A

acsihto0twc©

K

1

hatfnl

j

0d

Available online at www.sciencedirect.com

Talanta 75 (2008) 965–972

Development of a sensitive, rapid, biotin–streptavidin basedchemiluminescent enzyme immunoassay for human

thyroid stimulating hormone

Zhen Lin a,b, Xu Wang a,∗, Zhen-Jia Li c, Shi-Qi Ren a,b, Guo-Nan Chen b,Xi-Tang Ying c, Jin-Ming Lin a,∗

a The Key Laboratory of Bioorganic Phosphorus Chemistry & Chemical Biology, Department of Chemistry,Tsinghua University, Beijing 100084, China

b Department of Chemistry, Fuzhou University, Fuzhou 350002, Chinac Beijing Chemclin Biotech Co., Ltd., Beijing Academy of Science and Technology, Beijing 100094, China

Received 11 September 2007; received in revised form 19 December 2007; accepted 22 December 2007Available online 6 January 2008

bstract

A highly sensitive “two-site” chemiluminescent immunoassay specific for human thyroid stimulating hormone (TSH) was developed. The signalmplification was achieved via a biotin–streptavidin system (BSAS). The HRP–luminol–H2O2 chemiluminescent system with high sensitivity washosen as the detection system. Biotinylated anti-TSH monoclonal antibody (MAb) and HRP-labeled streptavidin were first synthesized. Then theignal amplification was achieved through the interaction between the biotinylated anti-TSH MAb and the HRP–streptavidin conjugate. The lightntensity developed was in proportion to the TSH present in the samples. The assay showed little cross-reactivity with three other glycoproteinormones (human chorionic gonadotropin (HCG), luteinizing hormone (LH), follicle stimulating hormone (FSH)) due to the high specificity ofhe antibody. The working range for human thyroid stimulating hormone was 0.1–40 mU L−1. Both the intra-assay and inter-assay coefficientsf variation were less than 10% for the BSAS based chemiluminescent enzyme immunoassay (CLEIA). The proposed assay had a sensitivity of.01 mU L−1 which was 10-fold higher than the HRP–MAb conjugate based TSH immunoassay. Thus the higher sensitivity facilitated the clinical

esting for thyroid states. The effects of several reaction parameters, such as incubation time, temperature, and reaction volume of the method,ere also studied. This method has been successfully applied to the evaluation of TSH in human serum. Compared with the commercial enzyme

hemiluminescent immunoassay, the correlation was satisfied.2008 Elsevier B.V. All rights reserved.

ignal

iirdes

eywords: Human thyroid stimulating hormone; Biotin–streptavidin system; S

. Introduction

Thyroid gland function is regulated by the thyroid stimulatingormone (TSH), a hormone secreted by pituitary gland and withmolecular weight of approximately 28,000 Da. TSH secre-

ion is in turn regulated by hypothalamic thyroliberin and by a

eedback-inhibiting loop in which free hormones (triiodothyro-ine (T3), thyroxine (T4)) act at both pituitary and hypothalamicevels [1]. Sensitive TSH measurement for trace amount of TSH

∗ Corresponding authors. Tel.: +86 10 62792343; fax: +86 10 62792343.E-mail addresses: wang-x@mail.tsinghua.edu.cn (X. Wang),

mlin@mail.tsinghua.edu.cn (J.-M. Lin).

vhh

mini

039-9140/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.talanta.2007.12.043

amplification; Chemiluminescent immunoassay

n human fluids is utilized as a “first line” thyroid test [2,3] and isn great need to distinguish hyperthyroid and primary hypothy-oid states from euthyroidism, as well as to facilitate the clinicalecision-making. Usually, elevated TSH levels indicate an inad-quate thyroid hormone production, while suppressed levelsuggest excessive unregulated production of hormone [4,5]. Byirtue of its obvious importance for the clinical diagnosis, thereas been an ongoing effort to develop analytical methods withigh sensitivity for TSH in human body fluids.

