development of new methodologies for plasma mass...

Development of new methodologies for biomolecules determination based on the use of immunoassays and inductively coupled plasma mass spectrometry detection

Emma Pérez Hernández

https://www.ua.es/

http://www.eltallerdigital.com/es/index.html

Departamento de Química Analítica, Nutrición y Bromatología

Facultad de Ciencias

Development of new methodologies for biomolecules

determination based on the use of immunoassays and

inductively coupled plasma mass spectrometry

detection

Emma Pérez Hernández

Tesis presentada para aspirar al grado de

DOCTOR/DOCTORA POR LA UNIVERSIDAD DE ALICANTE

MENCIÓN DE DOCTOR/DOCTORA INTERNACIONAL

DOCTORADO en CIENCIAS EXPERIMENTALES Y BIOSANITARIAS

Dirigida por:

Juan Mora Pastor Guillermo Grindlay Lledó

Catedrático de Universidad Profesor Contratado Doctor

Departamento de Química Analítica,

Nutrición y Bromatología



JUAN MORA PASTOR, Catedrático de Universidad del Departamento de

Química Analítica, Nutrición y Bromatología y GUILLERMO GRINDLAY

LLEDÓ, Profesor contratado doctor del Departamento de Química Analítica,

Nutrición y Bromatología, ambos de la Universidad de Alicante.

CERTIFICAN QUE:

Dña. Emma Pérez Hernández, Licenciada en Química, ha realizado en

el grupo de investigación de Espectrometría Atómica Analítica perteneciente al

Departamento de Química Analítica, Nutrición y Bromatología de la Universidad

de Alicante, bajo nuestra dirección, el trabajo que lleva por título:

“Development of new methodologies for biomolecules determination

based on the use of immunoassays and inductively coupled plasma mass

spectrometry detection”, que constituye su memoria para aspirar al grado de

doctora, reuniendo, a nuestro juicio, las condiciones necesarias para ser

presentada y defendida ante el tribunal correspondiente.

Y para que conste a los efectos oportunos, en cumplimiento de la

legislación vigente, firmamos el presente certificado en Alicante a 11 de enero

de 2018.

Juan Mora Pastor Guillermo Grindlay Lledó

Catedrático de Universidad Profesor Contratado Doctor





To my family

Índices

III

ÍNDICE GENERAL

RESUMEN/SUMMARY

Resumen ............................................................................................................ 1

Summary ............................................................................................................ 3

OBJETIVOS Y ESTRUCTURA GENERAL DE LA TESIS DOCTORAL.

1. Objetivo de la Tesis Doctoral. ..................................................................... 9

2. Estructura de la Tesis Doctoral. .................................................................. 9

CAPÍTULO 1. Introducción general.

1. Introducción. .............................................................................................. 17

2. Conceptos básicos de inmunoensayos. .................................................... 19

2.1. Anticuerpos. ......................................................................................................... 19

2.2. Tipos de inmunoensayos. ................................................................................. 22

2.2.1. Inmunoensayos no-competitivos. .............................................................. 22

2.2.2. Inmunoensayos competitivos. ................................................................... 23

2.2.3. Immunoensayos en fase heterogénea y homogénea. ................................ 25

3. La Espectrometría de Masas de Plasma de Acoplamiento Inductivo como

herramienta de cuantificación de biomoléculas. ............................................... 26

3.1. Conceptos básicos de ICP-MS. ........................................................................ 29

3.2. Estrategias analíticas para la cuantificación de biomoléculas mediante ICP-

MS………….. ............................................................................................................ 31

3.3. Limitaciones de los inmunoensayos con detección mediante ICP-MS. .......... 37

4. Referencias. ................................................................................................. 38

CAPÍTULO 2. Determination of aflatoxin M1 in milk samples by means of

an inductively coupled plasma mass spectrometry-based immunoassay.

1. Introduction. .................................................................................................. 49

2. Experimental. ............................................................................................... 51

IV

2.1. Reagents and materials. ...................................................................................... 51

2.2. Buffers and solutions. .......................................................................................... 52

2.3. Immunoassay procedure. ..................................................................................... 53

2.4. ICP-MS instrumentation. ..................................................................................... 55

2.5. Calibration. .......................................................................................................... 57

2.6. Samples. ............................................................................................................... 58

2.7. Sample preparation. ............................................................................................. 58

3. Results and discussion. ................................................................................ 59

3.1. Immunoassay optimization. ................................................................................. 59

3.2. Method validation. ............................................................................................... 64

3.3. Comparison with other methodologies. ............................................................... 66

4. Conclusions. ................................................................................................. 69

5. References. .................................................................................................. 70

Supplementary data. ........................................................................................ 77

CAPÍTULO 3. Evaluation of different competitive immunoassay for

aflatoxin M1 determination in milk samples by means of inductively

coupled plasma mass spectrometry.

1. Introduction. .............................................................................................. 81

2. Experimental. ............................................................................................... 83

2.1. Reagents and materials. ....................................................................................... 83

2.2. Buffers and solutions. .......................................................................................... 84

2.3. Immunoassay procedures .................................................................................... 85

2.3.1 Antibody binding inhibition assay (ABIA). .................................................... 87

2.3.2 Capture inhibition assay (CIA). .................................................................... 88

2.3.3. Capture bridge inhibition assay (CBIA). ...................................................... 88

2.4. Instrumentation. ................................................................................................... 89

2.5. Calibration. .......................................................................................................... 91

2.6. Samples. ............................................................................................................... 91

2.7. Sample preparation. ............................................................................................. 92

3. Results and discussion. ................................................................................ 93

V

3.1. Optimization of the immunoassay procedures. ................................................... 93

3.1.1. Antibody Binding Inhibition Assay (ABIA). .................................................. 94

3.1.2. Capture Inhibition Assay. ............................................................................. 98

3.2. Aflatoxin M1 analysis in milk samples. ............................................................. 102

3.3. Comparison of different competitive immunoassay formats for AFM1

determination. ........................................................................................................... 105

4. Conclusions. ............................................................................................... 107

5. References. ................................................................................................ 109

CAPÍTULO 4. A sensitive size-exclusion inductively coupled plasma mass

spectrometry multiplexed assay for cancer biomarkers using antibodies

conjugated with a lanthanide-labelled polymer.

1. Introduction. ............................................................................................ 117

2. Experimental. .......................................................................................... 119

2.1. Reagents and materials. ................................................................................. 119

2.2. Buffers. .......................................................................................................... 121

2.3. Serum samples. .............................................................................................. 121

2.4. Instrumentation. ............................................................................................. 121

2.4.1. Size Exclusion Chromatography. ........................................................... 121

2.4.2. Inductively Coupled Plasma Mass Spectrometry. .................................. 122

2.5. Antibody labelling procedure. ....................................................................... 123

2.5.1. Partial reduction of the antibody. .............................................................. 124

2.5.2. Antibody labelling via the polymer labelling kit. ................................... 124

2.5.3. Antibody labelling via DOTA-chelate complexes. .................................. 125

2.6. Determination of the antibody labelling degree. .............................................. 125

2.6.1. Protein quantification. ................................................................................ 125

2.6.2. ICP-MS analysis of metal content. ............................................................. 125

2.7. Immunoassay procedure. ................................................................................... 126

3. Results and discussion. .............................................................................. 126

3.1. Preliminary studies with lanthanide-labelled polymer in SEC-ICPMS. ........... 126

VI

3.2. Analysis of cancer biomarkers in human serum by m eans o f S EC-ICPMS a nd

polymer-labelled antibodies. .................................................................................... 131

3.2.1. Optimization of polymer-labelled antibodies synthesis. ............................. 131

3.2.2. Influence of the incubation medium on immunocomplex formation........... 134

3.3.3. Optimization of the concentration of the polymer-labelled antibody. ........ 137

3.3.4. Figures of merit. ......................................................................................... 142

3.3.5. Comparison with other methodologies. ...................................................... 144

4. Conclusions. ............................................................................................... 144

5. References. ................................................................................................ 147

CAPÍTULO 5. Conclusiones generales.

Conclusiones………………………………………………………………………..155

CAPÍTULO 6. Futuros estudios

Futuros estudios…………………………………………………………………….159

VII

ÍNDICE DE FIGURAS

Figura 1.1. Esquema de la inmunoglobulina G (IgG). Las cadenas ligera (azul

claro) y pesada (azul oscuro) están conectadas por enlaces disulfuro. Las

partes con secuencia de aminoácidos variable representan los sitios de

unión al antígeno. VH: parte variable de la cadena pesada; VL: parte

variable de la cadena ligera. Tras la digestión con papaína, el anticuerpo se

disocia en tres partes: dos fragmentos Fab (unión al antígeno) y un

fragmento Fc (región cristalizable)…………………………………………….21

Figura 1.2. Inmunoensayo no- competitivo……………………………………….23

Figura 1.3. Inmunoensayo no-competitivo para la detección de anticuerpos...24

Figura 1.4. Inmunoensayo competitivo……………………………………………25

Figura 1.5. Inmunoensayo homogéneo…………………………………………...26

Figura 1.6. Procesos que sufre la muestra al ser introducida en el seno del

plasma…………………………………………………………………………….31

Figura 1.7. Número de publicaciones en la bibliografía que emplean ICP-MS

como detector de inmunoensayos para el análisis de biomoléculas. Motor

de búsqueda: Scopus. Palabras clave: ICP-MS y proteins/elemental

tags/immunoassays (enero 2018)…………………………………….……….35

Figura 1.8. Metodología de análisis en los inmunoensayos que utilizan ICP-MS

como sistema de detección………………………………………………….…35

Figura 1.9. Ligando bifuncional empleado para marcar anticuerpos. El quelato

se une a un grupo reactivo, lo que permite la conjugación con el anticuerpo.

Se introduce un espaciador entre el quelato y el grupo reactivo para

mejorar la reactividad. M3+ es típicamente un ion lantánido…………….….37

VIII

Figure 2.1. Scheme of the competitive ICPMS-based immunoassay for AFM1

analysis in milk samples..............................................................................54

Figure 2.2. Aflatoxin M1 calibration curve using different pAb concentrations. (-

�-) 0.5 µg mL-1 pAb concentration/incubation time 1h; (-�-) 0.25 µg mL-1

pAb concentration/incubation time 2.5 h. Aflatoxin M1-BSA concentration:

0.35 ng mL-1; secondary Ab dilution factor: 1:2000; streptavidin-Au

nanoparticles conjugate concentration: 0.08 µg mL-1..................................62

Figure S2.1. Influence of the streptavidin-Au nanoparticles conjugate

concentration on the 197Au+ normalized signal and signal standard deviation

in ICP-MS. Aflatoxin M1-BSA concentration: 0.35 ng mL-1; pAb

concentration: 0.5 µg mL-1; secondary antibody dilution factor: 1:2000;

incubation time in the microtiterplate of AFM1 standards and pAb solution: 1

h…………………………………………………………………………………...78

Figure 3.1. Scheme of the different competitive immunoassays formats tested

in this work……………………………………………………………………….87

Figure 3.2. Influence of the α-AFM1 pAb concentration on the limits of detection

of AFM1 operating with different streptavidin-Au nanoparticles conjugates.

(�) 40 nm (�) 80 nm. Aflatoxin M1-BSA concentration: 0.35 ng L-1;

secondary Ab concentration: 500 ng mL-1; streptavidin-nanoparticle

conjugate concentration: 0.08 µg mL-1.………………………………...……..96

Figure 3.3. Aflatoxin M1 calibration curve using different streptavidin-Au

nanoparticles conjugate for the CBIA procedure. (-�-) 40 nm; (-�-) 80 nm.

Aflatoxin M1-BSA concentration: 1.0 ng mL-1; streptavidin-Au nanoparticles

conjugate concentration: 0.08 µg mL-1………………………..……………..102

IX

Figure 4.1. SEC-ICPMS chromatograms of a goat polyclonal antimouse IgG

antibody (pAb) labelled with 165Ho polymer reagents (black line) and 165Ho

DOTA chelate complex (red line). pAb nominal concentration: 10 µg mL-1,

column: Superose 6 Increase 10/300 GL……………………………………127

Figure 4.2. SEC-ICPMS chromatograms obtained after incubation of a mouse

IgG1 antibody solution with a goat polyclonal antimouse IgG antibody (pAb)

labelled with (A) 165Ho polymer reagents and (B) 165Ho DOTA chelate

complex. (1) High molecular weight immunocomplex; (2) low molecular

weight immunocomplex; (3) unreacted labelled pAb; (4) free lanthanide

label. pAb nominal concentration: 10 µg mL-1; antigen concentration: 10 µg

mL-1; incubation medium: 100 mM ammonium acetate; column: Superose 6

Increase 10/300 GL………………………………………………………..…..129

Figure 4.3. SEC-ICPMS chromatograms obtained after incubation of a human

serum sample spiked with 50 ng mL-1 CEA and with its corresponding 165Ho

polymer labelled mAb at a nominal concentration of: (A) 6 ng mL-1 or (B) 2

µg mL-1. Column: Superose 6 Increase 10/300 GL………………………...138


serum sample spiked with 50 ng mL-1 of sErbB2, CA 15.3 or CA 125

antigen with its corresponding polymer-labelled antibody at a nominal

concentration of 2 µg mL-1. Column: Superose 6 Increase 10/300

GL………………….....................................................................................141

XI

ÍNDICE DE TABLAS

Tabla 1.1. Comparativa entre los diferentes tipos de inmunoensayos……...…30

Table 2.1. Operating conditions employed in ICP-MS…………………………...55

Table 2.2. Aflatoxin M1 recovery assay for different kinds of milk

samples……………………………………………………………………….….65

Table 2.3. Comparison of diverse analytical methodologies proposed in the

literature for AFM1 determination in milk samples.......................................68

Table S2.1. Results of the checkboard titration experiments to optimize AFM1-

BSA and pAb concentration. Secondary Ab dilution factor: 1:2000;

streptavidin-Au nanoparticles conjugate concentration: 0.16 µg mL-1;

incubation time in the microtiterplate of AFM1 standards and pAb solution: 1

h…………………………………………………………………………………...77

Table 3.1. Operating conditions employed in ICP-MS………………………..….90

Table 3.2. Influence of the streptavidin-nanoparticles conjugate on ABIA

optimum experimental conditions, LoDs and dynamic range………………95

Table 3.3. Optimum experimental conditions, limits of detection and dynamic

range for CIA and CBIA formats…………………………………..…………...99

Table 3.4. Recovery analysis of AFM1 by ABIA and CBIA methodologies.

Sample pretratment: ABIA: dilution + acetonitrile extraction; CBIA:

immunocolumns…………………………………………………..……………104

Table 3.5. Comparison of diverse analytical methodologies propose in the

literature for AFM1 determination in milk samples…………………….……108

Table 4.1. Operating conditions of SEC-ICPMS………………………………..123

XII

Table 4.2. Labelling degree of the different mAbs using polymer-reagents and

DOTA-chelate complexes………………………………………………..……133

Table 4.3. Influence of the incubation medium on the HMW and LMW

immunocomplexes integrated signals obtained for CEA. Antibody nominal

concentration: 1 µg mL-1; CEA concentration: 50 ng mL-1 (mean ± t·s·n1/2, n

= 3, P = 95%).…………………………………………………………………..136

Table 4.4. Influence of the CEA concentration on the immunocomplexes

integrated signals after incubation with 165Ho polymer-labelled mAb at

nominal concentrations of 6 ng mL-1 or 2 µg mL-1. Incubation medium:

human serum. (mean ± t·s·n1/2, n = 3, P = 95%)……………...……………139

Table 4.5. Recovery values for the CEA, sErbB2, CA 15.3 and CA 125

biomakers of interest using polymer labelling kit (mean ± t·s·n1/2, n = 3, P =

95%)……………………………………………………………………………..142

Table 4.6. Sensitivity, LoDs and linear dynamic range obtained for CEA,

sErbB2, CA 15.3 and CA 125 analysis using both polymer labelling kit and

DOTA-chelate complexes *(mean ± t·s·n1/2, n = 3, P = 95%)……………..145

Table 4.7. Comparison of different methods for CEA, sErbB2, CA 15.3 and CA

125 analysis…………………………………………………………………….146

XIII

GLOSARIO DE ACRÓNIMOS Y TÉRMINOS

Término Descripción

Ab Antibody

Ag-Ab Antigen-antibody

AFB1 Aflatoxin B1

AFM1 Aflatoxin M1

AFM1-BSA Aflatoxin M1-Bovine serum albumin conjugate

CA 15.3 Cancer antigen 15.3

CA 125 Cancer antigen 125

CEA Carcinoembryonic antigen

BSA Bovine serum albumin

DOTA 1,4,7,10-tetraazacyclododecane-N,N',N'',N'''- tetraacetic acid

DMSO Dimethyl sulfoxide

DNA Deoxyribonucleic acid

DTPA Diethylentriamine-N,N,N',N'',N''-pentaacetic acid

EDTA Ethylenediaminetetraacetic acid disodium salt

ELISA Enzyme linked immunosorbent assay

ESI Electrospray ionozation

FIA Flow injection analysis

HMW High molecular weight

HPLC High-performance liquid chromatography

ICP-MS Inductively coupled plasma mass spectrometry

IgG Immunoglobulin G

IgG1 Immunoglobulin G subclass 1

XIV

LoD Limit of detection

lLoQ Lower quantification limit

LMW Low molecular weight

mAb Monoclonal antibody

MALDI Matrix-Assisted Laser Desorption/Ionization

mRNA Messenger ribonucleic acid

MS Mass spectrometry

pAb Polyclonal antibody

PBS Phosphate buffer solution

PEEK Polyether ether ketone

RIA Radioimmunoassays

RNA Ribonucleic acid

SEC Size-exclusion chromatography

sErbB2 Soluble form of human epidermal growth factor receptor 2

TCEP Tris (2-carboxyethyl) - phosphine hydrochloride

Tris Tris(hydroxymethyl)aminomethane

Tween 20 Polyethylene glycol sorbitan monolaurate

uLoQ Upper quantification limit

UV-Vis Ultraviolet–visible

Resumen/Summary

1

RESUMEN

La Espectrometría de Masas mediante ionización en Plasma Acoplado

Inductivamente (ICP-MS) es una técnica de análisis elemental que permite

cuantificar la mayor parte de elementos de la tabla periódica de forma rápida y

a niveles de ultratraza. A lo largo de la última década, y debido a sus

características analíticas, ICP-MS se ha utilizado como detector de

inmunoensayos para la determinación de bioméculas. A pesar de los avances

que se han producido, este tipo de estrategias todavía no están exentas de

inconvenientes. Entre los retos que aún están por resolver, se encuentran: (i)

desarrollo de metodologías para la determinación de haptenos (biomoléculas

de bajo peso molecular < 10 kDa), (ii) falta de criterio para seleccionar el tipo

de inmunoensayo más adecuado para una aplicación bioanalítica dada, sobre

todo cuando se trabaja con haptenos; y (iii) escaso aprovechamiento del

potencial del ICP-MS (sensibilidad, límites de detección, capacidad de análisis

multielemental, etc.). En la presente Tesis Doctoral se han desarrollado nuevos

métodos de análisis de biomoléculas con detección mediante ICP-MS tratando

de salvar los inconvenientes citados con anterioridad. Los métodos se han

aplicado al análisis de tóxinas en alimentos y de biomarcadores oncológicos en

fluidos biológicos (suero).

En primer lugar, debido a la gran preocupación existente por la seguridad

alimentaria y el efecto de la alimentación sobre la salud humana, se desarrolló

una nueva metodología para la cuantificación de la aflatoxina M1 (hapteno) en

muestras de leche que cumpla los niveles de seguridad requeridos por las

actuales políticas internacionales. Por otro lado, se compararó diferentes

Resumen/Summary

2

formatos de inmunoensayos competitivos para la determinación de esta misma

aflatoxina en leche. Además se estudió el efecto de la naturaleza y del tamaño

de las nanopartículas (es decir, Ag / Au, 80 nm / 40 nm) sobre los parámetros

analíticos. Finalmente, se presentó un nuevo enfoque con el objeto de

aumentar la sensibilidad de los inmunoensayos homogéneos con detección

mediante ICP-MS, con la introducción de un alto número de etiquetas por Ab

monoclonal mediante el empleo de polímeros como portadores de múltiples

iones lantánidos. Esta metodología se aplicó a la determinación simultánea de

4 biomarcadores de cáncer (CEA, sErbB2, CA 15.3 and CA 125) en suero

humano.

