intact cellular TfR.

sTfR reference interval.
Data from 225 subjects were used in the statistical analysis (Table 1 ). Seventy-five of the 300 subjects recruited were omitted from the calculation, because either data were missing, entry criteria were not observed, or the complete blood counts and iron status assays indicated anemia, abnormalities of erythrocyte production, iron deficiency, or iron overload (Table 2 ).

Individual [sTfR] ranged from 7.6 to 37.7 nmol/L (mean = 19.6 ± 5.0); their apparent conformation to gaussian distribution (Fig. 1 ) was verified with a Shapiro–Wilk test (W = 0.976, P = 0.1). There was no correlation with age (range 19–79 years, r = 0.008, P = 0.901), and there were no statistically significant differences between men and women (consistent with previous reports (23)(24)) or between pre- and postmenopausal women.

View larger version (41K): [in this window] [in a new window]   Figure 1. Histogram and fitted gaussian distribution curve of [sTfR] in healthy adults. The bars represent observed values; the curve shows the expected values for a gaussian distribution.

Two statistically significant differences were observed, however. First, black subjects had ~9% higher values for [sTfR] than did Caucasians, Asians, and Hispanics, whereas values for the latter three groups did not differ significantly. The difference may be related to the well-known but unexplained difference in hemoglobin concentrations ([Hb]) in blacks vs Caucasians (concentrations in blacks being ~5 g/L less (25)). Second, subjects residing at high altitude had concentrations ~9% higher than those nearer to sea level. To eliminate the possibility that the difference at high altitude could be due to the inclusion in Denver of individuals who might be regarded as anemic, the mean [sTfR] was recalculated for this site, but with the lower limits of the reference ranges for [Hb] at the University of Colorado Hospital (135 and 145 g/L for women and men, respectively) being used as the cutoff values for inclusion in this study. The effect on the mean [sTfR] was minimal: 20.8 nmol/L with lower cutoff, 20.9 with higher cutoff. Because these two effects, race and altitude, were independent and additive, separate 95% reference intervals (mean ± 2 SD) were calculated for black and nonblack subjects at low and high altitudes (Table 3 ).

The correlations between [sTfR] and the variables related to erythrocyte production were weak but statistically significant: Hb (r = 0.304, P <0.001, Fig. 2 ), hematocrit (r = 0.319, P <0.001), and erythrocyte count (r = 0.380, P <0.0001). This trend was expected, because serum [sTfR] in healthy individuals reflects erythropoietic rate rather than iron status. Examination of individual factors, however, showed inconsistencies: [Hb] was much higher in the men than in the women (157 vs 137 g/L, respectively), as expected, but [sTfR] values were similar; black subjects had lower [Hb] than nonblacks (144 vs 148 g/L) but higher [sTfR]; and subjects in Denver had both higher [Hb] than those at low altitude (148 vs 144 g/L) and higher [sTfR].

View larger version (22K): [in this window] [in a new window]   Figure 2. Correlation of [sTfR] with [hemoglobin] in healthy adults. Shown are the linear regression curve and the 95% confidence intervals of the regression for the measured values (r = 0.304, P <0.001).

Of 300 subjects recruited for this study, many of whom were employees of major health facilities, 25% were excluded for hematological reasons. In this study we set rigorous inclusion criteria to ensure a truly "normal" population. For more-routine determinations of the reference interval for [sTfR], only individuals with iron deficiency or an abnormal erythrocyte profile need to be excluded (see Table 2 ).

With a sufficiently large population, an alternative method (26) may be used, relying on the exclusion of outliers to determine the reference interval and thereby avoiding the need for other blood tests and postrecruitment exclusion criteria. We tested this with the sTfR measurements from all 300 subjects. Sequential removal of significant (P <0.01) outliers until no more could be eliminated removed 15 of the values as outliers. The reference interval was then taken as the mean ± 2 SD of the 285 values remaining, for a 95% reference interval of 10.3–29.1 nmol/L (mean [sTfR] 19.7 nmol/L). This is very close to the mean (19.6 nmol/L) and 95% range (9.6–29.6 nmol/L) obtained by carefully excluding from the study any individuals with abnormal blood test results.

	Acknowledgments

 
We thank Douglas Hawkins (University of Minnesota, Department of Applied Statistics) for help with statistical analysis. This study was sponsored by R&D Systems, Inc., Minneapolis. P.A.S. is a coinventor on patents dealing with sTfR that are owned by the University of Colorado.

	Footnotes

 
¹ Nonstandard abbreviations: TfR, transferrin receptor; sTfR, soluble transferrin receptor; PBS, phosphate-buffered saline; SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoreses; Hb, hemoglobin.