The traditional analytical methods for TSH include radioim-

unoassay (RIA), immunoradiometric assay (IRMA [3]),

mmunoenzymetric assay (IEMA) [6–9], immunochemilumi-escent assay (ICMA) [10–14], time-resolved fluorescencemmunoassay [15], bioluminescent immunoassay [16,17] as

9 nta 75

wRtw[amsiaHbafllihws

(narClfe

wnmpawsatbctttiBr

hcCTwwMiTi

0bd6tt

2

2

HAl9Cpvtc(cD

2

vLaCBc(wmGwq

tbp0i0l02

2

66 Z. Lin et al. / Tala

ell as some homogeneous particle based immunoassays [18].IA which used competitive mode required long incubation

ime and was relatively labor intensive. 125I-labeled TSHas less stable than the 125I-labeled immunoglobulin (IgG)

6,19]. The sensitivity of the method (0.5–1 mU L−1) precludedccurate measurement of TSH in euthyroid subjects and in pri-ary hyperthyroid. Although there was an improvement in the

ensitivity of IRMA, there still needed a great effort to elim-nate radioactive materials in the assay. ICMA, which usedcridinium ester or luminol derivatives, has high sensitivity.owever, a precise and automatic injection system is neededecause of the flash-type chemiluminescent reaction. Expensivend complex instrumentations are necessary for time-resolveduorescence immunoassay. Although some methods such as

uminescent oxygen channeling assay (LOCTTM) [20] and somemmunoassays with novel chemiluminescent labels [21] haveigh sensitivity, it is still necessary to develop various methodsith high sensitivity and simple operation, which are especially

uitable for the daily clinical analysis.In recent years, chemiluminescence enzyme immunoassay

CLEIA) has been widely used in the research of clinical diag-osis because of its advantages such as no radioactive pollution,cceptable sensitivity, rapid turnaround time, wide dynamicange, and glow-type chemiluminescent reaction. However,LEIA reported for TSH were not so satisfied because of their

imited sensitivity in the range of 0.04–4.5 mU L−1 [7–9]. There-ore, a methodology with sensitivity improvement for CLEIA toxtend its clinical application was in urgent needed.

Avidin, a glycoprotein found in egg white with moleculareight of 67 kDa, contains four binding sites with an extraordi-ary affinity (dissociation constant: about 10−15 M) for the smallolecule vitamin, biotin. Biotin could be easily covalently cou-

led to proteins, enabling a solid binding between the proteinsnd avidin. The biotin–streptavidin system (BSAS) has beenidely used in immunohistochemistry [22,23] and immunoas-

ay [24–28] because of its high specificity [29] and strongffinity [30]. Being attractive for researchers, excellent sensi-ivity could be achieved through signal amplification introducedy BSAS. In such a case, numbers of active biotins and enzymesould be conjugated to per antibody and per streptavidin, respec-ively, enabling more enzyme molecules catalyzing the substratehan the non-BSAS system [30,31]. Obviously, signal amplifica-ion could be implemented. Although the significance of BSASn highly sensitive analysis, unfortunately, the combination ofSAS and CLEIA for the clinical determination of TSH was

arely reported.In the present work, the combination of BSAS with the

orseradish peroxidase (HRP)–luminol–H2O2 chemilumines-ence system was performed to develop a highly sensitiveLEIA for TSH in human serum. Herein, biotinylated anti-SH monoclonal antibody (MAb) and HRP-labeled streptavidinere first synthesized. Then a signal amplification systemas achieved through the interaction between the biotinylated

Ab and the HRP–streptavidin conjugate. The utility of BSAS

ncreased the amount of enzyme linked to the immunocomplex.hus the light intensity was amplified and the sensitivity was

mproved.

2

c

(2008) 965–972

The BSAS based method showed a sensitivity of.01 mU L−1 versus 0.1 mU L−1 of the HRP–MAb conjugateased TSH immunoassay, meeting the demands of the clinicaliagnosis. The assay for TSH was completed within 70 min (with0-min immunoreaction and 5-min chemiluminescent incuba-ion). The use of 96-well microplate also achieved the highhroughput screening.

. Experimental

.1. Apparatus

Chemiluminescence microplate reader was purchased fromamamatsu photons Technology Co. Ltd. (Beijing, China).utomatic plate washer was got from Beijing Tuopu Ana-

ytical Instruments Co. Ltd. (Type: DEM-�)(Beijing, China).6-well microplates were from ShenZhen Jincanhua Industryo. Ltd. (Shenzhen, China). Electronic balance (AUY-120) wasurchased from Shimadzu Co. Ltd. (Japan). A model XW-80Aortex mixer (Jingke Industrial, Shanghai, China) was used forhorough mixing of substances. The incubation procedures werearried out using constant temperature incubator (HH.W21-Cr)Chang’an scientific instrument factory, Beijing, China). Single-hannel volume-adjustable micropipets were purchased fromragon Medical & Instruments Co. Ltd. (Shanghai, China).