Resumen/Summary

3

SUMMARY

Inductively coupled plasma mass spectrometry (ICP-MS) is an elemental

analytical technique that permits quantify most of the elements of the Periodic

Table at ultratrace levels. Over the last decade, and due to its features, ICP-MS

has been used as a detector in immunoassays for the analysis of biomolecules.

Despite the fact that these strategies are in good state of development, there is

still a long way from exploiting the full potential of this technique. Among the

challenges which have still to be solved: (i) development of methodologies for

hapten analysis (low molecular weight biomolecules < 10 kDa); (ii) lack of

criteria to select the most suitable immunoassay format for a given bioanalytical

application; especially for hapten determination and (iii) under-utilisation of the

potential of ICP-MS (sensibility, limit of detection, multi-element capability, etc.).

In the following Ph.D. work, the line of research dealing with the developed of

new analytical methods for biomolecules analysis that allow to solve the

limitations previously reported that currently have the immunoassays with ICP-

MS detection is presented. The methods have been applied to the analysis of

toxins in food and to the analysis of oncological biomarkers in biological fluids

(serum).

Firstly, due to the great concern about food safety and its influence on

human health, a new methodology for aflatoxin M1 quantification in milk

samples at the security levels required by the current international policies with

accuracy and precision has been developed. On the other hand, different

competitive immunoassay formats have been evaluated. Again, aflatoxin M1

has been selected as model analyte to evaluate the benefits and drawbacks of

Resumen/Summary

4

the different approaches. In addition, the effect of the nanoparticle metal label

(i.e. Ag/Au; 80 nm/40 nm) on the analytical figures of merit has been studied.

Finally, for the first time an approach aiming to the increase in sensitivity of the

homogeneous immunoassay with ICP-MS detection by the introduction of a

high number of tags into a monoclonal Ab has been presented, with the use of

polymer-based lanthanide group as useful carriers of multiple lanthanide ions.

This methodology has been applied for the simultaneous determination of four

cancer biomarkers (CEA, sErbB2, CA 15.3 and CA 125) in human serum.

Objetivos y Estructura general

de la Tesis Doctoral

Objetivos y Estructura general de la Tesis Doctoral

9

1. Objetivo de la Tesis Doctoral.

La presente Tesis Doctoral tiene como objetivo general el desarrollo de

nuevas metodologías basadas en el empleo de inmunoensayos y detección

mediante espectrometría de masas mediante ionización en plasma acoplado

inductivamente para la determinación de biomoléculas. Con ello se pretende

resolver algunas de las limitaciones que presentan este tipo de estrategias

analíticas, entre otras: (i) desarrollo de nuevas metodologías para el análisis de

haptenos (i.e., compuestos de bajo peso molecular, < 10 kDa); (ii) falta de

criterio para seleccionar el inmunoensayo más adecuado para una aplicación

bioanalítica dada; y (iii) escaso aprovechamiento del potencial de la técnica

(sensibilidad, límites de detección, capacidad de análisis multielemental, etc.).

Para ello, se plantean los siguientes objetivos específicos:

- Desarrollo de nuevas metodologías para la determinación de la

aflatoxina M1 (hapteno) en leche a niveles de ultra-traza.

- Desarrollo y evaluación de diferentes formatos de inmunoensayo

competitivo para el análisis de haptenos mediante ICP-MS.

- Desarrollo de una nueva metodología basada en inmunoensayos en

fase homogénea y detección mediante ICP-MS para el análisis

multicomponente de biomoléculas usando anticuerpos conjugados con

polímeros de lantánidos.

2. Estructura de la Tesis Doctoral.

La Presente Tesis Doctoral ha sido realizada en el grupo de investigación

de Espectrometría Atómica Analítica perteneciente al Departamento de


10

Química Analítica, Nutrición y Bromatología de la Universidad de Alicante.

Además, parte de la investigación ha sido desarrollada en colaboración con el

Laboratory of Bioinorganic Analytical and Environmental Chemistry (University

of Pau and Pays de l’Adour) dirigido por el Doctor Ryszard Lobinski.

Dado que la presente Tesis Doctoral opta a la Mención Internacional, los

capítulos de la misma correspondientes a los resultados obtenidos han sido

redactados en inglés, para así, cumplir con la normativa establecida.

Esta Tesis Doctoral se encuentra estructurada en un total de seis capítulos:

Capítulo 1. Introducción general.

Este capítulo presenta los conocimientos básicos que han hecho posible el

desarrollo de la presente Tesis Doctoral. Así, en primer lugar, se describen los

principios básicos de los inmunoensayos para seguir con la exposición de las

ventajas que ofrece el empleo de ICP-MS como sistema de detección frente a

los que se vienen empleando de forma tradicional en inmunoensayos. Además,

se exponen diferentes estrategias analíticas para la cuantificación de

biomoléculas mediante inmunoensayos con detección mediante ICP-MS.

Finalmente, se describen las limitaciones actuales de los inmunoensayos con

detección mediante ICP-MS, haciendo especial hincapié en aquellas que se

van a investigar en la presente Tesis doctoral.

Capítulo 2. Determination of aflatoxin M1 in milk samples by means of an

inductively coupled plasma mass spectrometry-based immunoassay.

Este trabajo describe un nuevo método para la determinación de la

aflatoxina M1 (AFM1) en niveles de ultra-traza en leche. La detección de la


11

AFM1 se lleva a cabo mediante un inmunoensayo competitivo que utiliza

anticuerpos secundarios biotinilados y estreptavidina conjugada con

nanopartículas de Au. Después de la adición de ácido, las nanopartículas se

descomponen y la señal de Au es registrada por medio de ICP-MS. Los

resultados demuestran que, en condiciones óptimas, el límite de detección del

inmunoensayo (0.005 g kg-1) es lo suficientemente bajo como para cuantificar

la AFM1 según las normativas internacionales actuales (incluida la europea que

es la más restrictiva). Además, queda patente, que el uso del ICP-MS para el

análisis de micotoxinas en alimentos tiene un gran potencial debido a que

presenta un mayor rango dinámico, señales de fondo más bajas e

independencia de la respuesta analítica a los tiempos de incubación o

almacenamiento con respecto a los métodos de detección convencionales.

Los resultados de este capítulo han dado lugar a las siguientes

contribuciones científicas:

Publicaciones:

1. E. Pérez, P. Martínez-Peinado, F. Marco, L. Gras, J. M. Sempere, J.

Mora, G. Grindlay, Determination of aflatoxin M1 in milk samples by

means of an inductively coupled plasma mass spectrometry-based

immunoassay, Food Chemistry 230 (2017) 721-727.

Comunicaciones orales:

2. G. Grindlay, E. Pérez, F. Marco, P. Peinado, J. Mora. Analysis of

aflatoxin M1 in milk by means of ICP-MS. EWCPS 2015 – European

Winter Conference on Plasma Spectrochemistry 2015. Münster

(Germany), 22-26th February 2015.

Comunicaciones (Póster):


12

3. E. Pérez, G. Grindlay, F. Marco, P. Martínez, J. Mora. Inductively

coupled plasma mass spectrometry as a tool for aflatoxin M1 detection in

milk samples. EUROANALYSIS 2015 Conference. Bordeaux (France),

6th-10th September 2015.

Capítulo 3. Evaluation of different competitive immunoassay for aflatoxin M1

determination in milk samples by means of inductively coupled plasma mass

spectrometry.

En este trabajo, se han comparado sistemáticamente diferentes formatos

de inmunoensayos competitivos para la determinación de haptenos mediante

ICP-MS utilizando nanopartículas de diferente naturaleza y tamaño. La

aflatoxina M1 ha sido seleccionada como analito modelo para evaluar las

ventajas e inconvenientes de los diferentes formatos estudiados, debido a su

alto riesgo para la salud y a los niveles de seguridad requeridos por las

actuales políticas internacionales. De acuerdo a los resultados obtenidos, el

inmunoensayo basado en el marcaje del anticuerpo, en lugar del antígeno,

parece ser el más adecuado para el análisis de haptenos mediante ICP-MS ya

que esta estrategia es la que mejor explota la gran capacidad de detección que

ofrece ICP-MS.



Publicaciones:

1. E. Pérez, F. Marco, P. Martínez-Peinado, J. Mora, G. Grindlay.

Evaluation of different inductively coupled plasma mass spectrometry-


13

based immunoassays for the determination of aflatoxin M1 in milk. (En

redacción).

Comunicaciones orales:

2. E. Pérez, G. Grindlay, F. Marco, P. Martínez-Peinado, J. Mora.

Evaluation of different inductively coupled plasma mass spectrometry-

based immunoassays for the determination of aflatoxin M1 in milk.

EWCPS 2017 – European Winter Conference on Plasma

Spectrochemistry 2017. Sankt Anton (Austria), 19-24th February 2017.

Capítulo 4. A sensitive size-exclusion inductively coupled plasma mass

spectrometry multiplexed assay for cancer biomarkers using antibodies

conjugated with a lanthanide-labelled polymer.

Este trabajo muestra, por primera vez, que anticuerpos marcados con

polímeros de lantánido pueden emplearse para la determinación simultánea de

cuatro biomarcadores de cáncer (CEA, sErbB2, CA 15.3 y CA 125) en suero

humano mediante cromatografía de exclusión por tamaño con detección de

ICP-MS. Los polímeros de lantánido llevan 30 veces más átomos del lantánido

correspondiente que el complejo lantánido-DOTA, utilizado tradicionalmente

para este fin; lo que ha dado como resultado una disminución de 10 veces en

los límites de detección. Además, la metodología analítica desarrollada, mejora

los límites de detección e incrementa el número de muestras que pueden ser

analizadas en un período de tiempo, al mismo tiempo que reduce los costes

operativos con respecto a los kits ELISA comerciales.




14

Publicaciones:

1. E. Pérez, K. Bierla, G. Grindlay, J. Szpunar, J. Mora, R. Lobinski. A

sensitive size-exclusion inductively coupled plasma mass spectrometry

multiplexed assay for cancer biomarkers using antibodies conjugated

with a lanthanide-labelled polymer. (Submitted Analytica Chimica Acta).

Capítulo 5. Conclusiones generales.

En este capítulo se presentan las conclusiones más relevantes de las

metodologías analíticas desarrolladas en la presente Tesis Doctoral.

Capítulo 6. Futuros estudios.

En este capítulo se presentan algunos de los posibles futuros estudios que

podrían derivar del trabajo desarrollado en la presente Tesis Doctoral.

Capítulo 1

Introducción general

Capítulo 1

17

Nothing should be done

if the consequences might not be serious.

George Bernard Shaw

1. Introducción.

La cuantificación de biomoléculas es de gran importancia en una amplia

rama de campos científicos (e.g. biología, medicina, toxicología, farmacología,

etc.) ya que las funciones específicas en una célula o en un organismo están

controladas por cambios en los niveles de expresión proteica bajo diferentes

condiciones fisiológicas [1-4]. Actualmente, la herramienta principal para

cuantificar biomoléculas son los inmunoensayos, donde el analito de interés se

cuantifica mediante la unión específica que se establece entre un anticuerpo y

su correspondiente antígeno. El uso de esta técnica se inició a finales de los

años 50 cuando S.A. Berson y R. Yalow utilizaron por primera vez anticuerpos

marcados con radioisótopos (radioinmunoensayos, RIA) para la cuantificación

de insulina [5]. Aunque los métodos de RIA son fiables, precisos y altamente

sensibles, sufren de los problemas relacionados con el uso de radioisótopos

(como son: riesgo para la salud, problemas de eliminación de residuos y

estabilidad limitada) que restringen su uso a laboratorios especializados. Como

consecuencia, se han propuesto inmunoensayos basados en una detección

alternativa [6] como los ensayos de inmunoabsorción ligado a enzimas [7,8],

inmunoensayos quimioluminiscentes [9,10], inmunoensayos

electroquimioluminiscentes [11,12,13], etc. En los últimos 50 años, se han


18

aplicado con éxito a un gran número de moléculas de diferente tamaño,

propiedades químicas y físicas, y actividad biológica. Debido a su alta

sensibilidad, alta especificidad y coste, los inmunoensayos se han convertido

en métodos rutinarios para la detección y cuantificación de cientos de

moléculas tanto endógenas de organismos vivos (e.g. enzimas, proteínas,

hormonas, etc.) como exógenas (e.g., toxinas, fármacos). Sin embargo, estas

técnicas convencionales de detección de inmunoensayos a menudo presentan

baja sensibilidad, intervalo dinámico limitado y no son siempre adecuadas para

realizar análisis multicomponente [14-17].

La espectrometría de masas de plasma acoplado inductivamente (ICP-MS)

es indudablemente el instrumento comercial más sensible y más utilizado para

la determinación de una amplia gama de metales y varios no metales [18-21].

Las ventajas del ICP-MS como detector elemental incluye: bajos límites de

detección (a niveles de pg mL-1 para la mayoría de los elementos), bajos

efectos de matriz, amplios intervalos dinámicos y alta resolución espectral para

elementos e isótopos [4,22-31]. Es por todo ello que actualmente el ICP-MS se

ha convertido en una de las técnicas con mayor potencial en inmunoanálisis

desde que fuera publicado por primera vez por Zhang et al. [32].

A continuación, se presentarán algunos conceptos básicos de

inmunoensayos para facilitar la comprensión de las investigaciones realizadas

en la presente Tesis Doctoral. Además, se describirán las ventajas del empleo

de ICP-MS como sistema de detección en inmunoensayos y las estrategias

analíticas que se han empleado hasta la fecha para cuantificar biomoléculas.

Finalmente, se expondrán los retos que presentan los inmunoensayos con

Capítulo 1

19

detección mediante ICP-MS en la actualidad, haciendo hincapié en los que se

van abordar en la Tesis Doctoral.

2. Conceptos básicos de inmunoensayos.

Un inmunoensayo es un tipo de prueba bioquímica que utiliza anticuerpos

contra un analito de interés como medio para detectar la presencia de este

último. El analito al que se unen los anticuerpos se llama antígeno; aunque

este término se refiere a aquella sustancia capaz de desencadenar la

formación de anticuerpos y puede causar una respuesta inmunitaria.

Generalmente, en un inmunoensayo, los anticuerpos se inmovilizan en una

superficie sólida y al ser muy selectivos, sólo se unen a su correspondiente

antígeno, incluso en matrices muy complejas. La detección del correspondiente

inmunocoplejo entre el anticuerpo y su analito siempre requiere del marcaje de

alguna de las especies con un heteroátomo que facilite la detección.

Los inmunoensayos destacan por su especificidad, sensibilidad y

flexibilidad únicas debido a tres importantes características de los anticuerpos:

- Su capacidad para unirse a una amplia gama de productos químicos, ya

sean naturales o artificiales; biomoléculas; células y virus.

- Su elevada especificidad.

- La fuerza de la unión entre un anticuerpo y su antígeno.

2.1. Anticuerpos.

La eficiencia de cualquier reacción inmune es altamente dependiente de la

especificidad y de la afinidad del anticuerpo por su antígeno. Es importante

tener en cuenta que la afinidad del anticuerpo es el factor limitante en una


20

reacción inmune. El tipo de anticuerpo utilizado en inmunoensayos es la

inmunoglobulina G (IgG), que representa el 75% de todas las inmunoglobulinas

séricas [33]. Consiste en dos cadenas pesadas idénticas (H) de 50 kDa y dos

cadenas ligeras idénticas (L) de 24 kDa [34]. Las cadenas ligera y pesada, así

como su homólogo idéntico, están unidas por fuerzas no covalentes y enlaces

disulfuro (ver Figura 1.1). Los extremos N-terminal de cada cadena ligera y

pesada son altamente variables y representan los dos sitios de unión al

antígeno. Los extremos C-terminal son constantes en su secuencia de

aminoácidos. Si el anticuerpo se hace reaccionar con la enzima papaína, éste

se divide en tres partes: dos fragmentos Fab (fragmento de unión al antígeno)

con actividad inmunogénica y un fragmento Fc (fragmento cristalizable).

La obtención de anticuerpos se puede realizar vacunando animales

huéspedes con el antígeno contra el que se quiere producir el anticuerpo

(inmunización) o mediante biotecnología. Los anticuerpos producidos pueden

ser de 3 tipos: policlonales, monoclonales y recombinantes. Los anticuerpos

policlonales se obtienen por inmunización de animales y son relativamente

baratos en grandes cantidades; sin embargo, pueden reaccionar de forma

cruzada con compuestos estructuralmente similares. Los anticuerpos

monoclonales se obtienen mediante biotecnología combinando un linfocito B

productor del anticuerpo específico de interés con una línea celular de

mieloma. Estos anticuerpos poseen características de unión específicas, ya

que no sólo reconocen a un mismo antígeno sino que también se unen al

mismo epítopo (porción de la molécula reconocida por el anticuerpo) del

antígeno, dando lugar a ensayos mucho más específicos. El principal

inconveniente de los anticuerpos monoclonales es que son caros de producir y

Capítulo 1

21

técnicamente exigente para generar y mantener una línea celular de hibridoma.

Por último, el desarrollo de la biología molecular ha permitido la producción de

anticuerpos en ausencia de inmunización, conocidos como anticuerpos

recombinantes. Los anticuerpos recombinantes ofrecen ventajas sobre los

anteriores en términos de facilidad de producción, mayor repertorio para

selección y versatilidad.

Figura 1.1. Esquema de la inmunoglobulina G (IgG). Las cadenas ligera (azul

claro) y pesada (azul oscuro) están conectadas por enlaces disulfuro. Las

partes con secuencia de aminoácidos variable representan los sitios de unión al

antígeno. VH: parte variable de la cadena pesada; VL: parte variable de la

cadena ligera. Tras la digestión con papaína, el anticuerpo se disocia en tres

partes: dos fragmentos Fab (unión al antígeno) y un fragmento Fc (región

cristalizable).


22

2.2. Tipos de inmunoensayos.

2.2.1. Inmunoensayos no-competitivos.

El inmunoensayo más sencillo (Figura 1.2) consiste en un anticuerpo,

inmovilizado en una superficie de plástico (e.g. un pocillo de una placa de

inmunológica), que captura el antígeno presente en la muestra. A continuación,

se utiliza un anticuerpo diferente, específico de otro epítopo, como base del

sistema de detección. Este segundo anticuerpo debe estar marcado con alguna

especie que facilite la detección del inmunocomplejo formado (trazador). La

especie que permite monitorizar el inmunocomplejo suele ser una enzima. Al

igual que los anticuerpos, las enzimas son proteínas que se unen a analitos

específicos; sin embargo, las enzimas también son capaces de catalizar

reacciones específicas. La molécula sobre la que actúa una enzima se llama

sustrato. Las enzimas empleadas como marcadores, con el sustrato

apropiado, pueden usarse para generar color o crear productos finales

fluorescentes o luminiscentes, que pueden medirse fácilmente mediante

equipos ópticos y electrónicos. Además, cada molécula de enzima puede

unirse a varias moléculas de sustrato, amplificando la señal. Este formato de

ensayo a menudo se conoce como ensayo de inmunoabsorción ligado a

enzimas (ELISA). En cualquier caso, independientente del trazador empleado,

la señal que proporciona el marcador es proporcional a la concentración de

antígeno presente en la muestra. Una etapa crítica del ensayo es la eliminación

del trazador no unido durante una la etapa lavado. El material (normalmente de

plástico) al que el anticuerpo de captura se une irreversiblemente se conoce

como la fase sólida. Debido a que los anticuerpos forman un sándwich

Capítulo 1

23

alrededor del antígeno, estos ensayos inmunométricos también son conocidos

como inmunoensayos tipo sándwich.

Figura 1.2. Inmunoensayo no- competitivo.

Otro tipo de ensayo inmunométrico consiste en detectar anticuerpos en una

muestra fijando el antígeno en el soporte. Esta aplicación es útil, por ejemplo,

para detectar la exposición previa a una enfermedad infecciosa específica. Las

proteínas que se producen en la superficie de un virus se inmovilizan en la fase

sólida y son las encargadas de capturar los anticuerpos específicos, para ese

virus, presentes en la muestra. En esta estrategia, como trazador podría usarse

un anticuerpo marcado (anticuerpo secundario) contra la región constante del

primer anticuerpo (Figura 1.3).

2.2.2. Inmunoensayos competitivos.

Los ensayos inmunométricos, descritos hasta el momento, funcionan bien

cuando el antígeno es una molécula grande con suficiente área superficial para

+ + +

Fase sólida recubierta de anticuerpo

Analito Anticuerpo marcado

Separación del anticuerpo marcado libre

Concentración de analito

Se

ña

l

Respuesta lineal

Saturación


24

acomodar dos moléculas de anticuerpo. Sin embargo, muchos inmunoensayos

tienen como objetivo determinar la presencia de moléculas pequeñas (tóxicos,

hormonas, etc.) que sólo presentan un epítopo. Estas moléculas se las conoce

como haptenos y para desarrollar anticuerpos específicos contra ellas

requieren conjugarlas con alguna proteína de forma que se estimule una

respuesta inmunitaria.