	References

Top Abstract Introduction Materials and Methods Results and Discussion References

 

1. Jing S, Trowbridge IS. Identification of the intermolecular disulfide bonds of the human transferrin receptor and its lipid-attachment site. EMBO J 1987;6:327-331. [ISI][Medline] [Order article via Infotrieve]
2. Harford JB, Rouault TA, Heubers HA, Klausner RD. Molecular mechanisms of iron metabolism. Stamatoyannopoulos G Nienhuis AW Majerus PW Varmus H eds. The molecular basis of blood diseases 2nd ed. 1994:351-378 Saunders Philadelphia. .
3. Harford JB, Rouault TA, Klausner RD. The control of cellular iron homeostasis. Brock JH Halliday JW Pippard MJ Powell LW eds. Iron metabolism in health and disease 1994:123-149 Saunders London. .
4. Baynes RD. Iron deficiency. Brock JH Halliday JW Pippard MJ Powell LW eds. Iron metabolism in health and disease 1994:190-225 Saunders London. .
5. Seligman PA, Schleicher RB, Allen RH. Isolation and characterization of the transferrin receptor from human placenta. J Biol Chem 1979;254:9943-9946. [Abstract/Free Full Text]
6. Schneider C, Owen MJ, Banville D, Williams JG. Primary structure of human transferrin receptor deduced from the mRNA sequence. Nature 1984;311:675-679. [Medline] [Order article via Infotrieve]
7. Shih YJ, Baynes RD, Hudson BG, Flowers CH, Skikne BS, Cook JD. Serum transferrin receptor is a truncated form of tissue receptor. J Biol Chem 1990;265:19077-19081. [Abstract/Free Full Text]
8. Baynes RD, Shih YJ, Hudson BG, Cook JD. Production of the serum form of the transferrin receptor by a cell membrane-associated serine protease. Proc Soc Exp Biol Med 1993;204:65-69. [Abstract]
9. ExPA. Sy World Wide Web molecular biology server of the Geneva University Hospital and the University of Geneva. Theoretical pI/Mw for protein sequence. http://expasy.hcuge.ch/cgi-bin/pi_tool..
10. Beguin Y, Huebers HA, Josephson B, Finch CA. Transferrin receptors in rat plasma. Proc Natl Acad Sci U S A 1988;85:637-640. [Abstract/Free Full Text]
11. Kohgo Y, Nishisato T, Kondo H, Tsushimi N, Niitsu Y, Urushizaki I. Circulating transferrin receptor in human serum. Br J Haematol 1986;91:277-281.
12. Skikne BS, Flowers CH, Cook JD. Serum transferrin receptor: a quantitative measure of tissue iron deficiency. Blood 1990;75:1870-1876. [Abstract/Free Full Text]
13. Thorstensen K, Romslo I. The transferrin receptor: its diagnostic value and it potential as therapeutic target. Scand J Clin Lab Invest 1993;Suppl215:113-120.
14. Ferguson BJ, Skikne BS, Simpson KM, Baynes RD, Cook JD. Serum transferrin receptor distinguishes the anemia of chronic disease from iron deficiency anemia. J Lab Clin Med 1992;119:385-390. [ISI][Medline] [Order article via Infotrieve]
15. Punnonen K, Irjala K, Rajamaki A. Iron-deficiency anemia is associated with high concentrations of transferrin receptor in serum. Clin Chem 1994;40:774-776. [Abstract/Free Full Text]
16. Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptor and its ratio to serum ferritin in the diagnosis of iron deficiency. Blood 1997;89:1052-1057. [Abstract/Free Full Text]
17. Pettersson T, Kivivuori SM, Slimes MA. Is serum transferrin receptor useful for detecting iron deficiency in anaemic patients with chronic inflammatory diseases?. Br J Rheumatol 1994;33:740-744. [Abstract/Free Full Text]
18. Lucas C, Bald LN, Fendly BM, Mora-Worms M, Figari IS, Patzer EJ, Palladino MA. The autocrine production of transforming growth factor-ß1 during lymphocyte activation: a study with a monoclonal antibody-based ELISA. J Immunol 1990;145:1415-1422. [Abstract]
19. Gefter ML, Margulies DH, Scharff MD. A simple method for polyethylene glycol promoted hybridization of mouse myeloma cells. Somatic Cell Genet 1977;3:231-236. [ISI][Medline] [Order article via Infotrieve]
20. Foung SKH, Sasaki DT, Grumet FC, Engleman EC. Production of functional human T-T hybridomasin selection medium lacking aminopterin and thymidine. Proc Natl Acad Sci U S A 1982;79:7484-7488. [Abstract/Free Full Text]
21. Turkewitz AP, Amatruca JF, Borhani D, Harrison SC, Schwartz AL. A high yield purification of the human transferrin receptor and properties of its major extracellular fragment. J Biol Chem 1988;263:8318-8325. [Abstract/Free Full Text]
22. . National Institutes of Health. Guidelines on the inclusion of women and minorities as subjects in clinical research. Fed Reg 1994;59(Part IV March 9):11146-11151.
23. Kohgo Y, Niitsu Y, Nishisato T, Kato J, Kondo H, Sasaki K, Urushizaki I. Quantitation and characterization of serum transferrin receptor in patients with anemias and polycythemias. Jpn J Med 1988;27:64-70. [Medline] [Order article via Infotrieve]
24. Huebers HA, Beguin Y, Pootrakul P, Einspahr D, Finch CA. Intact transferrin receptors in human plasma and their relation to erythropoiesis. Blood 1990;75:102-107. [Abstract/Free Full Text]
25. Perry GS, Byers T, Yip R, Margen S. Iron nutrition does not account for the hemoglobin differences between blacks and whites. J Nutr 1992;122:1417-1424.
26. Hawkins DM. Outliers. Kotz S Johnson NL eds. Encyclopedia of statistical science 1985;Vol. 6:539-543 Wiley New York. .