.2. Chemicals and solutions

(+)-Biotin-N-hydroxysuccinimide ester (BNHSE), strepta-idin and HRP were purchased from Sigma–Aldrich Co. (St.ouis, MO, USA). Anti-TSH MAbs were obtained from Fitzger-ld Industries International, Inc. (Clone: M94205 and M94206,A, USA). TSH antigen with standard grade was from Maineiotechnology Services, Inc. (Portland, USA). Commercialhemiluminescence kit was purchased from Monobind, Inc.CA, USA). Polyoxyethylenesorbitan monolaurate (Tween-20)as from Sigma Chemical Co. (MO, USA). Bovine serum albu-in (BSA) and hydrolyzed gelatin were from Merck (Darmstadt,ermany). Luminol, chemiluminescent enhancer, and H2O2ere obtained from Monobind Inc. (CA, USA). EZTM biotinuantitation kit was purchased from Pierce (USA).

Highly purified distilled and deionized water was usedhroughout. The coating solution was 0.05 mol L−1 carbonateuffer (pH 9.5). The blocking solution was 0.01 mol L−1 phos-hate buffer (pH 7.4) with 1% (w/v) BSA, 2.5% (w/v) sucrose,.5% (w/v) gelatin, and 0.1% (v/v) proclin-300. The wash-ng solution was 0.01 mol L−1 phosphate buffer (pH 7.4) with.1% (v/v) Tween-20. The dilution solution for the biotiny-ated anti-TSH MAb and the HRP–streptavidin conjugate was.01 mol L−1 phosphate buffer (pH 7.4) with 0.1‰ (v/v) Tween-0, 1‰ (w/v) BSA, and 0.5% (v/v) mouse serum.

.3. Procedure

.3.1. Biotinylation of anti-TSH MAbAnti-TSH MAb was first dialyzed against 0.05 mol L−1

arbonate buffer for 4 h. BNHSE was dissolved in dimethyl sul-

nta 75 (2008) 965–972 967

fmtAbwu

2

oTdfBdtattwts

2

MTiaftcu

2H

TaAtw9at(4pracs

2

i

otHfifs1AwbMj

2o

tSmaaai

2

(lo

3

3

Z. Lin et al. / Tala

oxide, and then was added 20-, 40-, 50-, 100- and 150-foldolar excess against anti-TSH MAb in the dialysis solu-

ion. The solution was put still at room temperature for 2 h.fter that, it was dialyzed against 0.01 mol L−1 phosphateuffer to remove unreacted materials. Finally, the conjugateas preserved in glycerol and stored at 4 ◦C for furtherse.

.3.2. Calibrator preparationThe calibrators were prepared by spiking different amounts

f TSH into TSH-free serum, and then were stored at 4 ◦C.hen their concentrations were calibrated against national stan-ards. TSH national standards (600 mU/ampoule) were boughtrom National Institute for The Control of Pharmaceutical andiological Product (NICPBP). 600 mU of lyophilized TSH wasissolved in TSH-free serum and was diluted to serial concen-rations. The concentrations of the calibrators were determineds the unknown samples by the use of national standards for fiveimes. The average values were considered as the calibrators’rue concentrations. The final concentrations of these calibratorsere 0.1, 0.5, 2.5, 10, and 40 mU L−1, respectively. Meanwhile,

he prepared calibrators were store at 37 ◦C to investigate thetability.

.3.3. Immobilization of anti-TSH MAb on the microplateAnti-TSH MAb was first diluted by the coating buffer. Then

Ab solution of 110 �L was pipetted into each well on the plate.he plate was placed still for overnight at 4 ◦C to accomplish the

mmobilization. After that, 300 �L of the blocking buffer wasdded into each well and the plate was left at room temperatureor 3 h in order to block the free sites on the plate. Subsequently,he solution in the well was aspirated and the plate was desic-ated. Finally, the plate was vacuumized and stored for furtherse.

.3.4. Synthesis of HRP–MAb conjugate andRP–streptavidin conjugate0.5 mg of HRP was dissolved in 500 �L of deionized water.

hen freshly prepared 0.1 mol L−1 sodium periodate was addednd the solution was stirred at room temperature for 30 min.fter that, glycol (0.16 mol L−1) solution was added to cease

he oxidation reaction. Meanwhile, 0.5 mg of anti-TSH MAbas first dialyzed against 0.01 mol L−1 carbonate buffer (pH.5) for 4 h, and then was added into HRP solution mentionedbove. The mixture was stirred for 2 h at room temperature. Afterhat, 50 �L of freshly prepared sodium borohydride solution4 mg mL−1) was added and the mixture obtained was placed at◦C over a period of 2 h. The solution was then dialysed againsthosphate buffer (pH 7.4) for 24 h to remove the unreacted mate-ials. Finally, the conjugate was preserved in glycerol and storedt 4 ◦C. The procedure for the synthesis of HRP–streptavidinonjugate was the same except substituting anti-TSH MAb withtreptavidin.