Figura 1.3. Inmunoensayo no-competitivo para la detección de anticuerpos.

En este tipo de ensayos, se utiliza normalmente un anticuerpo presente en una

concentración limitada (Figura 1.4). El otro reactivo clave, el trazador, consiste

en el antígeno marcado con, por ejemplo, un radioisótopo o una enzima. La

cantidad de trazador que se une al anticuerpo es indirectamente proporcional a

la concentración de antígeno presente en la muestra. En este tipo de ensayo,

conocido como inmunoensayo de tipo competitivo, las concentraciones de

anticuerpo y trazador son críticas, a diferencia de los inmunoensayos de tipo

sándwich donde se usa un exceso de anticuerpo respecto del antígeno. Es

importante señalar que este tipo de inmunoensayo también se puede llevar a

+ + +

Fase sólida recubierta de antígeno

Anticuerpo Anticuerpo secundario

Separación del anticuerpo marcado libre

Concentración de anticuerpo

Se

ña

l

Respuesta lineal

Saturación

Capítulo 1

25

cabo fijando la molécula de analito modificada en el soporte y utilizando el

anticuerpo como trazador. En ocasiones, para evitar que el marcado del

anticuerpo altere la afinidad con el antígeno se utilizan sistemas auxiliares

como un anticuerpo secundario.

Figura 1.4. Inmunoensayo competitivo.

2.2.3. Immunoensayos en fase heterogénea y homogénea.

Todos los inmunoensayos descritos hasta el momento dependían de la

separación física del trazador no unido antes de medir la señal correspondiente

al trazador unido. Sin una etapa de separación, la intensidad de señal sería

siempre la misma, independientemente de la concentración de analito. Estos

formatos de ensayo son todos ejemplos de inmunoensayos de tipo

heterogéneo. Por el contrario, los inmunoensayos de tipo homogéneo se

basan en un cambio en la actividad del inmunorreactivo marcado que se

produce cuando el antígeno se une al anticuerpo para formar el

inmunocomplejo y no requieren ningún paso de separación (Figura 1.5).

+ + +

Fase sólida recubierta de anticuerpo

Analito Analitomarcado(trazador)

Separación del analitomarcado libre


Se

ña

l


26

Figura 1.5. Inmunoensayo homogéneo.

Aunque la etapa de separación complica el procedimiento, los inmunoensayos

de tipo heterogéneo a menudo proporcionan límites de detección superiores

respecto a los inmunoensayos de tipo homogéneo, convirtiéndolos en los

métodos más frecuentemente adoptados.

3. La Espectrometría de Masas de Plasma de Acoplamiento

Inductivo como herramienta de cuantificación de biomoléculas.

En los últimos años, el desarrollo de nuevas herramientas de bioanálisis ha

permitido avanzar enormemente en campos como la metalómica, proteómica y

metabolómica. Gracias a ello, los laboratorios de análisis clínico, por ejemplo,

pueden controlar los niveles de biomarcadores tanto endógenos (e.g.,

proteínas cuyo contenido en fluidos biológicos se pueda ver alterado por

procesos bioquímicos o enfermedades como cáncer o infecciones) como

exógenos (e.g., compuestos tóxicos o toxinas), lo que permite mejorar el

diagnóstico, tratamiento y evolución de diversas enfermedades. También

+ + +

Anticuerpo Antígeno Antígenomarcado

Generación de señal por launión del antígeno marcadoal anticuerpo (no se requiereseparación)


Se

ña

l

Capítulo 1

27

facilita los estudios farmacocinéticos de los tratamientos recibidos, lo que

permite su ajuste o el diseño de fármacos más seguros.

Muchas de las biomoléculas de interés en proteómica y metabolómica se

suelen determinar mediante métodos de inmunoensayos basados en detección

mediante absorción en UV-Vis (ELISA), fluorescencia, o quimioluminiscencia,

entre otros. Sin embargo, estas técnicas convencionales de detección de

inmunoensayos a menudo no poseen un amplio intervalo dinámico y presentan

baja sensibilidad. Además, se produce la superposición de señales, no

pudiendo ser utilizados para la cuantificación multianalítica simultánea. La

Tabla 1.1 resume una comparativa de los inmunoensayos de uso común.

El ICP-MS es indudablemente el instrumento comercial más sensible para

la determinación de una amplia gama de metales y varios no metales

[18,19,20,21]. El gran atractivo del ICP-MS viene determinado por su: (i)

elevada sensibilidad y bajos límites de detección (ng Kg-1); (ii) amplio intervalo

dinámico (iii) capacidad multielemental; (iv) información isotópica; (v) robustez;

y (vi) facilidad de acoplamiento a técnicas cromatográficas. Esto ha hecho que

actualmente se haya convertido en una de las técnicas con mayor potencial en

inmunoanálisis desde fuera publicado por primera vez por Zhang et al. [32].

Intro

du

cción

gen

eral

Tab

la 1.1. Co

mp

ara

tiva

entre

los d

ifere

nte

s tip

os d

e in

mu

no

en

sa

yos.

D

etección

del in

mu

no

ensayo

R

adio

isóto

pos

UV

-Vis

Q

uim

iolu

min

iscencia

F

luore

ccencia

E

lectro

qu

imio

lum

inis

cencia

IC

P-M

S

Año d

e

inve

nció

n

1959

[5]

1971

[7]

1976

[9]

1979

[35

] 1991

[11

] 2001

[32

]

Sensib

ilidad

A

lta

Med

ia

Alta

Alta

Alta

Alta

Cara

cte

rístic

as

del tra

zador

Radio

activo

P

rodu

zca s

eñal

colo

rimétric

a

Pro

du

zca s

eñal

quim

iolu

min

iscente

Pro

du

zca s

eñal

fluore

scente

Pro

du

zca s

eñal

ele

ctro

quim

iolu

min

iscente

Sin

requerim

iento

especia

l

Vid

a ú

til de lo

s

reactiv

os

Corta

M

ed

ia

Med

ia

Med

ia

Med

ia

Larg

a

Esta

bilid

ad

M

ed

ia

Alta

M

ed

ia

Med

ia

Med

ia

Alta

Pote

ncia

l para

multiá

na

lisis

B

ajo

B

ajo

Bajo

Bajo

Bajo

Excele

nte

Coste

de

intru

menta

ció

n y

opera

tivo

Med

io

Bajo

M

ed

io

Med

io

Med

io

Alto

Capítulo 1

29

3.1. Conceptos básicos de ICP-MS.

El ICP-MS se introdujo a principios de la década de 1980 [36], y hoy en día

se ha convertido en una poderosa técnica para la determinación de elementos

traza, minoritarios y mayoritarios en una variedad de muestras [37,38,39]. En

esta técnica, la muestra se introduce en forma de aerosol líquido en la base de

un plasma de argón; siendo este último la fuente de excitación y/o ionización

más común en Espectrometría Atómica. A diferencia de las fuentes de

ionización suave, como la ionización por electroespray (ESI) y la

desorción/ionización láser asistida por matriz (MALDI), el ICP opera a elevada

temperatura, entre 6.000 y 10.000 K [37]; esta energía es suficiente para la

desolvatización, volatilización de la muestra, atomización e ionización de los

átomos formados (Figura 1.6). Los iones generados en el plasma se introducen

en el analizador de masas, donde los iones se separan por su relación masa-

carga (m / z). Tras ello, son detectados y finalmente convertidos en una señal

eléctrica.

Figura 1.6. Procesos que sufre la muestra al ser introducida en el seno del

plasma.

Aerosol con la muestra

RecombinaciónM+ + 1e- MM+ + O MO+

DesolvataciónH2O (l) H2O (g) Vaporización

MX (s) MX (g)

AtomizaciónMX (g) M + X

Ionización M M+

Analizador de masas

Analitopresente como M+


30

En un primer momento, la aplicación de ICP-MS en bioanálisis estuvo

limitada al análisis elemental inorgánico y, dadas las limitaciones de la técnica

para la determinación de elementos como C, O, N, etc., al de aquellas especies

orgánicas que contenían en su estructura algún metal, metaloide y/o no metal.

Así, por ejemplo, se aplicó a la determinación de proteínas que contenían

cisteína o metionina a partir de la señal de S [40]. De igual forma, se aplicó a la

determinación de otras especies que contenían heteroátomos como: (i) P (e.g.

proteínas fosforiladas y ADN) [41]; (ii) Se (e.g. selenoproteinas) [42]; (iii) As

(e.g. arsenoazúcares) [43]; y (v) metales (e.g. enzimas y metaloproteínas)

[44,45]. El gran atractivo de este tipo de aplicaciones radica en que, al conocer

de antemano el número de heteroátomos por molécula, se pueden emplear las

estrategias de calibración habituales en análisis inorgánico (i.e. patrones

externos, adición de estándar y dilución isotópica). No obstante, este tipo de

análisis presenta diversas limitaciones que afectan a la exactitud de los

resultados y a los límites detección. Así, la mayor parte de los heteroátomos

presentes en las biomoléculas (e.g., P, S, Se, As, etc.) son muy poco sensibles

en ICP-MS por su elevado potencial de ionización (> 9 eV). Por otro lado, la

determinación de este tipo de elementos en matrices biológicas está afectada

por interferencias espectrales y no espectrales [40,41,42,46] que limitan sus

aplicaciones. En comparación, un analizador de sector magnético posee una

alta capacidad de resolución para distinguir los analitos de las inferencias,

incluso en muestras de matrices complejas. Sin embargo, el uso de ICP-MS de

alta resolución es limitado debido a su alto coste. Para superar las

interferencias en el cuadrupolo del ICP-MS, se desarrolló el uso de una célula

de colisión / reacción combinado con el ICP-MS [47]. Los detalles sobre su

Capítulo 1

31

desarrollo y su aplicación pueden encontrarse en la bibliografía y no se entrará

en más detalle dado que no es objeto de este estudio [37,47,48]. Finalmente es

importante indicar que el ICP y otras fuentes de ionización blandas (por

ejemplo, ESI o MALDI) son realmente técnicas complementarias [25]. La

información estructural se obtiene preferiblemente por medio de ESI o MALDI-

MS, mientras que el ICP-MS es ideal para detectar y cuantificar metales en

proteínas, incluso cuando están presentes en niveles muy bajos de

concentración.

3.2. Estrategias analíticas para la cuantificación de biomoléculas

mediante ICP-MS.

Como se mencionó anteriormente, los métodos basados en detección

mediante ICP-MS ofrecen conceptos simples de cuantificación, efectos de

matriz bajos en comparación con las técnicas bioanalíticas convencionales, y

límites de detección biológicamente relevantes (LoD) en el intervalo bajo pg g-1.

Sin embargo, la mayor parte de analitos de interés en bioanálisis no contienen

heteroátomos detectables mediante ICP-MS. Es por ello que, para tratar de

aprovechar las ventajas de la técnica en este campo se han investigado

diferentes metodologías de marcaje de biomoléculas con heteroátomos [30]. La

clave para una cuantificación fiable es que se produzca una reacción de

marcaje [28] bajo las siguientes condiciones generales:

- formación de un enlace estable entre la etiqueta y la proteína (o

escindible a propósito),

- reacción completa (cerca del 100%),


32

- reacción específica (reacción sólo con el grupo funcional objetivo, sin

reacciones laterales),

- reproducibilidad, y

- las condiciones de reacción no deberían interferir en las siguientes

etapas analíticas.

Además, las reacciones rápidas y evitar el uso de un exceso de reactivo,

debido a que implica menores señales de fondo y coste, son favorables. Así, se

han empleado métodos de derivatización utilizando: (i) I (las proteínas que

contienen histidina y tirosina pueden ser marcadas en un anillo bencénico con

este no-metal [49]); (ii) compuestos organometálicos [28] (los grupos tiol de la

cisteína pueden ser marcados con compuestos organomercúricos [50]); y (iii)

quelatos de lantánidos (DOTA/DTPA) [51] (complejos altamente estables que

pueden añadirse a las proteínas a través de los residuos tiol y amino de los

aminoácidos). El principal inconveniente de este tipo de estrategias es que, al

basarse en reacciones químicas, es difícil controlar la selectividad y el grado de

marcaje del analito en mezclas complejas. Una estrategia cuyo interés y

aplicaciones se ha extendido en los últimos años en el campo del bioanálisis

(Figura 1.7) es el marcaje indirecto de biomoléculas mediante reacciones

antígeno-anticuerpo en las que el anticuerpo se funcionaliza previamente con

un heteroátomo detectable mediante ICP-MS [52,53].

Capítulo 1

33

Figura 1.7. Número de publicaciones en la bibliografía que emplean ICP-MS

como detector de inmunoensayos para el análisis de biomoléculas. Motor de

búsqueda: Scopus. Palabras clave: ICP-MS y proteins/elemental

tags/immunoassays (enero 2018).

La Figura 1.8 muestra un esquema del modo de trabajo empleado

habitualmente para cuantificar biomoléculas mediante el uso combinado de

inmunoensayos e ICP-MS.

Figura 1.8. Metodología de análisis en los inmunoensayos que utilizan ICP-MS

como sistema de detección.

0

10

20

30

40

50

Núm

ero

de p

ublic

acio

nes

Año

BIOMOLÉCULA ICP-MS

Tipo

FormatoAnticuerpo

Heteroátomo

Nanopartículas Lantánidos

1. FUNCIONALIZAR ANTICUERPO

2. INMUNOENSAYO 3. DETECCIÓN

CompetitivoNo competitivo

Nebulizador

Señal

Ablación laser

Pulsos

Single Particle

F. HeterogéneaF. Homogénea

Otros


34

En primer lugar, se debe funcionalizar el anticuerpo (o el trazador) con un

heteroátomo a través de los enlaces disulfuro de los residuos de cisteína o los

grupos amino de los aminoácidos. Aunque sobre el papel se puede utilizar casi

cualquier elemento de la tabla periódica para funcionalizar anticuerpos, la gran

parte de los estudios realizados hasta el momento se centran en el uso de

nanopartículas de Au y/o Ag así como en quelatos de lantánidos. Estos últimos

consisten en un grupo reactivo, un enlazador o espaciador y un quelato, tal y

como se muestra esquemáticamente en la Figura 1.9. Estos ligandos

bifuncionales a menudo se unen covalentemente a grupos -amino de residuos

de lisina y a extremos N-terminales de grupos -amino a pH alcalinos. Los

residuos de isotiocianatobencilo (SCN) y éster de N-hidroxisuccinimida (NHS)

son frecuentemente los grupos reactivos utilizados para este tipo de marcaje

químico. Alternativamente, los residuos de maleimidoetilacetamida (maleimido)

se conjugan a los residuos de sulfhidrilo después de una reducción selectiva de

los puentes disulfuro de los residuos de cisteína del anticuerpo. Esta etapa de

reducción es crítica, ya que puede afectar la eficacia de unión del anticuerpo.

Con respecto al grupo quelante, diferentes compuestos lineales o macrocíclicos

basados en el ácido poliaminocarboxílico se suelen utilizar para acomplejar el

metal. Estos grupos quelantes se seleccionan en función de la carga de este

último. Además, deben tener una alta constante de equilibrio para resistir al

intercambio metálico. Los compuestos más usados son: ácido dietilen-triamino-

tetraacético (DTTA), ácido dietilentriaminapentaacético (DTPA) y ácido

1,4,7,10-tetraazaciclododecano-1,4,7,10- tetraacético (DOTA). Los lantánidos

poseen una respuesta altamente sensible en ICP-MS debido a: (i) su baja

abundancia en la naturaleza y, en consecuencia, la señal de fondo en las

Capítulo 1

35

matrices de muestras biológicas es baja; (ii) las interferencias poliatómicas rara

vez son significativas; y, (iii) la eficacia de ionización de los lantánidos es alta

debido a que el primer potencial de ionización es bajo. Además, los lantánidos

tienen propiedades químicas similares y, por lo tanto, son muy adecuados para

desarrollar ensayos multicomponente basados en detección mediante ICP-MS.

Figura 1.9. Ligando bifuncional empleado para marcar anticuerpos. El quelato

se une a un grupo reactivo, lo que permite la conjugación con el anticuerpo. Se

introduce un espaciador entre el quelato y el grupo reactivo para mejorar la

reactividad. M3+ es típicamente un ion lantánido.

Las nanopartículas se suelen emplear en ensayos monocomponente por su

capacidad para amplificar la señal analítica, debido a que incluyen miles de

átomos por conjugado. El tamaño de partícula puede variar entre el más

pequeño de 2nm (clúster de oro) a más de 100nm (oro coloidal), dependiendo

de la aplicación. Especialmente las nanopartículas de Au son a menudo

empleadas como marcadores de anticuerpos y enzimas debido a su robustez y

fácil manejo. Las nanopartículas están conjugadas con grupos reactivos como

residuos de maleimido o NHS para el marcaje, así como también ligandos

Quelato

Espaciador

Ligando bifuncional

Grupo reactivo


36

bifuncionales. Pero hay algunos puntos críticos a tener en cuenta durante el

uso de nanopartículas de Au: el oro tiene una gran afinidad por las superficies

y, por lo tanto, las etapas de bloqueo y lavado son críticas. Además, es difícil

sintetizar nanopartículas de tamaño uniforme, las cuales son necesarias

cuando la finalidad es la cuantificación. No obstante, existen estrategias

prometedoras con lantánidos que permiten mejorar significativamente la

sensibilidad y los límites de detección (polímeros [54] y nanopartículas de

lantánidos [55]).

Una vez funcionalizado el anticuerpo con el heteroátomo, se procede a

llevar a cabo el inmunoensayo con la biomolécula de interés. Finalmente, la

última etapa del análisis es la detección del analito a través del heteroátomo

mediante ICP-MS. Normalmente, y dado que el volumen de muestra disponible

en un inmunoensayo no es muy elevado (100-200 µL), la muestra se analiza en

forma líquida con un sistema de inyección en flujo acoplado a un nebulizador y

a una cámara de nebulización [52,53]. No obstante, cuando se trabaja con

electroforesis en gel o directamente en tejidos se puede utilizar un sistema de

ablación laser. En este caso, además de estudiar la distribución de una

determinada especie, se mejora la velocidad de análisis y se evita la dilución

del analito (mayor sensibilidad). Finalmente, cabe indicar, que algunos autores

han sugerido el empleo del modo single particle cuando se trabaja con

nanopartículas [56]. En este caso, la concentración de analito está relacionada

con la frecuencia con la que se detectan los pulsos de señal de las

nanopartículas al atomizarse e ionizarse. Sin embargo, a pesar de las

potenciales ventajas que puede presentar (e.g. menores límites de detección),

este modo de medida está muy poco estudiado hasta el momento.

Capítulo 1

37

3.3. Limitaciones de los inmunoensayos con detección mediante

ICP-MS.

Tal y como acabamos de revisar, los estudios realizados hasta el momento

nos permiten afirmar que las metodologías de bioanálisis basadas en el empleo

de ICP-MS se encuentran en un buen estado de desarrollo. Sin embargo,

todavía se está muy lejos de conseguir explotar todo el potencial de la técnica,

ya que existen importantes retos aún por resolver [52,53]. Así, en lo que se

refiere al tipo de analito, la mayor parte de aplicaciones desarrolladas hasta el

momento se han centrado en la cuantificación de biomoléculas de elevado

peso molecular (proteínas, marcadores tumorales, enzimas, ADN, etc). Las

aplicaciones a biomoléculas sencillas de bajo peso molecular (haptenos; masa

molecular <10000 Da) han sido mucho más escasas ya que, al no provocar

respuesta inmunológica, es más complicado obtener anticuerpos para este tipo

de analito. Además, hace necesario el uso de inmunoensayos de tipo

competitivo [57,58] que, desde el punto de vista analítico, son claramente

menos atractivos que los ensayos tipo sándwich. De hecho, para este tipo de

analito se suelen utilizar técnicas cromatográficas acopladas a Espectrometría

de Masas en lugar de inmunoensayos. Además, en la bibliografía tampoco es

posible encontrar un estudio sistemático que compare las ventajas e

inconvenientes de los distintos formatos de inmunoensayos competitivos

comúnmente empleados para el análisis de haptenos. En cuanto al tipo de

heteroátomo elegido, las especies más empleadas por su disponibilidad

comercial y prestaciones analíticas son los quelatos de lantánidos y las

nanopartículas metálicas de Au/Ag [53]. Sin embargo, el efecto de la naturaleza

y del tamaño de las nanopartículas sobre los parámetros analíticos de ensayos


38

competitivos con detección mediante ICP-MS no ha sido estudiado en

profundidad. Otro aspecto a tener en cuenta es que la gran parte de

inmunoensayos se suele realizar en fase heterogénea ya que proporciona una

mayor capacidad de detección que los ensayos en fase homogénea. No

obstante, estos últimos presentan el gran atractivo que se pueden realizar en

una sola etapa con la consiguiente mejora en la velocidad de análisis, coste y

reproducibilidad al no utilizar placas u otro tipo de soporte. En este tipo de

ensayos, se requiere de una separación cromatográfica previa para separar el

inmunocomplejo del anticuerpo que no ha reaccionado [59]. Hasta la fecha, se

han realizado estudios en fase homogénea utilizando quelatos de lantánidos o

In [59,60,61]. Si bien los resultados son prometedores, sería interesante

evaluar el uso de polímeros de lantánidos ya que permitiría aumentar la

sensibilidad debido a la presencia de más de un lantánido por cada etiqueta en

el anticuerpo marcado. Además, a pesar de que la mayor parte de los estudios

realizados resaltan las capacidades de análisis multicomponente de ICP-MS,

en muy contadas ocasiones se aprovecha dicho potencial.

4. Referencias.

[1] M. Hamdan & P.G. Righetti. Modern strategies for protein quantification in

proteome analysis: Advantages and limitations. Mass Spectrom. Rev. 21

(2002) 287–302.

[2] S.E. Ong & M. Mann. Mass spectrometry-based proteomics turns

quantitative. Nat. Chem. Biol. 1 (2005) 252–262.

Capítulo 1

39

[3] J.S. Becker & N. Jakubowski. The synergy of elemental and biomolecular

mass spectrometry: New analytical strategies in life sciences. Chem. Soc.

Rev. 38 (2009) 1969–1983.

[4] S. Mounicou, J. Szpunar & R. Lobinski. Metallomics: The concept and

methodology. Chem. Soc. Rev. 38 (2009) 1119–1138.

[5] R.S. Yalow & S.A. Berson. Assay of plasma insulin in human subjects by

immunological methods. Nature 184 (1959) 1648–1649.

[6] D.S. Hage. Immunoassays. Anal. Chem. 71 (1999) 294R–304R.

[7] E. Engvall & P. Perlmann. Enzyme-linked immunosorbent assay (ELISA)

quantitative assay of immunoglobulin-G. Immunochemistry 8 (1971) 871-

874.

[8] R.M. Lequin. Enzyme immunoassay (EIA)/enzyme-linked immunosorbent

assay (ELISA). Clin. Chem. 51 (2005) 2415–2418.

[9] H.R. Schroeder, P.O. Vogelhut, R.J. Carrico, R.C. Bogulaski & R.T. Buckler.

Competitive protein-binding assay for biotin monitored by

chemiluminescence. Anal. Chem. 48 (1976) 1933–1937.

[10] L.X. Zhao, L. Sun & X.G. Chu. Chemiluminescence immunoassay. Trac-

Trends Anal. Chem. 28 (2009a) 404–415.

[11] G.F. Blackburn, H.P. Shah, J.H. Kenten, J. Leland, R.A. Kamin, J. Link, J.

Peterman, M.J. Powell, A. Shah, D.B. Talley, S.K. Tyagi, E. Wilkins, T.G.

Wu & R.J. Massey. Electrochemiluminescence detection for development of

immunoassays and DNA probe assays for clinical diagnostics. Clin. Chem.

37 (1991) 1534–1539.


40

[12] J.H. Kenten, J. Casadei, J. Link, S. Lupold, J. Willey, M. Powell, A. Rees &

R. Massey. Rapid electrochemiluminescence assays of polymerase

chainreaction products. Clin. Chem. 37 (1991) 1626–1632.

[13] R.J. Forster, P. Bertoncello & T.E. Keyes. Electrogenerated

chemiluminescence. Annu. Rev. Anal. Chem. 2 (2009) 359–385.

[14] M.Y. Han, X.H. Gao, J.Z. Su & S. Nie. Quantum-dot-tagged microbeads for

multiplexed optical coding of biomolecules. Nat. Biotechnol. 19 (2001) 631–

635.

[15] P.K. Chattopadhyay, D.A. Price, T.F. Harper, M.R. Betts, J. Yu, E. Gostick,

S.P. Perfetto, P. Goepfert, R.A. Koup, S.C. De Rosa, M.P. Bruchez & M.

Roederer. Quantum dot semiconductor nanocrystals for

immunophenotyping by polychromatic flow cytometry. Nat. Med. 12 (2006)

972–977.

[16] V. Biju, T. Itoh & M. Ishikawa. Delivering quantum dots to cells:

Bioconjugated quantum dots for targeted and nonspecific extracellular and

intracellular imaging. Chem. Soc. Rev. 39 (2010) 3031–3056.

[17] P. Wu, L.N. Miao, H.F. Wang, X.G. Shao & X.P. Yan. A multidimensional

sensing device for the discrimination of proteins based on

manganesedoped ZnS quantum dots. Angew Chem. Int. Ed. 50 (2011)

8118–8121.

[18] N.H. Bings, A. Bogaerts & J.A.C. Broekaert. Atomic spectroscopy: A

review. Anal. Chem. 82 (2010) 4653–4681.

[19] C.B. Zheng, L. Yang, R.E. Sturgeon & X.D. Hou. UV photochemical vapor

generation sample introduction for determination of Ni, Fe, and Se in

Capítulo 1

41

biological tissue by isotope dilution ICPMS. Anal. Chem. 82 (2010) 3899–

3904.

[20] Y. Gao, X.H. Peng, Z.M. Shi, R.X. Zhang, X.F. Xia, F.R. Yue & R. Liu.

Determination of mercury in alcoholic drinks by ICP-MS after matrix

assisted photochemical vapor generation. Atom. Spectrosc. 33 (2012) 73–

77.

[21] D. Profrock & A. Prange. Inductively coupled plasma-mass spectrometry

(ICP-MS) for quantitative analysis in environmental and life sciences: A

review of challenges, solutions, and trends. Appl. Spectrosc. 66 (2012)

843–868.

[22] V.I. Baranov, Z.A. Quinn, D.R. Bandura & S.D. Tanner. The potential for

elemental analysis in biotechnology. J. Anal. At. Spectrom. 17 (2002b)

1148–1152.

[23] J. Bettmer, N. Jakubowski & A. Prange. Elemental tagging in inorganic

mass spectrometric bioanalysis. Anal. Bioanal. Chem. 386 (2006) 7–11.

[24] A. Prange & D. Proefrock. Chemical labels and natural element tags for the

quantitative analysis of bio-molecules. J. Anal. At. Spectrom. 23 (2008)

432–459.

[25] J.S. Becker & N. Jakubowski. The synergy of elemental and biomolecular

mass spectrometry: new analytical strategies in life sciences. Chem. Soc.

Rev. 38 (2009) 1969–1983.

[26] J. Bettmer, M. Montes-Bayón, J. Ruiz Encinar, M.L. Fernádez, M.

Fernández de la Campa & A. Sanz-Medel. The emerging role of ICP-MS

proteomics analysis. Journal of proteomics 72 (2009) 989-1005.


42

[27] M. Careri, L. Elviri & A. Mangia. Element-tagged immunoassay with

inductively coupled plasma mass spectrometry for multianalyte detection.

Anal. Bioanal. Chem. 393 (2009) 57–61.

[28] S. Bomke, M. Sperling & U. Karst, Organometallic derivatizing agents in

bioanalysis. Anal. Bioanal. Chem. 397 (2010) 3483–3494.

[29] A. Sanz-Medel. ICP-MS for multiplex absolute determinations of proteins.

Anal. Bioanal. Chem. 398 (2010) 1853–1859.

[30] A. Tholey & D. Schaumlöffel. Metal labeling for quantitative protein and

proteome analysis using inductively-coupled plasma mass spectrometry.

Trends Anal. Chem. 29 (2010) 399-408.

[31] M. Wang, W.Y. Feng, Y.L. Zhao, Z.F. Chai. ICP-MS-based strategies for

protein quantification. Mass Spectrom. Rev. 29 (2010) 326–348.

[32] C. Zhang, F.B. Wu, Y.Y. Zhang, X. Wang & X.R. Zhang. A novel

combination of immunoreaction and ICP-MS as a hyphenated technique for

the determination of thyroid-stimulating hormone (TSH) in human serum. J.

Anal. At. Spectrom. 16 (2001) 1393–1396.

[33] L.C. Junqueira, J. Carneiro & T.H. Schiebler. Histologie, 4th ed. Springer-

Verlag, Berlin Heidelberg, 1996.

[34] F. Lottspeich & J.W. Engels. Bioanalytik, 2nd ed. Elsevier GmbH, 2006.

[35] E. Soini & I. Hemmila. Fluoroimmunoassay—Present status and key

problems. Clin. Chem. 25 (1979) 353–361.

[36] R.S. Houk, V.A. Tassel, G.D. Flesch, H.J. Svec, A.L. Gray & C.E. Taylor.

Inductively coupled argon plasma as an ion source for mass spectrometric

determination of trace elements. Anal. Chem. 53 (1980) 2283–2289.

Capítulo 1

43

[37] Nelms SM, editor. 2005. Inductively coupled plasma mass spectrometry

handbook. Oxford: Blackwell Publishing Ltd. p. 485.

[38] D. Beauchemin. Inductively coupled plasma mass spectrometry. Anal.

Chem. 78 (2006) 4111–4135.

[39] A.A. Ammann. Inductively coupled plasma mass spectrometry (ICPMS):

Aversatile tool. J Mass Spectrom. 42 (2007) 419–427.

[40] M. Wind, A. Wegener, A. Eisenmenger, R. Keller & W.D. Lehman. Sulfur as

the key element for quantitative protein analysis by capillary liquid

chromatography coupled to element mass spectrometry. Angewandte

Chemie-Internationl Edition 42 (2003) 3425-3427.

[41] P. Marshall, O. Heudi, S. Bains, H. N. Freeman, F. Abou-Shakra & K.

Reardon. The determination of protein phosphorylation on electrophoresis

gel blots by laser ablation inductively coupled plasma-mass spectrometry.

Analyst 127 (2002) 459-461.

[42] J.R. Encinar, L. Ouerdane, W. Buchmann, J. Tortajada, R. Lobinski & J.

Szpunar. Identification of water-soluble selenium-containing proteins in

selenized yeast by size-exclusion-reversed-phase HPLC/ICPMS followed

by MALDI-TOF and electrospray Q-TOF mass spectrometry Anal. Chem.

75 (2003) 3765-3774.

[43] A. Raab, P. Fecher & J. Feldmann. Determination of arsenic in algae -

Results of an interlaboratory trial: Determination of arsenic species in the

water-soluble fraction. Microchim. Acta 151 (2005) 153-166.

[44] E. del Castillo Busto, M. Montes-Bayón, E. Blanco González, J. Meija & A.

Sanz-Medel, Strategies to study human serum transferrin isoforms using

integrated liquid chromatography ICPMS, MALDI-TOF, and ESI-Q-TOF


44

detection: Application to chronic alcohol abuse. Anal. Chem. 77 (2005)

5615-5621.

[45] M. Montes-Bayón, D. Pröfock, A. Sanz-Medel & A. Prange. Direct

comparison of capillary electrophoresis and capillary liquid chromatography

hyphenated to collision-cell inductively coupled plasma mass spectrometry

for the investigation of Cd-, Cu- and Zn-containing metalloproteins. J.

Chromatogr. A 1114 (2006) 138-144.

[46] G. Grindlay, J. Mora, M.T.C. de Loos-Vollebregt & F. Vanhaecke. A

systematic study on the influence of carbon on the behavior of hard-to-

ionize elements in inductively coupled plasma-mass spectrometry

Spectrochim. Acta Part B 86 (2013) 42-49.

[47] S.D. Tanner, V.I. Baranov & D.R. Bandura. Reaction cells and collision

cells for ICP-MS: A tutorial review. Spectrochim. Acta Part B 57 (2002)

1361–1452.

[48] D.W. Koppenaal, G.C. Eiden & C.J.Barinaga. Collision and reaction cells in

atomic mass spectrometry: Development, status, and applications. J. Anal.

At. Spectrom. 19 (2004) 561–570.

[49] N. Jakubowski, J. Messerschdmidt, M. Garijo Añorbe, L. Waentig, H.

Hayen & P.H. Roos. Labelling of proteins by use of iodination and detection

by ICP-MS. J. Anal. At. Spectrom. 2008 (23) 1487-1496.

[50] Y. Guo, L. Chen, L. Yang & Q. Wang. Counting Sulfhydryls and Disulfide

Bonds in Peptides and Proteins Using Mercurial Ions as an MS-Tag. J. Am.

Soc. Mass Spectrom. 19 (2008) 1108-1113.

Capítulo 1

45

[51] G. Schwarz, L. Mueller, S. Beck & M. W. Linscheid. DOTA based metal

labels for protein quantification: A review. J. Anal. At. Spectrom. 29 (2014)

221-233.

[52] C. Giesen, L. Waentig, U. Panne & N. Jakubowski. History of inductively

coupled plasma mass spectrometry-based immunoassays. Spectrochim.

Acta Part B 76 (2012) 27-39.

[53] R. Liu, P. Wu, L. Yang, X. Hou & Y. Lv. Inductively coupled plasma mass

spectrometry-based immunoassay: a review. Mass Spectrom. 33 (2014)

373-393.

[54] L. Mueller, A.J. Herrmann, S. Techritz, U. Panne & N. Jakuwbowski.

Quantitative characterization of single cells by use of immunocytochemistry

combined with multiplex LA-ICP-MS. Anal. Bioanal. Chem. 409 (2017)

3667-3676.

[55] Z. Liu, B. Yang, B. Chen, M. He & B. Hu. Upconversion nanoparticle as

elemental tag for the determination of alpha-fetoprotein in human serum by

inductively coupled plasma mass spectrometry. Analyst 142 (2017) 197-

205.

[56] G. Han, Z. Xing, Y. Dong, S. Zhang & X. Zhang. One-step homogeneous

DNA assay with single-nanoparticle detection. Angew Chem Int Ed 50

(2011) 3462-3465.

[57] A.R. Montoro Bustos, L. Trapiella-Alfonso, J. Ruiz Encina, J.M. Costa-

Ferández, R. Pereiro & A. Sanz-Medel. Elemental and molecular detection

for quantum dots-based immunoassays: a critical appraisal. Biosensors and

bioelectronics 33 (2012) 165-171.


46

[58] P. Jarujamrus, R. Chawengkirttikul, J. Shiowatana & A. Siripyanoud.

Towards chloramphenicol detection by inductively coupled plasma mass

spectrometry (ICP-MS) linked immunoassay using gold nanoparticles

(AuNPs) as element tags. J. Anal. At. Spectrom. 27 (2012) 884-890.

[59] M. Terenghi, L. Elviri, M. Careri, A. Mangia & R. Lobinski. Multiplexed

determination of protein biomarkers using metal-tagged antibodies and size

exclusion chromatography−inductively coupled plasma mass spectrometry.

Anal. Chem. 81 (2009) 9440-9448.

[60] L. López-Fernández, E. Blanco-González & J. Bettmer. Determination of

specific DNA sequences and their hybridisation processes by elemental

labelling followed by SEC-ICP-MS detection. Analyst 139 (2014) 3423-

3428.

[61] S. Hann, T. Falta, K. Boeck, M. Sulyok & G. Koellensperger. On-line fast

column switching SEC × IC separation combined with ICP-MS detection for

mapping metallodrug-biomolecule interaction. J. Anal. At. Spectrom. 25

(2010) 861-866.

Capítulo 2

Determination of aflatoxin M1 in milk

samples by means of an inductively coupled

plasma mass spectrometry-based

immunoassay

Chapter 2

49

1. Introduction.

Aflatoxins are secondary metabolites produced by different filamentous

fungi (mainly Aspergillus species) and they are known to represent a high risk

for human health due to their mutagenic and teratogenic effects. These

substances could be found in different kinds of food and animal feeds (e.g.

cereals, cocoa, coffee, etc.) that have been in contact with fungi through the

food chain under high temperature and humidity conditions [1].

Aflatoxin M1 (AFM1) is the most significant aflatoxin in milk and dairy

products. This compound is the hydroxylated form of the aflatoxin B1 (AFB1) and

it is usually present in milk when animals have been fed with feedstuffs

containing AFB1 [2]. Aflatoxin M1 has been classified as Group 2 human

carcinogen by the International Agency of Research on Cancer [3]. For this

reason, and taking into account the significance of milk and milk products in

human diet (especially for children), the maximum allowed levels of AFM1 are

strictly regulated worldwide [2,4]. Food and Drug Administration from USA limits

the concentration of AFM1 in milk and processed milk products at 0.50 µg kg-1

[5]. However, European Community Legislation is even more restrictive and

does not allow AFM1 levels in milk and infant formula above 0.050 and 0.025 µg

kg-1, respectively [6,7].

Aflatoxin M1 determination is usually carried out by means of high-

performance liquid chromatography (HPLC) or immunoassays after an

extraction treatment to reduce matrix effects and pre-concentrate the analyte

[8,9]. HPLC is considered the reference method for AFM1 analysis

[10,11,12,13]. The detection of AFM1 is generally achieved by means of both

fluorescence and mass spectrometry. However, HPLC analysis requires

Determination of aflatoxin M1 in milk samples by means of an inductively coupled plasma mass spectrometry-based immunoassay

50

laborious sample preparation treatments to reduce matrix effects and improve

analytical figures of merit. On the other hand, immunoassays (mainly Enzyme

Linked Immunosorbent Assay, i.e., ELISA) are widely used for screening

purposes due to their high sample throughput, simplicity and low budget

[14,15,16,17].

Inductively coupled plasma mass spectrometry (ICP-MS) is a powerful

technique for inorganic analysis due to its: (i) low limits of detection (LoD)

(usually in the µg kg-1-ng kg-1 range), (ii) good precision; (iii) multi-element

capability, (iv) high dynamic range; and (v) the possibility to obtain analyte

isotopic information [18]. Traditionally, the analysis of organic molecules by ICP-

MS has been limited to those analytes containing metals, metalloids and some

non-metals (e.g. P or S) because of the difficulty to utilize C, H, N and O for

quantification purposes at ultra-trace levels. However, it has been demonstrated

that ICP-MS can be used as a detector for all kinds of organic molecules (e.g.

proteins, mRNA, DNA, etc.) after a derivatization procedure with a heteroatom

or a compound containing a heteroatom [19,20]. In this context, ICP-MS has

been employed as a detector of proteins and biomolecules in immunoassays in

view of it is quite straightforward to functionalized antibodies with elements

detectable by this technique [21,22]. In general, antibodies (or any other

species present in the immunoassay) are conjugated with elements presumably

not present in biological samples, such as lanthanide based chelates or Au

nanoparticles. The latter approach is especially advantageous to amplify the

analytical response because of the significant number of quantities of Au atoms

in each nanoparticle. The use of ICP-MS as a detector in immunoassays affords

several attractive features such as: (i) specificity to heteroatom detection; (ii)

Chapter 2

51

compound-independent detection sensitivity; (iii) high elemental sensitivity and

dynamic range; (iv) robustness (complex sample pre-treatments are not

required to diminish matrix effects); and (vii) multielement capabilities, since the

antibodies can be conjugated with different heteroatoms and detected in a

single run. In spite of the above mentioned features, the use of ICPMS-based

immunoassays in food analysis has been limited so far. Nonetheless, these

methods have been successfully applied to quantify peanut allergens [23],

ochratoxine A in wine [24] and progesterone in milk [25].

The goal of this work is to develop a new procedure to quantify AFM1 in milk

samples by ICP-MS at the security levels required by the current international

policies with accuracy and precision. The proposed methodology is based on a

competitive immunoassay using secondary biotinylated antibodies and

streptavidin-Au nanoparticles conjugate followed by Au detection by ICP-MS.

2. Experimental.

2.1. Reagents and materials.

All solutions were prepared using ultrapure water (Milli-Q water purification

system, Millipore Inc., Paris, France).