.3.5. Immunoassay proceduresThe schematic illustration of the proposed BSAS CLEIA

s presented in Fig. 1. Calibrators or samples with a volume

apBl

Fig. 1. Schematic illustration of the proposed BSAS CLEIA.

f 50 �L were added together with 50 �L of the mixture ofhe HRP–streptavidin conjugate and biotinylated MAb. TheRP–streptavidin conjugate and the biotinylated MAb wererst mixed with the same volume. After incubation at 37 ◦Cor 60 min, the microplate was washed by washing solution toeparate the free compounds from the bound immunocomplex.00 �L chemiluminescent substrate was pipetted into each well.fter 5 min incubation at room temperature, the light intensityas measured. The working scheme of the HRP–MAb conjugateased CLEIA was the same except that HRP-labeled anti-TSHAb was used instead of mixture of the HRP–streptavidin con-

ugate and the biotinylated MAb.

.3.6. Procedure for quantitation of labeling ratio of molesf biotinylated MAb

The procedure was performed according to the kit introduc-ion. The reagent was first equilibrated to room temperature.olution containing the biotinylated MAb was added into theixture of 4-hydroxyazobenzene-2-carboxylic acid (HABA)

nd avidin. Both the absorbance of the mixture of HABA/avidinnd the mixture added with biotinylated MAb was measuredt 500 nm. Then calculation was performed based on the kitntroduction.

.4. Data analysis

The chemiluminescence intensity was expressed by the RLUrelative light unit). Standard curves were got by plotting theogarithm of chemiluminescence intensity against the logarithmf corresponding sample concentration.

. Results and discussion

.1. Principle of the BSAS based CLEIA

The proposed method was a noncompetitive assay that used

n excess of anti-TSH MAb coated on the solid phase. Therinciple of the BSAS based CLEIA is presented in Fig. 2.iotin was first covalently coupled to the anti-TSH MAb. HRP-

abeled streptavidin was synthesized too. After a sandwiched

968 Z. Lin et al. / Talanta 75 (2008) 965–972

petit

iMHerswrsfcCetT

3B

iktcd

3c

tTstt

wskrf

3a

ll

Fig. 2. The principle of the noncom

mmunocomplex was formed among the capture anti-TSHAb, TSH antigen, the biotinylated anti-TSH MAb, and theRP–streptavidin conjugate, numbers of active biotins and

nzymes were conjugated to per antibody and per streptavidin,espectively, enabling more enzyme molecules catalyzing theubstrate to emit photons [30,31]. Thus sensitivity improvementas obtained through signal amplification. Then the unbound

eagents were removed by a washing step. Subsequently, sub-trate was added and the light intensity was measured. It wasound that the light intensity was in proportional to the TSHoncentrations. The principle of the HRP–MAb conjugate basedLEIA was the same as the described BSAS based CLEIAxcept that the sandwiched immunocomplex was formed amonghe capture anti-TSH MAb, TSH antigen and HRP-labeled anti-SH MAb.

.2. Parameters affecting the sensitivity of the proposedSAS CLEIA

The sensitivity of an immunoassay depends on both method

tself and a series of experimental parameters, which include theinetics of the immunoreaction and chemiluminescent reaction,he preparation and the concentration of immunoreagents, theondition of solid phase as well as some experimental proce-ures. These parameters were discussed as following.

Hiatb

ive BSAS based CLEIA for TSH.

.2.1. The kinetics of immunoreaction andhemiluminescent reaction

The Effect of immunoreaction and chemiluminescent reac-ion time were studied to improve the immunoassay sensitivity.aking both CL intensity and signal to noise ratio into con-ideration, the immunoreaction reached an equilibrium whenhe incubation time was longer than 60 min. RLU leveled offhereafter. Hence 60 min was chosen.

HRP has been widely used as the labeled enzyme in CLEIAith the development of substrates (luminol, iso-luminol) and

eries of chemiluminescent enhancers [32,33]. The overallinetic of light emission was studied in the experiment. RLUeached maximum when the incubation time was 5 min. There-ore, incubation time of 5 min was selected.

.2.2. Effect of ratio of biotin to anti-TSH MAb on thentibody biotinylation process

It is known that excess biotinylation may inactivate a bio-ogical molecule and meanwhile, insufficient biotinylation mayead to weak interaction and therefore, low amplification effect.ence, the effect of various biotinylation on the conjugate activ-

ty was studied. The results showed that using the biotin withn amount of 150- or 100-fold molar excess against MAb inac-ivated the biological activity of the antibody obviously. Thusiotin with an amount of 40-fold molar excess against MAb

Z. Lin et al. / Talanta 75 (2008) 965–972 969

Table 1Effect of different blocking conditionsa

Blocking reagent

Phosphate buffer spiked with BSA and gelatin Phosphate buffer spiked with BSA

Blocking condition 4 ◦C (overnight) 37 ◦C (1 h) Room temperature (3 h) Room temperature (3 h)RR

cc

3

ds

b0tiwwo

ibwiRpstF4s

3d

bbrMja

ocooitrsc

AHt

3

emHtThe effect of the mouse sera with concentrations from 0.1 to0.5% (V/V) was investigated. In the whole tested range, the lin-ear coefficient showed little changes and all were higher than0.9985. When the mouse serum concentration was higher than

Fig. 3. Optimization of the dilution level of the biotinylated MAb and the

LU(S0) 836 1501LU(S5)/RLU(S0) 677 320

a All the calibrators were analyzed using 2-well replicated during three runs.

oncentration was selected, and was used in the subsequentondition optimization testing.

.2.3. Effect of solid immobilization conditionsThe sensitivity of an assay depends on the solid phase con-

ition to a great extent. Ideal solid phase should be consistent,table and has low background signal.

The buffers frequently used for coating are 0.05 mol L−1 car-onate buffer (pH 9.6), 0.01 mol L−1 phosphate (pH 7.4), and.01 mol L−1 citric buffer (pH 4.8) [34]. MAb was diluted withhese buffers for immobilization to investigate the effect of coat-ng buffer. When 0.05 mol L−1 carbonate buffer was used, RLUas the highest among the three aforementioned coating buffersith the same MAb concentration, exhibiting maximum amountf MAb immobilization.

BSA and hydrolyzed gelatin were commonly used as block-ng reagents to block the free sites on a solid phase. Two kinds oflocking reagents (one was mixture of BSA and gelatin, the otheras BSA only) were used as the blocking buffer. The block-

ng buffer without gelatin had a higher background signal withLU of 2700, resulting in worse sensitivity for the assay. Hence,hosphate buffer spiked with BSA and gelatin was chosen as theuitable blocking solution. The effect of coating and blockingime and temperature on the immunoassay was also studied.inally, microplate was coated at 4 ◦C overnight, blocked at◦C overnight or 3 h at room temperature with phosphate buffer

piked with BSA and gelatin (Table 1).

.2.4. Effect of biotinylated MAb concentration andilution level of HRP–streptavidin conjugate

The binding of HRP to MAb is through the interactionetween streptavidin and biotin. Streptavidin possesses fourinding sites per molecular for biotin. So the relative molaratio of the HRP–streptavidin conjugate to biotinylated anti-TSH

Ab is of great importance. Excessive HRP–streptavidin con-ugate or biotinylated anti-TSH MAb may lead to non-specificbsorption, deteriorating the assay sensitivity.

Concentrations of biotinylated MAb and dilution levelsf HRP–streptavidin conjugate were determined using thealibrators. The signal to background ratio was a functionf both biotinylated MAb concentration and dilution levelsf HRP–streptavidin conjugate. As shown in Fig. 3A, RLUncreased with the concentration of biotinylated MAb. When

he concentration of biotinylated MAb was 1.25 �g mL−1, RLUeached its peak value and the signal to background ratio alsotayed a plateau. Therefore, the following experiment was exe-uted with concentration of 1.25 �g mL−1 for biotinylated MAb.

Hmccc

723 2700768 232

s can be seen from Fig. 3B, the dilution level of 1:3000 forRP–streptavidin conjugate was chosen as taking both the signal

o background ratio and RLU into consideration.

.2.5. Effect of heterophile antibodiesThe human anti-mouse antibodies, which belong to het-

rophile antibody, may frequently bind to Fc fragments ofurine IgGs, resulting in abnormal TSH concentration [35].erein, mouse serum was added into the antibody dilution solu-

ion and enzyme solution to “neutralize” anti-mouse antibodies.

RP–streptavidin conjugate. The signal to background ratio is the chemilu-inescent intensity ratio obtained between the two calibrators with the TSH

oncentration of 40 and 0 mU L−1. (A) Optimization of the biotinylated MAboncentration; (B) Optimization of the dilution level of the HRP–streptavidinonjugate. The immunoreaction time was 1 h.