Sodium carbonate, sodium hydrogen carbonate, monosodium phosphate,

disodium phosphate, sodium chloride, biotinylated goat α-rabbit immunoglubulin

G (IgG) secondary antibody (secondary Ab), aflatoxin M1 (AFM1) from

Aspergillus flavus, AFM1-Bovine serum albumin conjugate (AFM1-BSA),

streptavidin-40 nm Au nanoparticles conjugate from Streptomyces avidinii,

polyethylene glycol sorbitan monolaurate (Tween 20) and HPLC-grade

acetonitrile were purchased from Sigma-Aldrich (Steinheim, Germany). Bovine

serum albumin (BSA) was obtained from Biowest (Nuaillé, France) whereas α-


52

AFM1 primary rabbit polyclonal antibody (pAb) was obtained from Agrisera

(Vännas, Sweeden). Iridium 1000 mg L-1 stock solution was provided by Merck

(Darmstadt, Germany). Thiourea, 69% w w-1 nitric acid and 35% w w-1

hydrochloric acid were purchased from Panreac (Barcelona, Spain).

F16 maxisorp polystyrene microtiter plates were obtained from Thermo-

Scientific (Roskilde, Denmark).

2.2. Buffers and solutions.

Standard stock solution of AFM1 (10 µg mL-1) was prepared in pure

acetonitrile in an amber vial. AFM1-BSA was dissolved in 2 mL of phosphate

buffer solution (PBS, 10 mol L-1 monosodium phosphate, 2 mmol L-1 disodium

phosphate, 154 mmol L-1 sodium chloride, pH 7.6) for a final concentration of

500 µg mL-1. Both solutions were kept at -200C. Primary rabbit polyclonal

antibody was dissolved in 500 µL ultrapure water and kept at 40C.

The following solutions were employed in the ICPMS-based immunoassay:

(a) carbonate/bicarbonate buffer solution (15 mmol L-1 sodium carbonate and

35 mmol L-1 sodium hydrogen carbonate, pH 9.6); (b) 1% w V-1 BSA in a PBS

solution for plate blocking; (c) 1% w V-1 BSA and 0.05% V V-1 Tween 20 in PBS

as incubation medium; (d) 0.05% V V-1 Tween 20 in PBS for washing microtiter

plate wells medium and (d) 4% V V-1 nitric acid and 12% V V-1 hydrochloric acid

for Au-nanoparticles digestion.

Chapter 2

53

2.3. Immunoassay procedure.

The analysis of AFM1 by ICP-MS is based on a competitive immunoassay

[24] in which varying amounts of free AFM1 inhibit the binding of specific pAb to

the solid phase coated with AFM1-BSA conjugate using secondary biotinylated

antibodies and streptavidin- Au nanoparticles conjugate for ICP-MS detection

(see Figure 2.1). First of all, the polystyrene microtiter plate wells were coated

with 100 µL of the appropriate AFM1-BSA concentration in carbonate-

bicarbonate buffer (step 1a). After 1 h incubation at room temperature, wells

were washed three times and blocked with 1% w V-1 BSA in PBS for 1 h at

room temperature. Simultaneously, samples or AFM1 standards were mixed

with the pAb solution (step 1b). The mixture was incubated 1 h at room

temperature and then 100 µL of it were transferred to the plate wells for another

incubation step of 2.5 h at room temperature (step 2). After washing -three

times in order to eliminate the antigen-antibody complexes present in the

solution as well as the free pAb, the microwell plates were sequentially

incubated with 100 µL of a secondary Ab solution (step 3) and then with 100 µL

of the streptavidin-Au nanoparticles conjugate solution (step 4). The incubation

time for the previous steps was 1 h at room temperature followed by three

washing steps. Finally, before ICP-MS analysis, Au nanoparticles were digested

(step 5) with 150 µL of the digestion acid mixture and spiked with 50 µL of a

1.0% w V-1 thiourea solution containing 2.5 µg L-1 Ir. All the AFM1 standards

were analyzed in triplicate wells whereas samples containing unknown AFM1

amounts in quintuplicate wells.


54

Figure 2.1. Scheme of the competitive ICPMS-based immunoassay for AFM1

analysis in milk samples.

+

Anti -AFM1 antibody

Y

Y

AFM1

Microtiterplate

YY

Microtiterplate

Y

YY

Free Anti-AFM1 IgG interacts withfree AFM1-BSA in the microtiterplate

Y YY Y

Y YY Y

Streptavidine-Au addition

MicrotiterplateY Y

Y

Biotinilated secondaryantibody addition

Y

Y

ICP-MS

Microtiterplate

Microtiterplate

Microtiterplate

AFM1-BSA

+

Step 2

Step 1a Step 1b

+

Step 3

Step 4

Step 5Au nanoparticles digestion

Incubation at room temperature (1 h)

Y

Acid mixtureInternal standard (Ir)

Incubation at room temperature (1h)

Incubation at room temperature (2.5 h)

Incubation at room temperature (1h)

Incubation at at room temperature (1h)

Chapter 2

55

2.4. ICP-MS instrumentation.

According to the immunoassay procedure described above, AFM1 is

quantified by means of ICP-MS using the 197Au+ signal. Despite the poor Au

ionization in the plasma because of its high ionization potential (9.23 eV) [26],

the use of Au nanoparticles is especially advantageous to amplify the analytical

response due to the high number of atoms present in each nanoparticle. ICP-

MS measurements were performed by means of a 7700x quadrupole-ICP-MS

system (Agilent, Santa Clara, USA). Operating conditions were daily optimized

to maximize 197Au+ following the instrument user’s guide (Table 2.1).

Table 2.1. Operating conditions employed in ICP-MS.

Agilent 7700x ICP-MS

Plasma forward power (W) 1550

Argon flow rate (L min-1):

Plasma 15

Auxiliary 0.9

Nebulizer 1.01

Sample introduction system

Nebulizer OneNeb micronebulizer

Spray chamber Double pass

Carrier flow rate (mL min−1) 0.6

Dwell time (µs) 0.5

Number of sweeps 100

Replicates 90

Signal nature Area


56

On account of the limited volume of sample available in the immunoassay

(200 µL), a micronebulizer (OneNeb, Ingeniatrics, Sevilla, Spain) coupled to a

double pass quartz spray chamber (Agilent, Santa Clara, USA) was selected as

the sample introduction system. Using this configuration, 197Au+ sensitivity for a

sample uptake rate of 0.6 mL min-1 was 3-fold higher than the one obtained

using the standard nebulizer provided with the instrument (i.e., Micromist, Glass

Expansion, Australia). These results were the expected taking into account the

higher aerosol generation efficiency of the former nebulizer operating at sample

uptake rates in the order of µL min-1 [27]. Initially, samples were tried to be

introduced into the spectrometer using self-aspirating conditions (i.e., without

using a peristaltic-pump) as suggested by Giesen et al. [24]. However, signal

reproducibility was poor (relative standard deviation, RSD, in the 5-8% range)

and strong Au memory effects were registered. In fact, wash-out times (defined

as the time required for reaching the 1% of the stable signal after blank

introduction) of around 300 s for 1 µg Au L-1 were obtained. This behavior could

be attributed to the Au surface sticky nature [21] and the low sample uptake rate

used [28]. In order to improve the reproducibility and the sample throughput, a

flow injection analysis (FIA) procedure was employed. In this operating mode,

samples were introduced into a carrier solution controlled by a peristaltic pump

(Model Minipulse 3, Gilson, France) with the aid of a V-451 flow injection

manifold (Upchurch Scientific, Silsden, United Kingdom) equipped with a 75 µL

loop valve and a syringe. Carrier flow rate was set at 0.6 mL min-1 for high

throughput analysis. To minimize Au wash-out times, different carrier solutions

(i.e. water, diluted nitric acid, etc.) were tested. Finally, a 1% V V-1 hydrochloric

acid and 1% w V-1 thiourea mixture was chosen since it provided the lowest

Chapter 2

57

wash-out times (i.e., lower than 40 s) [29]. Nonetheless, further improvement

was feasible spiking the digestion acid mixture for Au nanoparticles with 1% w

V-1 thiourea. Operating this way, no differences on wash-out times between Au

and other elements (e.g. Mn, Ir, etc.) were observed (25 s). Finally, because of

the discontinuous sample introduction mode of the FIA device, a peak-shape

signal is obtained. Microsoft Excel software was employed to integrate 197Au+

signals manually.

2.5. Calibration.

Aflatoxin M1 determination was performed by means of a calibration curve

built with the 197Au+ ICP-MS signal response of AFM1 standards of

concentrations ranging from 0.001 to 5 µg kg-1. To improve accuracy and

precision, Ir signal (193Ir+) was employed as internal standard for Au

measurements. Ionization potential and m/z values for Ir are closed to the Au

ones and, hence, matrix and drifts effects are expected to be similar for both

elements [30]. Therefore, the 197Au+ and 193Ir+ signal ratio was really employed

to build the calibration curve and quantify AFM1. Iridium was added to the

standards and the unknown samples with the acid mixture employed to digest

Au nanoparticles after the immunoassay procedure (section 2.3.) for a final

concentration of 2.5 µg L-1.

Finally, it is important to remark that, unlike the conventional ICP-MS

analysis, the use of a competitive immunoassay makes that a high ICP-MS

signal is related to a low AFM1 concentration. Thus, for instance, when no AFM1

is present in the sample, all the pAb is retained in the microtiter plate and, as a

consequence, the Au signal is maximum. Because of the sigmoidal curve

response of the competitive immunoassay procedure, analyte determination


58

was circumscribed to the curve section where there was a lineal relationship

between the analyte signal and AFM1 concentration logarithm (i.e. 0.010-2.5 µg

kg-1 under optimum assay conditions).

2.6. Samples.

The AFM1 certified reference material of whole milk powder, ERM-BD284

(certified at 0.44±0.06 µg kg-1), was purchased from the Institute for Reference

Materials and Measurements (European Commission Joint Research Centre,

Geel, Belgium). In addition, five cow milk samples (i.e. raw, pasteurized and

ultrahigh temperature pasteurized) were obtained from retail markets and

supermarkets, stored at 40C and analyzed before their respective expiration

dates.

2.7. Sample preparation.

The ERM-BD284 milk powder (10.0 g) was suspended in 50 mL of ultrapure

water previously heated up to 500C using a stirring rod until a homogeneous

mixture was obtained. After that, the solution was cooled and then diluted to

100 mL using ultrapure water. So, AFM1 concentration in the reconstituted milk

was of 0.044 µg L-1.

Both the certified and the commercial milk samples were pre-treated before

the immunoassay using the procedure described by Huang et al. [31] with some

minor modifications. Thus, 200 µL of milk were mixed with 800 µL of acetonitrile

in an Eppendorf tube to extract AFM1 and remove matrix components. The

extraction was performed using a vortex mixer for 2 min and sonicating the

mixture for 30 min (Vibramix, J.P. Selecta S.A., Barcelona, Spain). Then, the

extracts were centrifuged at 12100 x g for 10 min at 40C (A 5804R Centrifuge,

Chapter 2

59

Eppendorf, Hamburg, Germany). After that, 800 µL of the supernatant were

collected and evaporated up to dryness (miVac Quattro concentrator, Genevac

Ltd, Suffolk, UK). The residue was reconstituted with 100 µL of PBS and then

analyzed.

3. Results and discussion.

3.1. Immunoassay optimization.

The competitive immunoassay described in the present paper has been

developed with a polyclonal antibody produced in rabbits by immunization with

BSA haptenized with AFM1. Considering that not only may this IgG preparation

contain antibodies which react with BSA but also bovine milk contains a

significant amount of this compound (≈1.2 % w V-1), an excess of BSA (1% w V-

1) was included in all the incubation media (see section 2.2) in order to

neutralize any antibody activity specific to BSA in the assay. According to the

supplier, the pAb is highly specific for AFM1 determination with low cross-

reactivity against other aflatoxins (AFB1 2%; aflatoxin B2 0.4%; aflatoxin G1

0.4% and aflatoxin G2 0.1%).

Variables selected for the immunoassay optimization were the

concentration of: (i) AFM1-BSA conjugate; (ii) pAb; and (iii) streptavidin-Au

nanoparticles conjugate. The concentration of the secondary biotinylated

antibody was not optimized and the dilution factor recommended by the

manufacturer (1:2000) was employed. The optimal conditions for the inhibition

assay were chosen by checkerboard titration experiments as described for

ELISA [32,33]. Briefly, decreasing amounts of AFM1 antigen (AFM1-BSA) were

bound to microtiter wells and then incubated with serial concentrations of the

pAb. Inductively coupled plasma mass spectrometry readouts were evaluated to


60

identify the optimal amount of AFM1-BSA per well as well as the corresponding

concentration of pAb. Primary polyclonal antibody was prepared by serial

concentrations from 0.125 to 2 µg mL-1 whereas AFM1-BSA concentration was

modified between 0.07 and 10 ng mL-1. For these experiments, streptavidin-Au

nanoparticles conjugate concentration was kept constant at 0.08 µg mL-1. In

titration experiments, optimal conditions were defined as the amount of

reagents (both AFM1-BSA and pAb) producing signal-to-background ratio in

ICP-MS close to the 80% of the maximal signal at plateau, characteristic of an

antibody excess conditions [33]. The rationale for this criterion was that an

assay dependent on inhibition of antibody binding, sensitivity would be maximal

at the lowest concentration of antibody producing a consistent readout (i.e.

minimal variability in replicates and an adequate dose-response relationship,

when tested with the lowest possible amount of AFM1-BSA bound to solid

phase). Experimental results showed that optimum response was obtained for

0.35 ng L-1 of AFM1-BSA and a pAb concentration of 0.5 µg mL-1 (see Table

S2.1, supplementary material). Once the optimum AFM1-BSA and pAb

concentration were selected, streptavidin-Au nanoparticles conjugate

concentration was optimized following a similar procedure to ensure the

maximum 197Au+ response by the mass spectrometer. To this end, this

parameter was modified between 0.02 and 0.32 µg mL-1. As expected, when

the streptavidin-Au nanoparticles conjugate concentration was decreased, the

Au response in ICP-MS improved. A 6-fold signal improvement was registered

when the concentration was varied from 0.02 to 0.32 µg mL-1 (see Figure S2.1,

supplementary material). On the other hand, the precision in ICP-MS was also

observed to be dependent on the streptavidin- Au nanoparticles conjugate

Chapter 2

61

concentration employed. Gold signal RSD was improved up to 2-3% when

streptavidin-Au nanoparticles conjugate concentration was decreased down

0.08 µg mL-1. No further improvement of the RSD was obtained when higher

concentrations were used. Though higher streptavidin-Au nanoparticles

conjugate concentrations provide higher Au signals in ICP-MS, no real

improvement on analytical figures of merit was obtained (this topic will be

discussed in detail below). For this reason, 0.08 µg mL-1 streptavidin- Au

nanoparticles conjugate concentration was selected for further studies as a

compromise between analytical performance and cost.

For calibration purposes, AFM1 standards were treated under the optimum

competitive immunoassay experimental conditions (Figure 2.2). As it has been

described in the experimental section, the sample containing the AFM1 and the

pAb solution were incubated together (Figure 2.1, step 1.b) and then the

mixture was transferred to the microtiter plate for a second incubation step

(Figure 2.1, step 2). Initially, the incubation time of each step was 1 h. Analytical

response in Figure 2.2 is expressed as the inhibition factor, defined as (S0-

S)·100/S0 where S0 is the maximal signal obtained in wells with no inhibition

(i.e. no AFM1 was added) and S is the signal observed for each sample or

standard preparation. Figure 2.2 shows that the concentration of AFM1 giving

rise to a 50% inhibition factor (i.e., half maximal inhibitory concentration, IC50)

for the ICPMS-based immunoassay was 6.4 µg kg-1 whereas the limit of

detection (LoD) (calculated as three times the standard deviation of the signal of

15 blank replicates [16]) was around 0.070 µg kg-1, low enough to analyze

AFM1 according to USA legislation (AFM1max level: 0.500 µg kg-1). However,

unless a pre-concentration step was implemented, the immunoassay could not


62

be applied to the analysis of milk samples (AFM1max level: 0.050 µg kg-1) and

infant formula (AFM1max level: 0.025 µg kg-1) under the more restrictive EU policy.

In order to reduce the minimum AFM1 concentration level detectable and avoid

costly and long sample pretreatments based on immunocolums for AFM1

preconcentration [10,13], the immunoassay optimization procedure was revised.

Figure 2.2. Aflatoxin M1 calibration curve using different pAb concentrations. (-

�-) 0.5 µg mL-1 pAb concentration/incubation time 1h; (-�-) 0.25 µg mL-1 pAb

concentration/incubation time 2.5 h. Aflatoxin M1-BSA concentration: 0.35 ng

mL-1; secondary Ab dilution factor: 1:2000; streptavidin-Au nanoparticles

conjugate concentration: 0.08 µg mL-1.

Competitive immunoassay LoD is strongly linked to both primary antibody

concentration and detector capability to recognize the primary antibody retained

in the microtiter plate after the incubation step with the analyte containing

sample. Thus, pAb concentration could be decreased to make the

0

20

40

60

80

100

120

0.0001 0.001 0.01 0.1 1 10 100 1000

Inh

ibti

on

(%

)

AFM1 concentration (µg kg-1)

EU limit USA limit

Chapter 2

63

immunoassay more sensitive to low AFM1 levels. However, this means that pAb

retained in the microtiter plate is less significant, thus making detection more

difficult or even impossible. Accordingly, LoDs for the ICPMS-based

immunoassay are expected to improve by increasing the efficiency of pAb

retention in the microtiter plate and/or detection by ICP-MS. First of all, the

incubation time of the AFM1 and the pAb in the microtiter plate coated with

AFM1-BSA (step 2) was increased from 1 to 2.5 h with the aim to favor pAb

retention on it. Operating in this way, it was feasible to reduce pAb

concentration from 0.5 to 0.25 µg mL-1, thus improving sensitivity and LoDs

without compromising robustness. As a consequence, the IC50 was

approximately 10 times lower (0.42 µg kg-1) and LoD decreased down to 0.005

µg kg-1 (Figure 2.2). The use of higher incubation times together lower pAb

concentration (< 0.25 µg mL-1) was not further explored since sample

throughput is negatively affected and LoD were low enough to quantify AFM1

according to the current international policies for this analyte. Alternatively, the

feasibility of using lower streptavidin-Au nanoparticles conjugate concentration

to improve ICP-MS detection was checked but no real improvement was

obtained on LoDs. It should be taking into account that secondary Ab has a

limited amount of biotin moieties and, as a consequence, signal amplification

with streptavidin-Au nanoparticles conjugate is limited and could not

compensate the lower amount of pAb retained in the microtiter plate. Probably,

LoD could be improved using a sector field instrument or with a heteroatom-tag

with a higher sensitivity in ICP-MS in order to use lower pAb concentrations.


64

3.2. Method validation.

The analytical methodology was evaluated according to European

Conformity guidelines for analytical methods of food contaminants and

mycotoxins [6,34]. First of all, method accuracy and precision was checked by

analyzing an AFM1 certified reference material (ERM-BD284 milk powder, 0.044

± 0.006 µg kg-1) and different milk samples spiked with known amounts of

AFM1. Recovery test was performed spiking milk samples with AFM1 standard

for a final concentration of 0.030 and 0.080 µg kg-1. All the samples were

analyzed following the optimized immunoassay after an extraction treatment

with acetonitrile to mitigate the effects of matrix components (e.g. proteins, fats

etc.) on the antibody reaction. In fact, direct analysis of milk samples produced

systematically AFM1 recovery values between 200-300%. Sample dilution to

mitigate matrix effects was not explored due to its negative impact on LoDs.

Table 2.2 shows the recovery values for AFM1 in the certified reference

material and in the spiked milk samples after the sample pre-treatment with

acetonitrile. There is observed to be a good agreement between experimental

and theoretical values. Aflatoxin M1 recovery values ranged between 80% and

102%. These values were within the limits established by the EU for analyte

concentrations below 1 µg kg-1 (-40%/+20%). The repeatability (intra-assay

precision) of the method was determined by analyzing five replicates of each

sample on the same day. The RSD of the AFM1 concentration levels was within

the 5%-15% range. These values are the typical for immunoassays. The

immunoassay reproducibility (inter-assay precision) of the proposed

methodology was evaluated as the RSD of the measurements obtained for five

independent immunoassays performed in five different days. The average of

Chapter 2

65

AFM1 concentration obtained for the certified reference material was 0.042 ±

0.008 µg kg-1 which highlights the reproducibility of the assay. Similar

conclusions were obtained for the spiked milk samples.

Table 2.2. Aflatoxin M1 recovery assay for different kinds of milk samples.

AFM1 concentration (µg kg-1) Recovery (%)*

Sample Certified/Spike Experimental

Certified milk powder 0.044 ± 0.006 0.038 ± 0.006 87 ± 10

Raw milk 0.030 0.027 ± 0.003 90 ± 10

0.080 0.078 ± 0.004 93 ± 5

Pasteurized milk 0.030 0.024 ± 0.005 80 ± 17

0.080 0.078 ± 0.002 98 ± 2

UHT whole milk 0.030 0.026 ± 0.003 87 ± 8

0.080 0.079 ± 0.010 99 ± 10

UHT whole milk (2) 0.030 0.028 ± 0.003 93 ± 7

0.080 0.082 ± 0.008 102 ± 10

Replicates = 5, * mean ± standard deviation

Finally, the lower and the upper quantification limits of the immunoassay (lLoQ

and uLoQ) were estimated [35]. The lLoQ was defined as the analyte

concentration that has a response at least 3 times that of a blank sample and

repeatability lower than 20%. Similarly, the uLoQ was defined as the highest

concentration standard that signal response has repeatability lower than 20%.

The lower and the upper quantification limits of the immunoassay experimental

values were 0.012 µg kg-1 and 2.5 µg kg-1, respectively. These results


66

demonstrate the suitability of the ICPMS-based immunoassay for AFM1

determination in milk samples according to the current security policies for this

analyte. In fact, this methodology was applied to the analysis of different milk

samples (i.e. raw, pasteurized or ultra-high temperature pasteurized) but no

AFM1 was detected (< 0.005 µg kg-1).

3.3. Comparison with other methodologies.

Analytical figures of merit of the ICPMS-based immunoassay have been

compared with those previously reported in the literature (Table 2.3). In general,

the methodology developed in this work affords similar results to those obtained

with chromatographic-based approaches or even other immunoassays.

Comparing to chromatographic methods, poorer precisions are obtained (RSD

nearly 2 to 3-fold higher), but no laborious sample pretreatments based on solid

phase extraction or immunocolumns are required to preconcentrate and purify

the AFM1. Sample pretreatment proposed in the present work is quite simple

and it is only focused to eliminate the most significant matrix components

without compromising sample throughput. Nevertheless, a 2-fold AFM1

preconcentration factor is obtained operating on this way. On the other hand,

due to immunoassay simplicity, reagents and solvent requirements (as well as

wastes) are minimized. As regards other immunoassay procedures, ICPMS

detection offers a wider linear range. Moreover, the proposed method not only

shows lower background and blank levels but also an independent analytical

response from incubation and storages times. In fact, standards and samples in

the microtiter plate can be stored 2-3 weeks after acid mixture addition without

significant effects on AFM1 results. However, sample throughput is partially

Chapter 2

67

degraded since microtiter plate wells are sequentially analyzed by the ICP-MS

(3-4 h to analyze a 96-well plate).

Up to date, the use of ICP-MS as a detector for mycotoxin analysis has

been limited. In fact, it has only previously used to quantify ochratoxin A levels

in wine by means of a competitive immunoassay methodology [24]. Though this

previous work is not focused on AFM1, it is worth to compare both

methodologies due to their similarities. Thus, Ochratoxin A detection was

accomplished using secondary antibodies functionalized with Au nanoparticles

and a sector field based ICP-MS. The limit of detection using this configuration

was stablished at 0.003 µg kg-1. A priori the high detection capabilities found

could be attributed (at least in part) to the high sensitivities afforded by sector

field mass spectrometers which makes feasible to use a low pAb concentration.

However, our work shows that AFM1 can also be quantified in the ng kg-1 range

using a quadrupole-based mass spectrometer with a two-step signal

amplification procedure based on a biotinylated secondary antibody and

streptavidin-Au nanoparticles conjugate. The possibility of using quadrupole-

based ICP-MS instruments to quantify toxin at ultra-trace levels is

advantageous since they are the most spread mass spectrometers worldwide.

Furthermore, precision for the AFM1 immunoassay was significantly better (5 -

15%) than that for the Ochratoxin A (5 - 40%). The origin of these differences is

not clear due to the different procedures implemented in each immunoassay.

Nonetheless, it points out that the precision achievable for the ICPMS-based

immunoassay is mainly limited by the immunological step since uncertainty

derived by the ICP-MS lays below 5%.

Determ

inatio

n o

f aflatoxin

M1 in

milk sam

ples b

y mean

s of an

ind

uctively co

up

led p

lasma m

ass spectro

metry-b

ased im

mu

no

assay

Tab

le 2.3. Co

mp

aris

on

of d

ive

rse

an

aly

tica

l me

thod

olo

gie

s p

rop

ose

d in

the

litera

ture

for th

e d

ete

rmin

atio

n o

f AF

M1 in

milk

sa

mp

les.

Meth

od

olo

gy

Reco

very

(%)

Precisio

n

(%)

LO

D

(µg kg

-1)

Lin

ear rang

e

(µg kg

-1) R

eference

Ind

irect co

mp

etitive

imm

uno

assa

y (ICP

-MS

) 8

0-1

02

5-1

5

0.0

05

0.0

10

-2.5

T

his

wo

rk

Liq

uid

-liqu

id e

xtractio

n H

PL

C-F

luo

resce

nce

7

3-9

9

2-7

0

.05

- [1

1]

IAC

-HP

LC

-FD

1

16

5 0

.010

0.0

1-0

.20

[36

]

MS

FE

-HP

LC

-FD

9

1-1

02

5 0

.005

0.0

15

-10

[37

]

SP

E-L

CM

S

78

-108

5-1

0 0

.010

0.0

20

-1 [3

8]

UP

HL

C-M

S/M

S

84

-97

13

0.0

10

- [1

2]

80

-110

<1

0 0

.005

0.0

25

-10

[13

]

Ind

irect co

mp

etitive

imm

uno

assa

y (EL

ISA

) 8

0-1

02

5-1

7

0.0

40

0.0

40

-0.5

00

[39

]

Dire

ct c

om

pe

titive E

LIS

A

90

-110

<1

0

0.0

03

- [1

5]

Dire

ct c

om

pe

titive E

LIS

A

93

-98

5-8

0

.008

0.0

04

-0.2

50

[40

]

Chapter 2

69

4. Conclusions.

This work demonstrates that ultra-trace AFM1 analysis in milk samples is

feasible using an ICPMS-based immunoassay. Analytical figures of merit of this

method fulfil the most restrictive current international policies for AFM1 analysis

in milk and dairy products. Despite the fact that sample throughput could be

deteriorated in relation to other immunoassay methodologies described in the

literature, the use of ICP-MS for mycotoxin analysis has a great potential in food

analysis due to a higher dynamic range, lower background levels and the

independence of analytical response from incubation or storage times. In this

regard, aflatoxin detection by means of ICP-MS could be still improved. Thus,

LoDs for the competitive immunoassay employed in this work strongly depend

on the primary antibody concentration and ICP-MS Au sensitivity to detect the

amount of primary antibody retained in the microtiter plater. Therefore, detection

capabilities could be probably improved (below ng L-1) using lower primary

antibody concentration with a sector field mass spectrometer and/or labelling

with higher sensitive hetereoatoms in ICP-MS. On the other hand, methodology

sample throughput could be significantly enhanced determining other

mycotoxins together AFM1 by means of antibodies functionalized with different

heteroatoms. Operating on this way ICP-MS multiplexing capabilities are

exploited and the use of a mass spectrometer could be more beneficial than

other immunoassay detection procedures (e.g. ELISA, etc.). These experiments

are currently being carried out in our laboratories.


70

5. References.

[1] R. Bhat, R.V. Rai, & A.A. Karim. Mycotoxins in food and feed: present status

and future concerns. Comprehensive reviews in food science and food

safety 9 (2010) 57-81.

[2] M.E. Flor-Flores, E. Lizarraga, A. López de Cerain & E. González-Peñas.

Presence of mycotoxins in animal milk: a review. Food Control 53 (2015)

163-176.

[3] International Agency for Research on Cancer. Monographs on the evaluation

of carcinogenic risks to humans. Some traditional herbal medicines, some

mycotoxins, naphthalene and styrene, World Health Organization (WHO),

Lyon, France, 2002, vol. 82, 171-174.

[4] A. Prandini, G. Tansini, S. Sigolo, L. Filippi, M. Laporta & G. Piva. On the

occurrence of aflatoxin M1 in milk and dairy products. Food and Chemical

Toxicology: an international journal published for the British industrial

biological research association 47 (2009) 984-991.

[5] FDA, Sec. 527.400 Whole milk, Low fat milk, skim milk – aflatoxin M1 (CPG

7106.10), FDA/ORA Compliancy Guides 2005.

[6] EC (2002). Commission Decision 2002/657/EC of 12 August 2002

implementing Council Directive 96/23/EC concerning the performance of

analytical methods and the interpretation of results, as amended by

Decision 2003/181/EC (4). Official Journal of the European Communities,

8–36.

[7] EC (2004). Commission Regulation, Regards Aflatoxins and Ochratoxin a in

Foods for Infants and Young Children, Official Journal of the European

Union L, 106, 5-24.

Chapter 2

71

[8] G.S. Shephard. Determination of mycotoxoins in human foods. Chemical

Society Reviews 37 (2008) 2468-2477.

[9] E. Reiter, J- Zentek & E. Razzazi. Review of sample preparation and

methods used for the analysis of aflatoxins in food and feed. Molecular

Nutrition Food Research 53 (2009) 508-524.

[10] S. Dragacci, F. Grosso & J. Gilbert. Immunoaffinity column cleanup with

liquid chromatography for determination of Alfatoxin M1 in liquid milk:

collaborative study. Journal of AOAC International 84 (2001) 437-443.

[11] P. Diniz Andrade, J. Laine Gomez da Silva & E. Dutra Caldas.

Simultaneous analysis of afaltoxins B1, B2, G1, G2, M1 and ochratoxin A in

brest milk by high-performance liquid chromatography/fluorescende after

liquid-liquid extraction with low temperature purification (LLE-LTP). Journal

of Chromatography A 1304 (2013) 61-68.

[12] M.M. Aguilera-Luiz, P. Plaza-Bolaños, R. Romero-González, J.L. Martínez

Vidal & A. Garrido Frenich. Comparison of the efficiency of different

extraction methods for the simultaneous determination of mycotoxins and

pesticides in milk samples by ultra high-performance liquid

chromatography-tandem mass spectrometry. Analytical Bionalytical

Chemistry 399 (2011) 2863-2875.

[13] E. Beltran, M. Ibañez, J.V. Sancho, M.A. Cortés, V. Yusa & F. Hernández.

UHPLC-MS/Ms highly sensitive determination of aflatoxin, the aflatoxin

metabolite M1 and ochratoxin A in baby food and milk. Food Chemistry, 126

(2011) 737-744.

[14] Y. Pico. (2007) Food Toxicants Analysis. (1st ed). Amsterdam, Elsevier,

(chapter 5)


72

[15] D. Guan, P. Li, W. Zhang, D. Zhang & J. Jiang. An ultra-sensitive

monoclonal antibody-based competitive enzyme immunoassay for aflatoxin

M1 in milk and infant milk products. Food Chemistry 125 (2011) 1359-1364.

[16] W. Jiang, Z. Wang, G. Nolke, J. Zhang, L. Niu & J. Shen. Simultaneous

determination of aflatoxin B1 and aflatoxin M1 in food matrices by enzyme-

linked immunosorbent assay. Food Analytical Methods 6 (2013) 767-774.

[17] M.M. Vdovenko, C.C. Lu, F.Y. Yu & I.Y. Sakharov. Development of

ultrasensitive direct chemiluminescent enzyme immunoassay for

determination of aflatoxin M1 in milk. Food Chemistry 158 (2014) 310-314.

[18] J. Bettmer, M. Montes-Bayón, J. Ruiz Encinar, M.L. Fernádez, M.

Fernández de la Campa & A. Sanz-Medel. The emerging role of ICP-MS

proteomics analysis. Journal of proteomics 72 (2009) 989-1005.

[19] A. Tholey & D. Schaumlöffel. Metal labeling for quantitative protein and

proteome analysis using inductively-coupled plasma mass spectrometry.

Trends in Analytical Chemistry 29 (2009) 399-408.

[20] D. Kretschy, G. Koellensperger & S. Hann. Elemental labelling combined

with liquid chromatography inductively coupled plasma mass spectrometry

for quantification of biomolecules: A review. Anal. Chim. Acta 750 (2012)

98-110.


coupled plasma mass spectrometry-based immunoassays. Spectrochimica.

Acta Part B 76 (2012) 27-39.


spectrometry-based immunoassay: a review, Mass Spectrometry Reviews

33 (2014) 373-393.

Chapter 2

73

[23] M. Careri, L. Elviri, A. Mangia & C. Mucchino. ICP-MS as a novel detection

system for quantitative element-tagged immunoassay of hidden peanut

allegerns in foods. Analytical Bionalytical Chemistry 387 (2007) 1851-1854.

[24] C. Giensen, N. Jakubowski, U. Panne, & M.G. Weller. Comparison of

ICPMS and photometric detection of an immunoassay for the determination

of ochratoxin A in wine. Journal of Analytical Atomic Spectrometry 25

(2010) 1567-1572.

[25] A.R. Montoro Bustos, L. Trapiella-Alfonso, J. Ruiz Encina, J.M. Costa-

Ferández, R. Pereiro & A. Sanz-Medel. Elemental and molecular detection

for quantum dots-based immunoassays: a critical appraisal. Biosensors and

bioelectronics 33 (2012) 165-171.

[26] NIST atomic spectradatabase (version 5), http://physics.nist.gov/cgi-

bin/ASD/ie.pl. Accessed 01.12.16.

[27] B. Almagro, A.F. Gañán-Calvo, M. Hidalgo & A. Canals. Flow focusing

pneumatic nebulizer in comparison with several micronebulizers in

inductively coupled plasma atomic emission spectrometry. Journal of

Analytical Atomic Spectrometry 21 (2006) 770-777.

[28] J. Mora, S. Maestre, V. Hernandis & J.L. Todolí. Liquid-sample introduction

in plasma spectrometry. Trends in Analytical Chemistry 22 (2003)123-132.

[29] W. Chen, P. Wee & I.D. Brindle. Elimination of the memory effects of gold,

mercury and silver in inductively coupled plasma atomic emission

spectroscopy. Journal of Analytical Atomic Spectrometry 15 (2000) 409-

413.

[30] F. Vanhaecke, H. Vanhoe, R. Dams & C. Vandecasteele. The use of

internal standards in ICP-MS. Talanta 39 (1992) 737-742.


74

[31] L.C. Huang, N. Zheng, B.Q. Zheng, F. Wen, J.B. Cheng, R.W. Han, X.M.

Xu, S.L. Li & J.Q. Wang. Simultaneous determination of aflatoxin M1,

ochratoxin A, zearalenone and -zearalenol in milk by UHPLC–MS/MS.

Food Chem. 146 (2014) 242–249.

[32] S.J. Gee, T. Miyamoto, M.H. Goodrow, D. Buster & B.D. Hammock.

Development of an enzyme-linked immunosorbent assay for the analysis of

the thiocarbamate herbicide molinate. Journal of Agricultural and Food

Chemistry 36 (1988) 863–870.

[33] J.R. Crowter (2001) The ELISA Guidebook. (2nd ed.) New Jersey, Humana

press.

[34] EC (2006). Commission Regulation (EC) No. 1881/2006 of 19 December

2006 setting maximum levels for certain contaminants in foodstuffs. Official

Journal of the European Union L, 364, 5–24.

[35] Guidance for Industry, Bioanalytical Method Validation, US Department of

Health and Human Services, Food and Drug Administration Centre for Drug

Evaluation and Research (CDER), Centre for Veterinary Medicine (CVM),

May 2001.

[36] A. Gurbay, S. Aydin, G. Girgin, A.B. Engin & G. Sahin. Assessment of

aflatoxin M1 levels in milk in Anakara, Turkey. Food Control 17 (2006) 1-4.

[37] M. Hashemi & Z. Taherimaslak. Determination of aflatoxin M1 in liquid milk

using high performance liquid chromatography with fluorescence detection

after magnetic solid phase extraction. RSD Advances 4 (2014) 33497-

33506.

Chapter 2

75

[38] L.K. Sørensen & T.H. Elæk. Determination of mycotoxins in bovine milk by

liquid chromatography tandem mass spectrometry. Journal of

Chromatography B 820 (2005) 183–196.

[39] S.C. Pei, Y.Y. Zhang, S.A. Eremin & W.J. Lee. Detection of aflatoxin M1 in

milk products from China by ELISA using monoclonal antibodies. Food

Control 20 (2009) 1080–1085.

[40] A. Radoi, M. Targa, B. Prieto-Simon & J.L. Marty. Enzyme-Linked

Immunosorbent Assay (ELISA) based on superparamagnetic nanoparticles

for aflatoxin M1 detection. Talanta 77 (2008) 138–143.

Chapter 2

77

Supplementary data.

Table S2.1. Results of the checkboard titration experiments to optimize AFM1-

BSA and pAb concentration. Secondary Ab dilution factor: 1:2000; streptavidin-

Au nanoparticles conjugate concentration: 0.16 µg mL-1; incubation time in the

microtiterplate of AFM1 standards and pAb solution: 1 h.

pAb concentration (µµµµg mL-1)

AFM1-BSA concentration (µg mL-1) 2 1 0.5 0.25 0.125

10 1.0 0.9 0.8 0.6 0.4

5 1.0 0.9 0.8 0.6 0.5

2.5 0.9 0.9 0.8 0.7 0.5

0.75 0.9 0.9 0.8 0.7 0.5

0.35 0.9 0.9 0.8 0.6 0.5

0.15 0.9 0.9 0.7 0.6 0.5

0.07 0.9 0.8 0.7 0.5 0.4


78

Figure S2.1. Influence of the streptavidin-Au nanoparticles conjugate

concentration on the 197Au+ normalized signal and signal standard deviation in

ICP-MS. Aflatoxin M1-BSA concentration: 0.35 ng mL-1; pAb concentration: 0.5

µg mL-1; secondary antibody dilution factor: 1:2000; incubation time in the

microtiterplate of AFM1 standards and pAb solution: 1 h.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

0

1

2

3

4

5

6

7

2 1 0.5 0.25 0.125

Sign

al st

anda

rd d

evia

tion

(%)

Nor

mal

ized

Au

sign

al

Streptavidin-Au concentration (µµµµg mL-1)

Capítulo 3

Evaluation of different competitive

immunoassay for aflatoxin M1 determination

in milk samples by means of inductively

coupled plasma mass spectrometry

Chapter 3

81

1. Introduction.

The significance of inductively coupled plasma mass spectrometry (ICP-

MS) for biomolecule analysis has exponentially increased over the years [1].

The use of ICP-MS in this field offers several attractive features regarding other

stablished approaches: (i) sensitivity; (ii) specificity; (iii) compound-independent

detection sensitivity; (iv) multi-element capabilities; (v) robustness; and (vi)

versatility to be coupled to separation techniques. Initially, the analysis was

limited to those species containing a heteroatom detectable by ICP-MS.

However, analytical figures of merit were severely compromised since the

heteroatoms naturally present in biomolecules (e.g. P, S, Se, As, etc.) suffer

from low sensitivity and (spectral and non-spectral) interferences due to matrix

components in biological samples (e.g. carbon, chloride, etc.) [1,2]. To improve

the analytical figures of merit as well as to address with the determination of

non-containing heteroatoms biomolecules, different labelling strategies have

been proposed in the literature. First, the biomolecule can be derivatized

through a chemical reaction with a heteroatom or an organometallic compound

[3,4]. Alternatively, the analyte of interest can be labelled indirectly by means of

immunoreaction using a heteroatom-labelled antibody [5,6]. To this end, non-

competitive immunoassays (i.e. sandwich type) have been traditionally

employed for the analysis of high molecular weight biomolecules such as

proteins, enzymes, etc. Dealing with haptens (i.e., biomolecules which

molecular weight is lower than 10 kDa), quantification is only accomplished by

means of competitive immunoassays since, given the volume of these

molecules, they only react with a single antibody. Competitive immunoassays

have been successfully reported for several hormones (thyroxine [7],

Evaluation of different competitive immunoassay for aflatoxin M1 determination in milk samples by means of inductively coupled plasma mass spectrometry

82

progesterone [8]), herbicides (2,4-dichlorophenoxyacetic acid [9]), antibiotics

(chloramphenicol [10]); and toxins (ochratoxin A [11], aflatoxin M1 (AFM1) [12])

determination.