970 Z. Lin et al. / Talanta 75 (2008) 965–972

F

0p

3

ssawRarT

3

3

rTlbd

3

lmte(td

3

mtp(w

m20tm

3H

3

tobutlTi

C4b

3

haewoz[it(

ig. 4. Effect of the substrate volume on the chemiluminescent immunoassay.

.5%, RLU decreased. Hence, the subsequent experiment waserformed with 0.5% mouse serum.

.2.6. Effect of substrate volumeIt was reported that the substrate sensitivity referred to the

ignal intensity produced by a unit of enzyme activity [36]. Theubstrate volume was directly related to the light intensity as wells the sensitivity of an assay. Therefore, the volumes of substrateere studied from 10 to 150 �L. As can be seen from Fig. 4,LU increased with increasing substrate volume up to 50 �Lnd leveled off thereafter. The value of the signal to backgroundatio presented the highest at the substrate volume of 100 �L.herefore, substrate volume of 100 �L was selected.

.3. Method evaluation

.3.1. Standard curve and stability of the calibratorsStandard curve obtained was y = 3.68 + 1.42x, with linear

ange of 0.1–40 mU L−1 and correlation coefficient of 0.9984.he calibrators were stored at 37 ◦C for 3 and 6 days and the

ight intensity was measured. The light intensity of these cali-rators after the investigation showed little change after 3 or 6ays, indicating enough stability of the calibrators.

.3.2. Sample dilutionSerial dilution of high TSH sample with human serum yielded

inear dilution profiles. It was an important evidence to check theajor interference factors and the matrix differences between

he standard matrix and human samples. The measured andxpected values were subjected to linear functional analysisFig. 5). The correlation coefficient was 0.9992, proving thathe dilution characteristic was satisfactory down to the assay’setection limit.

.3.3. SpecificityCross-reactivity is of special importance with respect to post-

enopausal women and some outpatients who receive chemical

reatment and have an abnormal concentration of pituitary glyco-rotein hormones. 500 U L−1 of human chorionic gonadotropinHCG) gave TSH reading of 0 mU L−1. Furthermore, pregnantomen have TSH value in the normal range with the proposed

rwba

Fig. 5. Linear dilution profiles.

ethod. 100 U L−1 of follicle stimulating hormone (FSH) and00 mU L−1 of luteinizing hormone (LH) gave TSH reading of.05 and 0.0 mU L−1, respectively. These showed the satisfac-ory specificity of the developed method due to the use of paired

onoclonal antibodies [37].

.4. Comparison between BSAS based CLEIA andRP–MAb based CLEIA

.4.1. RLU comparisonAvidin has an extraordinary affinity for biotin. The interac-

ion was rapid and was unaffected by most extreme pH and somerganic solvent. Hence, HRP–streptavidin can bind tightly withiotin-labeled antibody. The EZTM biotin quantitation kit wassed to quantitate biotinylation degree. The result showed thathe average biotinylation degree was 3.12, meaning that biotiny-ated antibody bound with several number of HRP–streptavidin.he more the HRP linked to per antibody, the higher the CL

ntensity was.RLU of the BSAS based CLEIA and HRP–MAb based

LEIA was compared using the calibrators (0.5, 2.5, 10, and0 mU L−1) and signal amplification was observed by BSASased CLEIA for all these tested calibrators (Fig. 6).

.4.2. SensitivitySensitivity is diagnostically useful in hyperthyroidism and

ypothalamic or pituitary hypothyroidism. Sensitivity is alson important factor that affects the reproducibility of an assay,specially for the samples with low TSH levels. Sensitivityas constructed with multiple measurements of zero pointf the standard curve and was defined as RLU signals forero point adding twice the standard deviation of the point38] (LOD = S0 + 2S.D.). Sensitivity calculated from the exper-mental result was 0.01 mU L−1 which was 10-fold higherhan the HRP-labeled MAb based TSH measurement system0.1 mU L−1).

The sensitivity of the method was also compared with other

eported CLEIA using HRP or alkaline phosphatase (ALP),hich was listed in Table 2. Higher sensitivity was achievedy the proposed method, suggesting the advantages of BSASmplification effect.

Z. Lin et al. / Talanta 75 (2008) 965–972 971

Fbf

3

sAoctiwb

Table 2Sensitivity comparison

Label Sensitivity (mU L−1) References

ALP 4.50 [7]HRP 0.06 [8]HH

fib

3

moeb

3

mw

TI

I

I

TR

B

H

ig. 6. Light comparison between the BSAS based CLEIA and the HRP–MAbased CLEIA. The dilution levels of the MAb immobilized on the solid phaseor the two methods were 1:1000.

.4.3. PrecisionSamples with various TSH concentrations were prepared by

piking different amounts of TSH to normal human serum pools.series of samples were measured eight times within one run to

btain the intra-assay precision. Inter-assay precision was cal-ulated by comparing value for a series of samples included in

hree assay runs. The results were presented in Table 3. Thentra-assay coefficients of variation for the BSAS based CLEIAere in all cases lower than 4.3% and that of inter-assay wereelow 10%. Meanwhile, the intra-assay and inter-assay coef-

cScd

able 3ntra-assay and inter-assay precision

Sample Me

ntra-assay (n = 8)BSAS based CLEIA 1 0.

2 2.3 9.

HRP–MAb based CLEIA 1 0.2 2.3 9.

nter-assay (n = 24)BSAS based CLEIA 1 0.

2 2.3 10.

HRP–MAb based CLEIA 1 0.2 2.3 9.

able 4ecovery of the BSAS based CLEIA and the HRP–MAb based CLEIA

Spiked TSH amount (mU L−1)

SAS based CLEIA 10.75.401.10

RP–MAb based CLEIA 20.113.1

2.10

RP 0.04 [9]RP 0.01 The proposed method

cients of variation for HRP–MAb based CLEIA were alsoelow 10%.

.4.4. RecoveryRecovery test is a useful approach to check the accuracy of a

ethod. Human serum pools were spiked with different amountsf TSH and the recoveries were measured (Table 4). The recov-ries were in the range of 98.3–102 and 101–114% for the BSASased CLEIA and the HRP–MAb based CLEIA, respectively.

.5. Serum sample analysis

The concentration of TSH in 40 serum samples was deter-ined with the developed BSAS based CLEIA and comparedith that obtained by a commercial chemiluminescent kit. Good

orrelation was obtained with a coefficient of 0.9842 (Fig. 7).everal patients samples with very low TSH concentrations thatould not be detected by the HRP–MAb based CLEIA showedetectable values by the proposed method, proving that the

an (mU L−1) Standard deviation C.V. (%)

47 0.02 4.357 0.05 2.094 0.14 1.4

54 0.03 5.624 0.13 5.823 0.21 2.2

45 0.04 8.970 0.17 6.31 0.23 2.3

51 0.05 9.824 0.14 6.346 0.32 3.4

Recovery (%) Average recovery (%)

111 93.4 98.1 101105 104 96.2 102

95.0 110 90.0 98.3

103 102 99.1 101104 102 113 106115 110 116 114

972 Z. Lin et al. / Talanta 75

Fk

dfi

4

sTtCbHatshirfmhc

A

fgD

R

[

[[

[

[

[

[[

[

[[

[

[

[[

[

[

[

[[[[[

[[[[

Anal. Bioanal. Chem. 387 (2007) 1965.

ig. 7. Correlation between the proposed method and Monobind commercialit.

eveloped method was able to distinguish the hyperthyroidismrom euthyroid samples, which were suitable for the daily clin-cal practice.

. Conclusions

The objective of the present work is to develop a highlyensitive CLEIA based on biotin–streptavidin system for TSH.he proposed method comprises MAb immobilized solid phase,

he biotinylated MAb, the HRP–streptavidin conjugate and aL detection system. Signal amplification was accomplishedy introducing the interaction between biotinylated MAb andRP–streptavidin conjugate. Features make this assay be suit-

ble for clinical practice, including high sensitivity, rapidurnaround time, simple operation protocol, and high throughputcreening. The sensitivity of the proposed method was 10-foldigher than the HRP–MAb based CLEIA. The intra-assay andnter-assay coefficients of variation were below 4.3 and 10%,espectively. The accuracy examination gave recoveries rangedrom 98.3 to 102%. The correlation study between the proposedethod and commercial kit was satisfied. The proposed method

as exhibited great potentiality in the daily clinical practice andould act as a good tool for TSH analysis.

cknowledgements

The authors gratefully acknowledge the financial supportrom the National Basic Research Program of China (973 Pro-ram, No. 2007CB714507) and National Key Technology R &Program (2006AA02Z4A8).

eferences

[1] D.A. Fisher, Clin. Chem. 42 (1996) 135.

[

[

(2008) 965–972

[2] A.E. Pekary, J.M. Hershman, Clin. Chem. 32 (1986) 1009.[3] M.R. Hopton, J.S. Harrop, Clin. Chem. 32 (1986) 691.[4] C.M. Dayan, Lancet 357 (2001) 619.[5] L.E. Braverman, Clin. Chem. 42 (1996) 174.[6] Y.C. Tseng, K.D. Burman, J.R. Baker Jr., L. Wartofsky, Clin. Chem. 31

(1985) 1131.[7] I. Bronstein, J.C. Voyta, G.H. Thorpe, L.J. Kricka, G. Armstrong, Clin.