Two different competitive assay formats have been mainly reported for

hapten analysis with ICP-MS detection so far. In the first approach, the sample

is spiked with a given amount of the specific labelled antibody and the mixture is

incubated in a solid phase coated with an antigen (analyte)-protein conjugate. A

competitive reaction is stablished between the antigen bound to the solid phase

and the antigen present in solution for the limited number of the binding sites of

the antibody. The final stage of the assay involves the measurement of the

signal related to the amount of the antibody retained on the solid phase.

Alternatively, the competitive reaction can be also stablished by the competition

between the labelled and the unlabelled antigen for the limited number of the

binding sites of the antibody coating the solid phase. Up to date, both

immunoassay formats have been indistinctively employed in the literature but

the benefits (and drawbacks) of each approach have not been systematically

investigated under a similar set of experimental conditions. Regardless of the

competitive immunoassay format, different heteroatoms have been employed in

the literature for hapten analysis: (i) Eu [8]; (ii) Fe [11]; (iii) Au nanoparticles

[10,12,13]; and (iv) Quantum-Dots [9]. A priori, the use of Au nanoparticles or

Quantum-Dots seems to be more advantageous to improve analytical figures of

merit (e.g. sensitivity, LoDs, etc.) of competitive immunoassays due to the high

number of atoms present in each nanoparticle. Nevertheless, no previous

attempt has been made so far to systematically evaluate the influence of the

nanoparticle characteristics operating with competitive immunoassays.

Chapter 3

83

The goal of this work is to systematically compare diverse competitive

immunoassay formats for hapten determination by means of ICP-MS detection

using nanoparticles of different nature and size. Aflatoxin M1 has been selected

as a model of small chemical analyte to evaluate the benefits and drawbacks of

the different approaches, due to its high health risk and restrictive legal

requirements. Competitive immunoassays based on the use of either antibody

or antigen-protein conjugate as tracer species have been investigated since

they are the most frequently employed. Nanoparticles covering different

elements (Ag/Au) and sizes (40/80 nm) have been used through this work.

Finally, the methodologies developed have been applied to the AFM1 analysis in

milk samples.

2. Experimental.


Anti-aflatoxin M1 primary rabbit polyclonal antibody (pAb) (1 mg mL-1) was

obtained from Agrisera (Vännas, Sweeden). According to the supplier, this

antibody is highly specific for AFM1 determination with low cross-reactivity

against other aflatoxins (aflatoxin B1 2%; aflatoxin B2 0.4%; aflatoxin G1 0.4%

and aflatoxin G2 0.1%).

Biotinylated goat α-rabbit IgG (whole molecule) secondary antibody

(secondary Ab), aflatoxin M1 (AFM1) from Aspergillus flavus (5 g mL-1), AFM1-

Bovine serum albumin conjugate (AFM1-BSA) (500 µg mL-1), sodium carbonate,

sodium hydrogen carbonate, monosodium phosphate, disodium phosphate,


84

sodium chloride, polyethylene glycol sorbitan monolaurate (Tween 20), HPLC-

grade acetonitrile were purchased from Sigma-Aldrich (Steinheim, Germany).

Bovine serum albumin (BSA) was purchased from Biowest (Nuaillé,

France). Streptavidin – (80/40 nm) Ag or (80/40 nm) Au nanoparticles conjugate

were obtained from Cytodiagnostics (Ontario, Canada).

Iridium and rhodium 1000 mg L-1 stock solution were provided by Merck

(Darmstadt, Germany). Thiourea, 69% w w-1 nitric acid and 35% w w-1

hydrochloric acid were purchased from Panreac (Barcelona, Spain).

All solutions were prepared using ultrapure water (Milli-Q water purification

system, Millipore Inc., Paris, France).

F16 maxisorp polystyrene microtiter plates were obtained from Thermo-

Scientific (Roskilde, Denmark). Immunoaffinity columns (IACs) Afla M1 (Vicam,

Watertown, MA) with a capacity of approximately 150 ng and based on

monoclonal antibodies were employed for AFM1 determination in milk samples.

Amicon Ultra-4 Centrifugal Filter Units with Ultracel-10 membrane (Merck

Millipore, Cork, Ireland) were used throughout the work for washing steps

during the biotin labelling of AFM1-BSA.

2.2. Buffers and solutions.

Standard stock solution of AFM1 (10 g L-1) was prepared in pure

acetonitrile. Aflatoxin M1-BSA was dissolved in 2 mL of phosphate buffer

solution (PBS; 10 mM NaH2PO4, 2 mM Na2HPO4, 154 mM NaCl, pH 7.6) for a

final concentration of 500 g mL-1. Both solutions were kept at -200C. Primary

rabbit polyclonal antibody was dissolved in 500 µL ultrapure water and kept at

40C.

Chapter 3

85

The following solutions were employed in the different immunoassays

tested: (a) as microtiter plate coating medium: carbonate buffer solution (15mM

Na2CO3 and 35 mM NaHCO3, pH 9.6); (b) as plate blocking medium: 1% w V-1

BSA in a PBS solution; (c) as incubation medium: 1% w V-1 BSA and 0.05% V

V-1 Tween 20 in PBS; (d) as washing medium: 0.05% V V-1 Tween 20 in PBS;

(e) as Au-nanoparticles digestion medium: 4% V V-1 nitric acid and 12% V V-1

hydrochloric acid; and (f) as Ag-nanoparticles digestion medium: 2 % V V-1 nitric

acid. It is important to remark that the use of BSA in the incubation medium is

critical to ensure immunoassay reproducibility. The pAb could react either with

the AFM1 or BSA. Therefore, an excess of BSA was included in the incubation

medium to eliminate pAbs without affinity to AFM1. Moreover, this strategy is

also useful to avoid potential matrix effects for AFM1 analysis in milk samples

since they contain significant amount of BSA.

2.3. Immunoassay procedures.

In the present work, different competitive immunoassay formats have been

tested for AFM1 determination with ICP-MS detection (Figure 3.1); namely: (i)

antibody binding inhibition assay (ABIA). This immunoassay is based on that

described in our previous work [12] in which pAb are employed as the tracer

specie. The competitive reaction is stablished between the AFM1 present in the

sample and the AFM1-BSA, coating the solid phase, for the limited number of

the binding sites of the pAb spiked in the sample solution. A biotinylated

secondary Ab and streptavidin-nanoparticles conjugate were employed to

detect the pAb retained on the solid phase; (ii) capture inhibition assay (CIA).

This procedure is based on that previously reported by Trapiella-Alfonso et al.


86

[13] in which a labelled antigen is used as the tracer specie. In this case, the

competitive reaction is stablished between the AFM1 present in the sample and

the biotinylated AFM1-BSA conjugates (spiked to the sample) for the limited

number of the binding sites of the pAb immobilized on the solid phase. The

biotinylated AFM1-BSA retained on the solid phase is quantified using

streptavidin-nanoparticles conjugates; and (iii) capture bridge inhibition assay

(CBIA). For the first time, this immunoassay procedure has been proposed for

hapten determination with ICP-MS detection. It is based on the CIA procedure

but includes some modifications to favor immunocomplex formation. A mixture

of the pAb and the biotinylated AFM1-BSA is spiked to the sample. A

competitive reaction is stablished between the biotinylated AFM1-BSA and the

AFM1 present in the sample for the limited number of the binding sites of the

pAb. Next, and given the bivalent nature of the antibodies, the AFM1-pAb and

the biotinylated AFM1-BSA-pAb complexes generated are retained by the

AFM1-BSA which coates the solid phase. Streptavidin-nanoparticles conjugates

are used again to detect the biotinylated AFM1-BSA-pAb complexes retained on

the solid phase.

For all strategies, the influence of the nanoparticle type (i.e. Ag and Au) and

size (i.e. 40 and 80 nm) on the analytical figures of merit have been

investigated.

The biotinylated AFM1-BSA required for CIA and CBIA procedures was

obtained using a commercial biotin labelling kit. Briefly, 2.3 mg of succinimidyl-

6-(biotinamido)hexanoate was dissolved in 500 µL of dimetylformamide and

then added to a 200 µL AFM1-BSA solution. The reaction was carried out on ice

for 2 hours in the dark and the excess of non-reactive biotin reagent was

Chapter 3

87

removed on a cutoff filter by centrifugation with PBS and recovered at 150 µg

mL-1. The biotinylated AFM1-BSA solution was divided into single working

aliquots and stored at -200C.

Figure 3.1. Scheme of the different competitive immunoassay formats tested in

this work.

2.3.1 Antibody binding inhibition assay (ABIA).

Plates were coated at room temperature with 100 µL/well of the appropriate

AFM1-BSA concentration in 0.05 M carbonate–bicarbonate buffer (pH 9.6). After

incubation for 1 h, plates were washed three times and then blocked (200

µL/well) for 1 h at room temperature. Simultaneously, samples or AFM1

A. Antibody binding inhibition assay (ABIA)

B. Capture inhibition assay (CIA)

C. Capture bridge inhibition assay (CIA)

++

Y

Step 1.1 Step 1.2 Step 1.3 Step 1.4 Step 1.5

Step 2.1 Step 2.2 Step 2.3 Step 2.4

+ Digestion+

ICP-MS

AFM1 Anti-AFM1 antibody (primary)

AFM1-BSA Biotynilated antibody (secundary) Biotynilated AFM1-BSAStreptavidin-nanoparticles

M

M M

+

YYYYYY

YYYYYYM

YYYYYY

M M

Step 3.1 Step 3.2 Step 3.3

+

+

+

Step 3.4Y Y

M M M

M

Y


88

standards were incubated with the appropriate concentration of the pAb solution

for 1 h at room temperature (Figure 3.1, step 1.1). After another washing step,

100 µL/well of the AFM1-pAb mixture was added and incubated for 2.5 h at

room temperature (Figure 3.1, step 1.2). Following a washing step (Figure 3.1,

step 1.3), the secondary Ab solution (100 µL/well) was added and incubated for

1 h at room temperature (Figure 3.1, step 1.4). The plates were washed again,

and 100 µL/well of streptavidin-nanoparticles conjugate solution was added and

incubated for 1 h at room temperature (Figure 3.1, step 1.5).

2.3.2 Capture inhibition assay (CIA).

Plates were coated at room temperature with 100 µL/well with the

appropriate pAb concentration in 0.05 M carbonate–bicarbonate buffer (pH 9.6).

After incubation for 1 h, plates were washed three times and then blocked (200

µL/well) for 1 h at room temperature. Next, after a washing step, samples or

AFM1 standards and the appropriate biotinylated AFM1-BSA concentration were

mixed (Figure 3.1, step 2.1) and transferred to the solid phase (Figure 3.1, step

2.2). The mixture was incubated for 1 h at room temperature and, after a

washing step (Figure 3.1, step 2.3), 100 µL/well of streptavidin-nanoparticles

conjugate solution was added and incubated for 1 h at room temperature

(Figure 3.1, step 2.4).

2.3.3. Capture bridge inhibition assay (CBIA).

Plates were coated at room temperature with 100 µL/well of 1 ng L-1 AFM1-

BSA in 0.05 M carbonate–bicarbonate buffer (pH 9.6). After incubation for 1 h,

plates were washed three times and then blocked (200 µL/well) for 1 h at room

Chapter 3

89

temperature. Then, after a washing step, samples or AFM1 standards, pAb and

biotynilated AFM1-BSA solutions were mixed (Figure 3.1, step 3.1) and

transferred to the solid phase (Figure 3.1, step 3.2). Next, the mixture was

incubated for 1 h at room temperature. Following a washing step (Figure 3.1,

step 3.3), 100 µL/well of streptavidin-nanoparticles conjugate solution was

added and incubated for 1 h at room temperature (Figure 3.1, step 3.4).

Finally, before ICP-MS analysis and with independence of the

immunoassay format employed, nanoparticles in the microtiterplate wells were

acid digested and spiked with 50 µL of a 1% w V-1 thiourea solution containing

2.5 µg L-1 of the corresponding internal standard (Rh/Ir) for a final volume of 200

µL. All AFM1 standards were analyzed in triplicate wells whereas samples

containing unknown AFM1 amounts in quintuplicate wells.

2.4. Instrumentation.

Aflatoxin M1 is indirectly quantified by means of ICP-MS using the signal of

the element present in the nanoparticle (107Ag+ and 197Au+). Experimental

measurements were performed by means of a 7700x quadrupole-ICP-MS

system (Agilent, Santa Clara, USA). The Table 3.1 shows the operating

conditions employed in this work. The performance of the system was checked

to respect the manufacturer indications for 1 g L-1 of 7Li+, 89Y+, and 205Tl+ in 2%

HNO3.

The sample was introduced in the equipment using a micronebulizer

(OneNeb, Ingeniatrics, Sevilla, Spain) coupled to a double pass quartz spray

chamber (Agilent, Santa Clara, USA). Because of the limited volume of sample


90

available in the immunoassay (200 µL), a V-451 flow injection manifold

(Upchurch Scientific, Silsden, United Kingdom) equipped with a 75 µL loop

valve was employed. The samples were introduced with the aid of a syringe into

a carrier solution controlled by a peristaltic pump (Model Minipulse 3, Gilson,

France). Carrier flow rate was set at 0.6 mL min-1 for high throughput analysis.

Two different carrier solutions were selected depending on the nanoparticle

type. A mixture of 1% V V-1 nitric acid and 1% w V-1 thiourea solution was

employed for Ag measurements whereas a 1% V V-1 hydrochloric acid and 1%

w V-1 thiourea mixture was employed as carrier for Au determination. Operating

in this way, no significant memory effects were registered in ICP-MS [14].

Table 3.1. Operating conditions employed in ICP-MS.

Agilent 7700x ICP-MS

Plasma forward power (W) 1550


Plasma 15

Auxiliary 0.9

Nebulizer 1.01


Nebulizer OneNeb micronebulizer

Spray chamber Double pass

Carrier flow rate (mL min−1) 0.6

Dwell time (µs) 0.5

Number of sweeps 100

Replicates 90

Signal nature Area

Chapter 3

91

To improve accuracy and precision, 103Rh+ and 193Ir+ were employed as internal

standards for Ag and Au measurements, respectively. These internal standards

match m/z values and ionization potentials and, hence, matrix and drifts effects

are expected to be similar [15]. Rhodium and Ir were added to the standards

and the unknown samples with the acid mixture employed to digest the

nanoparticles after the immunoassay procedure (section 2.3) for a final

concentration of 2.5 µg L-1. Microsoft Excel software was employed to integrate

peak signals manually, given the peak-shape signal obtained operating the FIA

device.

2.5. Calibration.

Irrespective of the immunoassay format employed, standard curves were

obtained by plotting the inhibition factor value against the logarithm of the

analyte concentration because of the sigmoidal curve response of the

competitive immunoassay procedure. The inhibition value is defined as (S0-

S)/S0 × 100, where S is the signal ratio between 107Ag+ or 197Au+ and its

corresponding internal standard for each sample or AFM1 standard solution and

S0 is the signal ratio when no AFM1 is added.

2.6. Samples.

A certified reference material of whole milk powder (ERM-BD283, European

Commission Joint Research Centre, Geel, Belgium) at 111 ± 18 ng kg-1 of AFM1

and five cow commercial milk samples from retail markets have been analyzed

through this work. All the samples were stored at 40C and analyzed before their

respective expiration dates.


92

2.7. Sample preparation.

The ERM-BD283 milk powder (10.0 g) was suspended in 50 mL of ultrapure

water previously heated up to 50°C using a stirring rod until a homogeneous

mixture was obtained. After that, the solution was cooled and then diluted to

100 mL using ultrapure water. So, AFM1 concentration in the reconstituted milk

was 11.1 ± 1.8 ng kg-1.

All the milk samples were pre-treated before the analysis to mitigate

interferences caused by matrix components (e.g. proteins, lipids, etc.) since,

otherwise, accuracy and precision is severely compromised. To this end, two

different approaches have been employed through this work depending on the

immunoassay methodology employed. On one hand, in the case of ABIA, the

samples were pretreated using the procedure described by Huang et al. [16]

with some minor modifications. Thus, 200 L of milk were mixed with 800 L of

acetonitrile in an Eppendorf tube to extract AFM1 and remove matrix

components. The extraction was performed using a vortex mixer for 2 min and

sonicating the mixture for 30 min (Vibramix, J.P. Selecta S.A., Barcelona,

Spain). Then, the extracts were centrifuged at 12100 x g for 10 min at 4°C (A

5804R Centrifuge, Hamburg, Germany). After that, 800 L of the supernatant

were collected and evaporated up to dryness (miVac Quattro concentrator,

Genevac Ltd, Suffolk, UK). The residue was reconstituted with 100 L of PBS

and then analyzed. On the other hand, in the case of CBIA, milk samples were

pretreated using immunocolumns following the procedure suggested by the

manufacturer with some minor modifications. Briefly, a 100 mL volume of liquid

milk was centrifuged at 1614 × g for 15 min, to separate the fat and thin upper

fat layer was discarded. Fifty mL of the milk were passed through

Chapter 3

93

immunoaffinity column at 1-2 drops/second by gravity. Then, the column was

washed with 10 mL water at a rate of 1-2 drops/second until air comes through

column. Aflatoxin M1 was eluted slowly from column with 4 mL pure acetonitrile

at 1 drop/2-3 seconds by gravity. Two hundred µL of the sample eluate were

evaporated up to dryness (miVac Quattro concentrator, Genevac Ltd, Suffolk,

UK). The residue was reconstituted with 100 L of PBS and then analyzed.


Competitive immunoassays described in the literature shows several

differences on the experimental setup according to the tracer specie monitored:

the specific antibody against the analyte [12,13] or the antigen-protein

conjugate [9,10]. For the first time, immunoassays based on both approaches

are compared under a similar set of experimental conditions for AFM1 analysis.

From now on, antibody inhibition immunoassay (ABIA) will denote the

procedure based on the antibody detection whereas capture inhibition assay

(CIA) will be used as the immunoassay format based on the antigen-protein

conjugate detection.

3.1. Optimization of the immunoassay procedures.

The optimization of the immunoassay formats tested were performed by

means of checkerboard titration experiments as previously described for

Enzyme Linked Immunosorbent Assay (ELISA) [17,18]. The optimal conditions

were defined as the amount of reagents producing signal-to-background ratio in

ICP-MS close to 80% of the maximal signal at plateau, characteristic of a tracer

excess conditions [18]. The rationale for this criterion was that an assay


94

dependent on inhibition of tracer binding, its sensitivity would be maximal at the

lowest concentration of the tracer producing a consistent readout.

3.1.1. Antibody Binding Inhibition Assay (ABIA).

In our previous work [12], ABIA experimental conditions were optimized

using streptavidin-40 nm Au nanoparticles conjugate for AFM1 determination in

milk samples. It was observed that the limit of detection (LoD) was strongly

related to the pAb concentration. As long as the heteroatom-label allows the

accurate detection of the pAb retained on the solid phase, pAb concentration

could be decreased to improve immunoassay detection capabilities to low AFM1

levels. In this context, high sensitive heteroatom labels could be required. Thus,

operating with nanoparticles, ICP-MS signal mainly depends on the

nanoparticle size and the properties of the element in the nanoparticle (e.g.

ionization energy). Therefore, for the first time, this works explores the influence

of nanoparticle characteristics on the analytical figures of merit for a competitive

immunoassay based on the use of antibodies as tracers. To this end,

nanoparticles of different elements (Au and Ag) and size (80 nm and 40/80 nm,

for Au and Ag, respectively) conjugated to streptavidin were investigated. For

each of them, the following concentrations were optimized: (i) pAb; (ii) AFM1-

BSA; and (iii) streptavidin-nanoparticles conjugates. The concentration of the

secondary Ab was not optimized, and the value recommended by the

manufacturer was employed (0.5 µg mL-1). For the sake of comparison,

previous data with streptavidin-40 nm Au nanoparticles conjugate was

employed.

Table 3.2 summarizes the optimum immunoassay conditions for each

streptavidin-nanoparticles conjugate.

Chapter 3

95

Table 3.2. Influence of the streptavidin-nanoparticles conjugate on ABIA

optimum experimental conditions, LoDs and dynamic range.