Chem. 37 (1989) 1441.[8] M.C. Sheppard, E.G. Blac, Clin. Chem. 33 (1987) 179.[9] A.J. Pamham, I.F. Tarbit, Clin. Chem. 33 (1987) 1421.10] M. Roden, P. Nowotny, U. Hollenstein, B. Schneider, H. Vierhapper, W.

Waldhausl, Clin. Chem. 39 (1993) 544.11] D.J. Berty, P.M.S. Clark, C.P. Price, Clin. Chem. 34 (1988) 2087.12] M.P. Bounaud, J.Y. Bounaud, M.H. Bouln-Plneau, L. Orget, F. Begon, Clin.

Chem. 33 (1987) 2096.13] D.S. Ross, G.H. Daniels, D. Gouveia, J. Clin. Endocrinol. Metab. 71 (1990)

764.14] C.A. Spencer, J.S. LoPresti, A. Patel, R.B. Guttler, A. Eigen, D. Shen, D.

Gray, J.T. Nicoloff, J. Clin. Endocrinol. Metab. 70 (1990) 453.15] A. Papanastasion-Diamandi, T.K. Christopoulos, E.P. Diamandis, Clin.

Chem. 38 (1992) 545.16] L.A. Frank, A.I. Petunin, E.S. Vysotski, Anal. Biochem. 325 (2004) 240.17] D.S. Sgoutas, T.E. Tuten, A.A. Verras, A. Love, E.G. Barton, Clin. Chem.

41 (1995) 1637.18] T.A. Wilkins, G. Brouwers, J.C. Mareschal, C.L. Cambiaso, Clin. Chem.

34 (1988) 1749.19] A.E. Pekary, J.M. Hershman, Clin. Chem. 30 (1984) 1213.20] E.F. Ullman, H. Kirakossian, A.C. Switchenko, J. Ishkanian, M. Ericson,

C.A. Wartchow, M. Pirio, J. Pease, B.R. Irvin, S. Singh, R. Singh, R. Patel,A. Dafforn, D. Davalian, C. Skold, N. Kurn, D.B. Wanger, Clin. Chem. 42(1996) 1518.

21] I. Weeks, M. Sturgess, K. Siddle, M.K. Jones, J.S. Woodhead, Clin.Endocrinol. 20 (1984) 489.

22] J.L. Guesdon, T. Ternynck, S. Avrames, J. Histochem. Cytochem. 27 (1979)1131.

23] G.C. Moriarty, G. Unabia, J. Histochem. Cytochem. 30 (1982) 713.24] R.B. Zhang, Y. Huang, C.P. Li, Y.B. Cui, World J. Gastroenterol. 10 (2004)

1369.25] A. Yu, T.T. Geng, Q. Fu, C. Chen, Y. Cui, J. Magn. Magn. Mater. 311 (2007)

421.26] L.V. Ding, Z.E. Selvanayagam, P. Gopalakrishnakone, K.H. Eng, J.

Immunol. Methods 260 (2002) 125.27] J.T. He, Z.H. Shi, J. Yan, M.P. Zhao, Z.Q. Guo, W.B. Chang, Talanta 65

(2005) 621.28] Y. Matsumoto, H. Kuramitz, S. Itoh, S. Tanaka, Anal. Sci. 21 (2005) 319.29] X.C. Zhang, Y.R. Lu, Mod. Prev. Med. 28 (2001) 485.30] E.P. Diamandis, T.K. Christopoulos, Clin. Chem. 37 (1991) 625.31] E.P. Diamandis, Clin. Chem. 35 (1989) 491.32] G.H.G. Thorpe, L.J. Kricka, S.B. Moseley, T.P. Whitehead, Clin. Chem.

31 (1985) 1335.33] Y. Dotsikas, Y.L. Loukas, Talanta 71 (2007) 906.34] L.X. Zhao, J.-M. Lin, Z.J. Li, Anal. Chim. Acta 118 (2005) 177.35] N. Despres, A.M. Grant, Clin. Chem. 44 (1998) 440.36] W. Dungchai, W. Siangproh, J.-M. Lin, O. Chailapakul, S. Lin, X.T. Ying,

37] Y. Iijima, Y. Endo, N. Hata, H. Fujita, M. Unoki, K. Miyal, Clin. Chem. 34(1988) 98.

38] K.V. Waite, G.F. Maberly, G. Ma, C.J. Eastman, Clin. Chem. 32 (1986)1966.