Ag Au

Parameter 40 nm 80 nm 40 nm 80 nm

AFM1-BSA (µg mL-1) 0.35 0.35 0.35 0.35

pAb (ng mL-1) 240 15 240 15

Biotinylated secondary-Ab (ng mL-1) 500

Streptavidin-Au nanoparticles (µµµµg mL-1) 0.08

LoD (ng L-1) 12 2 5 0.1

LLoQ -ULoQ (ng L-1) 30-5000 6-1250 10-2500 0.3-300

Firstly, the optimum AFM1-BSA and pAb concentration was investigated by

means of a checkerboard tritation experiments. The AFM1-BSA concentration

was modified between 0.07 and 10 µg mL-1 whereas the pAb concentration

ranged from 16 to 2000 ng mL-1. The streptavidin-nanoparticles conjugate

concentration was kept at 0.08 µg mL-1 for these experiments. The optimum

AFM1-BSA concentration was similar for all streptavidin-nanoparticles

conjugates (i.e. 0.35 µg mL-1), but significant differences were noted on the pAb

concentration. The optimum pAb concentration using both 80 nm Ag and Au

nanoparticles was 15 ng mL-1 whereas it was 250 ng mL-1 for 40 nm Ag

nanoparticles. Interestingly, the optimum pAb concentration using 40 nm Ag

nanoparticles was similar to that previously found for 40 nm Au counterparts

[12]. From these findings, it could be derived that for this assay format,

analytical figures of merit afforded were strongly dependent on the nanoparticle

size.


96

Figure 3.2 shows the influence of the pAb concentration on LoD operating

with streptavidin-80 and 40 nm Au nanoparticles conjugate.

Figure 3.2. Influence of the α-AFM1 pAb concentration on the limits of detection

of AFM1 operating with different streptavidin-Au nanoparticles conjugates. ()

40 nm () 80 nm. Aflatoxin M1-BSA concentration: 0.35 ng L-1; secondary Ab

concentration: 500 ng mL-1; streptavidin-nanoparticles conjugate concentration:

0.08 µg mL-1.

It is important to remark that the only difference between the experimental set

up of both Au nanoparticles was the pAb concentration. So, it was feasible to

directly evaluate the influence of the pAb concentration on the LoD. Limit of

detection was calculated as three times the standard deviation of the signal of

15 blank replicates [19]. Experimental data for 40 nm Au nanoparticles was

measured following previously reported experimental conditions [12]. As

expected, irrespective of the nanoparticle considered, LoD increased when the

pAb concentration was increased. Under optimum pAb concentration,

0.01

0.1

1

10

100

15 30 60 120 240 480 960

LoD

(ng

L-1)

Ab concentration (µg L-1)

Chapter 3

97

streptavidin-80 nm Au nanoparticles conjugate afforded a LoD of 0.1 ng L-1, i.e.

50-fold lower than that obtained with the 40 nm Au counterparts. Similar

findings were observed for Ag nanoparticles but the LoD obtained were

significantly higher than those obtained using Au nanoparticles (Table 3.2).

Thus, the LoDs for 80 nm and 40 nm Ag nanoparticles were 2 and 12 ng L-1,

respectively. The origin of these differences is not clear; since, for a given

nanoparticle size, optimum immunoassay conditions do not depend on the

element in the nanoparticle. Nevertheless, it was noted that the use of Ag

nanoparticles was tricky due to a build up of Ag metallic deposits into the

injector tube that could even lead to full blockage. This phenomenon was also

noticed for Au nanoparticles but the metallic deposit formation was significantly

lower. Finally, the influence of the streptavidin-nanoparticle conjugate

concentration on the analytical figures of merit was also investigated. This

parameter was modified between 0.02 µg mL-1 and 0.32 µg mL-1. It was

observed that the signal registered by the mass spectrometer increased with

the streptavidin-nanoparticle conjugate concentration but no changes were

produced either on the optimum pAb and AFM1-BSA concentration or the LoD

for the nanoparticles tested. These results were similar to those previously

reported for 40 nm Au nanoparticles [12]; thus, pointing out that this parameter

was not critical for the immunoassay performance. It should be considered that,

given the limited amount of biotin moieties in the secondary Ab, signal

amplification is not high enough to reduce the pAb concentration in the

immunoassay.

Finally, the lower and upper quantification limits (lLoQ and uLoQ,

respectively) obtained with each streptavidin-nanoparticle conjugate were


98

evaluated (Table 3.2). The lLoQ was defined as the analyte concentration that

has a response at least 3 times that of a blank sample and repeatability lower

than 20%. Similarly, the uLoQ was defined as the highest concentration

standard that signal response has repeatability lower than 20% [20]. As

expected by the changes on detection capabilities, lLoQ and uLoQ depends on

the streptativin-nanoparticle conjugate employed. For instance, lLoQ and uLoQ

experimental values for streptavidin-80 nm Au nanoparticle conjugates were 0.3

ng L-1 and 300 ng L-1, respectively. Nevertheless, the dynamic linear range for

the different nanoparticles tested was similar (3/2-order of magnitude).

3.1.2. Capture Inhibition Assay.

In this immunoassay format, unlike the previous one, the sample containing

AFM1 was spiked with biotinylated AFM1-BSA and both species compete for the

limited number of the binding sites of the pAb which coated the solid phase. The

variables selected for the optimization of this immunoassay were the

concentration of: (i) pAb; (ii) biotinylated AFM1-BSA; (iii) streptavidin-

nanoparticles conjugate; and (iv) the type and size of the nanoparticle in the

streptavidin-nanoparticles conjugate. Firstly, the pAb and the biotinylated AFM1-

BSA concentration was optimized by means of checkerboard titration

experiments. The pAb concentration was ranged from 0.6 µg mL-1 to 10 µg mL-1

whereas the biotinylated AFM1-BSA concentration was modified between 0.08

and 5 µg mL-1. For these experiments, the streptavidin-nanoparticles conjugate

concentration was kept constant at 0.08 µg mL-1. Experimental results showed

that the optimum response was obtained for 2.5 µg mL-1 pAb and 0.03 ng mL-1

biotinylated AFM1-BSA. Under these experimental conditions, a calibration

curve was built with AFM1 standards to evaluate the immunoassay detection

Chapter 3

99

capabilities and the dynamic range. It was observed that a LoD of 100 ng kg-1

was always obtained, regardless the size and the element of the nanoparticle

employed (Table 3.3).

Table 3.3. Optimum experimental conditions, limits of detection and dynamic

range for CIA and CBIA formats.

Parameter CIA CBIA

Nanoparticle Ag/Au (40/80 nm) Ag/Au (40/80 nm)

AFM1-BSA (µg mL-1) - 1

pAb (ng mL-1) 2500 2000

Biotinylated AFM1-BSA (µg mL-1) 0.03 0.04

Streptavidin-Au nanoparticles (µµµµg mL-1) 0.08 0.08

LoD (ng L-1) 100 5

LLoQ -ULoQ (ng L-1) 300-5000 15-2500

Though this value was similar to that previously reported for antigen-protein

conjugate employed as tracer for progesterone analysis [9], it was significantly

higher than the value obtained by means of ABIA. To improve the LoD, the

immunoassay optimization was repeated using higher streptavidin-

nanoparticles conjugate concentrations. The concentration of this reagent was

modified from 0.08 to 0.64 µg mL-1. As it was shown for ABIA, analyte signal

increased with streptavidin-nanoparticles conjugate concentration but no

significant changes were produced on the optimum pAb and biotinylated AFM1-

BSA concentration values previously obtained and, consequently, on the LoDs.

Therefore, the streptavidin-nanoparticles conjugate concentration was kept at

0.08 µg mL-1 to minimize operative cost. Alternatively, the influence of the


100

incubation time of the AFM1 and the biotinylated AFM1-BSA mixture on the

microtiter plate was studied since it favored tracer retention on the solid phase;

thus, making feasible to operate with lower tracer concentrations [12]. The

incubation time was increased from 1 to 10 hours but, again, no significant

improvement on the LoD was obtained. From these findings, it can be derived

that detection capabilities were limited by the immunoassay procedure itself. It

should be considered that, given the size of the BSA moieties (66 kDa), high

steric effects are expected when biotinylated AFM1-BSA molecules are

captured by the pAbs coating the microtiter plate. Obviously, steric effects are

more significant when decreasing the AFM1 content in the sample since more

biotinylated AFM1-BSA molecules are captured on the solid phase. To address

this issue, the use of antibodies containing a spacer was proposed. From a

practical point of view, the use of antibodies with a spacer would reduce sample

throughput due to an additional incubation step would be required and, hence,

this approach was discarded. Instead, some modifications were implemented in

this immunoassay format to favor the immunocomplex formation. To this end, a

mixture of the pAb and the biotinylated AFM1-BSA was initially spiked to the

sample containing the AFM1. A competitive immunoreaction was stablished

between the AFM1 and the biotinylated AFM1-BSA for the limited number of the

binding sites of the pAb. In this way, the immunocomplexes were generated in a

homogeneous phase rather than in a heteregoenus phase; thus, minimizing

steric effects. Finally, given that the antibodies are bivalent molecules, the

mixture was incubated in a microtiter plate coated with AFM1-BSA in order to

capture AFM1-BSA-pAb and AFM1-pAb immunocomplexes. According to this

scheme, AFM1-BSA coating the microtiter plate merely served as a capturing

Chapter 3

101

agent to fix the immunocomplexes formed on the solid phase. To our best

knowledge, this is the first time that this strategy is employed for hapten

analysis with ICP-MS detection. From now on, this immunoassay will be called

capture bridge immunoassay (CBIA).

The optimization of CBIA was carried out likewise the CIA procedure but

including the concentration of the AFM1-BSA coating the microtiter plate. After

some preliminary test, it was observed that this concentration was not critical for

the immunoassay. This reagent had to be in excess to maximize the

immunocomplex capture and, hence, this parameter was kept at 1 ng L-1. Next,

a checkerboard tritation procedure was employed to optimize the biotinylated

AFM1–BSA and the pAb concentration. To this end, the pAb concentration was

modified from 0.125 µg mL-1 to 2 µg mL-1 whereas the biotinylated AFM1 –BSA

concentration was tested from 0.01 µg mL-1 to 0.64 µg mL-1. A streptavidin-

nanoparticles conjugate concentration of 0.08 µg mL-1 was employed. A

satisfying compromise between an optimal analytical sensitivity with a

consistent readout was obtained using 2 µg mL-1 pAb and 0.04 µg mL-1

biotinylated AFM1-BSA for all streptavidin-nanoparticles conjugate. Under these

conditions, as it was previously noticed for the CIA procedure, LoDs for AFM1

determination were again independent on the streptavidin-nanoparticles

conjugate, both type and concentration. This phenomenon is clearly observed in

Figure 3.3 where the calibration curves obtained using different streptavidin-Au

nanoparticles conjugate are presented. Nevertheless, steric effects were less

significant with the new approach since the LoD for CBIA format (5 ng mL-1)

was improved 20-fold regarding CIA strategy. From these results, and given its

better analytical performance, CBIA strategy was selected for further studies.


102

Figure 3.3. Aflatoxin M1 calibration curve using different streptavidin-Au

nanoparticles conjugate for the CBIA procedure. (--) 40 nm; (--) 80 nm.

Aflatoxin M1-BSA concentration: 1.0 ng mL-1; streptavidin-Au nanoparticles

conjugate concentration: 0.08 µg mL-1.

3.2. Aflatoxin M1 analysis in milk samples.

Both ABIA and CBIA formats were validated for AFM1 analysis in milk

samples. The streptavidin-80 nm Au nanoparticles conjugate was used for ABIA

measurements due to its higher detection capabilities. As regards CBIA, though

the nanoparticle characteristics are not critical, 80 nm Au nanoparticles were

also employed for the sake of comparison. Method accuracy and precision were

checked by analyzing an AFM1 certified reference material (ERM-BD283 milk

powder, at 11.1 ± 1.8 ng kg-1) and different commercial milk samples spiked

with known amounts of AFM1. Recovery test was performed spiking milk

samples with AFM1 standard at two different levels: above (80 ng kg-1) and

below (30 ng kg-1), the limits stablished by the European Union policy.

0

20

40

60

80

100

120

1 10 100 1000 10000

Inhi

bitio

n (%

)

AFM1 concentration (ng Kg-1)

Chapter 3

103

Initially, all samples were analyzed after an extraction pretreatment with

acetonitrile to mitigate the effects of matrix components (e.g. proteins, fats, etc.)

on the immunoreaction following the optimized ABIA and CBIA procedures.

However, AFM1 recoveries were systematically higher than 100% for both

approaches. These results were totally unexpected since this approach was

previously employed for AFM1 analysis with the ABIA procedure using the

streptavidin-40 nm Au nanoparticles conjugate [12]. The higher matrix effects

registered for the ABIA format using streptavidin-80 nm Au nanoparticles

conjugate could be attributed to the lower pAb concentration employed since it

makes the immunoassay more sensitive to low AFM1 concentrations but also to

the matrix components. Given the LoD afforded by the ABIA procedure, sample

dilution was explored to mitigate matrix effects due to its simplicity and lower

cost regarding immunocolumns. To this end, before the extraction pretreatment,

milk samples were diluted with the appropriate amount of ultrapure water. Table

3.4 shows the recovery values for AFM1 determination in the certified reference

material and in the spiked milk samples by means of ABIA procedure using

streptavidin-80 nm Au nanoparticles conjugate after a 4-fold milk dilution. There

is observed to be a good agreement between the experimental and the

theoretical values. Aflatoxin M1 recovery values ranged between 97% and

107%. These values were within the limits established by the EU for analyte

concentrations below 1 µg kg-1 (-40%/+20%) [23,24]. It is important to remark

that the use of a lower dilution factor did not allow accurate AFM1

determinations. Taking into account the dilution factor applied, LoD for AFM1

determination in milk samples by means of ABIA format was 0.4 ng L-1. The

repeatability (intra-assay precision) was determined by analyzing five replicates


104

of each sample on the same day. The RSD of the AFM1 determination was

within the 5%-15%. The immunoassay reproducibility (inter-assay precision)

was evaluated as the RSD of the measurements obtained for five independent

immunoassays performed in five different days. The reproducibility was within

10-25% for the ABIA format.

Table 3.4. Recovery analysis of AFM1 by ABIA and CBIA methodologies.

Sample pretratment: ABIA: dilution + acetonitrile extraction; CBIA:

immunocolumns.

Immuoassay Sample

AFM1 concentration (ng kg-1)*

Recovery (%)* Certified

/expected Experimental

ABIA

Commercial whole milk 80 81 ± 9 101 ± 8


Commercial fresh milk 80 81 ± 4 107 ± 7


ERM®-BD283 11.1 ± 1.8 11 ± 1 103 ± 7

CBIA

Commercial whole milk 80 71 ±2 0 89 ± 20




ERM®-BD283 11.1 ± 1.8 11 ± 3 100 ± 28

* mean ± standard deviation, replicates = 5.

As regards the CBIA format, the poor recoveries obtained for milk analysis after

the extraction pretreatment pointed out that this immunoassay was quite

sensitive to matrix components. It should be taking into account that matrix

components affect both the biotinylated AFM1-BSA-pAb immunocomplex

Chapter 3

105

formation and its retention on the microtiter plate. Unlike ABIA, dilution strategy

could not be employed with this immunoassay since LoD did not allow to

quantify AFM1 according to European Union Policy. Consequently,

immunocolumns were required for AFM1 determination. This approach made

feasible AFM1 analysis but at the expense of a significant precision increase

(Table 3.4). The RSD for intra-assay and inter-assay determination of AFM1

was 30%. Due to the extraction-preconcentration pretreatment, LoD for CBIA

format was approximately improved 50-fold (i.e. 0.1 ng mL-1) and, consequently,

detection capabilities for CBIA strategy were similar to those shown by ABIA. It

is interesting to note that irrespective of the immunoassay format, LoD for AFM1

determination were significantly lower (30/45000-fold) than those previously

reported for other hapten analysis by means of competitive immunoassays with

ICP-MS detection.

Finally, both methodologies were applied to the analysis of commercial

products (i.e. raw, pasteurized and ultra-high temperature pasteurized milk)

obtained from retail markets and supermarkets. However, none of them showed

detectable levels of AFM1.

3.3. Comparison of different competitive immunoassay formats for

AFM1 determination.

Aflatoxin M1 has been classified as Group 2 human carcinogen by the

International Agency of Research on Cancer [21] and, consequently, the

maximum allowed levels of this substance are strictly regulated worldwide. For

instance, European Community legislation limits AFM1 levels in milk and infant

formula at 50 and 25 ng kg-1 [22,23] whereas Food and Drug Administration

from USA limits do allow AFM1 levels up to 500 ng kg-1 [24]. Therefore, the


106

analytical parameters (i.e., LoD, dynamic range, sample throughput, matrix

effects, etc.) of the different immunoassay formats developed have been

compared for AFM1 analysis.

From previous data shown in Sections 3.1 and 3.2, it could be derived that

both ABIA and CBIA could be employed for AFM1 analysis according to the

international policies. Nevertheless, the use of pAb as tracer specie (ABIA)

seems to be more advantageous than the use of the antigen-protein conjugates

as the tracer one (CBIA) for AFM1 analysis. Immunocomplex formation in the

ABIA format is not impeded by steric effects since, given the volume occupied

by the BSA residues on the solid phase, pAbs have enough space to form the

immunocomplex without interacting with its neighborhoods (Figure 3.1). In this

case, the use of nanoparticles with bigger diameter is indeed advantageous to

detect lower amounts of pAb retained on the solid phase. As regard CIA and

CBIA, steric effects control the immunocomplex formation between the

biotinylated AFM1-BSA and the pAbs and, consequently, no direct advantages

are derived from using different kind of nanoparticles for tracer detection.

Probably, steric effects explain the higher matrix effects and lower dynamic

range shown by CBIA regarding ABIA. On the other hand, despite ABIA

immunoassay procedure takes longer than CBIA (7 vs 4 hours), no real

advantage on sample throughput is derived from the latter since sample

preparation is more complex, time consuming and costly due to the use of

immunocolumns. It is important to note that cost derived from the use of 100-

fold less amount of pAb is lower for ABIA in comparison to CBIA. Nevertheless,

this feature is partially counterbalanced due to the use of a secondary Ab.

Chapter 3

107

Finally, analytical figures of merit of ABIA and CBIA procedures have been

compared with those previously reported in the literature (Table 3.5).

Comparing to other competitive immunoassays and sensors, the LoD and the

dynamic range are clearly improved by the immunoassays with ICP-MS

detection. Nevertheless, sample throughput is partially deteriorated due to the

sequential nature of the mass spectrometer. As regard chromatographic

methods, the ABIA format affords similar LoDs without requiring complex

sample pretreatments based on solid phase extraction or immunocolumns to

preconcentrate and purify the AFM1. In this regard, CBIA format is clearly less

attractive since immunocolumns are mandatory for AFM1 analysis. The main

benefit of mass spectrometry is the feasibility of multi-compound analysis.

4. Conclusions.

This work demonstrates that the tracer specie used (i.e. antibody or

antigen-protein) in competitive immunoassays with ICP-MS detection is critical

for AFM1 analysis. Using the antibody as tracer specie, the heteroatom-label

exerts a great influence on the immunoassay experimental conditions and,

consequently, on the analytical figures of merit. This type of immunoassay

affords better analytical figures of merit and lower matrix effects than those

based on the use of antigen-protein conjugate as tracer. The immunocomplex

formation in the latter strategy is severely hampered by steric effects caused by

the protein moiety in the antigen-protein conjugate. According to these results,

competitive immunoassays based on the use of antibodies as tracer specie

seem to be more suitable for hapten analysis by means of ICP-MS detection

since this strategy exploits better the detection capabilities afforded by the

technique.

Evalu

ation

of d

ifferent co

mp

etitive imm

un

oassay fo

r aflatoxin

M1 d

etermin

ation

in m

ilk samp

les by m

eans o

f ind

uctively co

up

led p

lasma m

ass sp

ectrom

etry

Tab

le 3.5. Co

mp

aris

on

of d

ive

rse

an

aly

tica

l me

thod

olo

gie

s p

rop

ose

d in

the

litera

ture

for A

FM

1 de

term

ina

tion

in m

ilk s

am

ple

s.

M

etho

do

log

y R

ecovery

(%)

Precisio

n

(%)

LO

D

(ng

kg-1)

Lin

ear rang

e

(ng

kg-1)

Referen

ce

Immunoassays

AB

IA- IC

PM

S

98-1

07

< 1

0

0.4

1-3

00

This

work

CB

IA-IC

PM

S

89-1

06

< 3

0

0.1

15-2

500

This

work

Dire

ct c

om

petitive

ELIS

A (U

V-V

is d

ete

ction)

90-1

10

< 1

0

3

- [2

5]

Dire

ct c

om

petitive

ELIS

A

(chem

ilum

inis

cente

dete

ctio

n)

80-1

20

< 1

0

1

2-7

.5

[26]

Indire

ct c

om

petitive

ELIS

A (U

V-V

is d

ete

ctio

n)

80-1

02

5-1

7

40

40-5

00

[27]

Liquid chromatography

Hig

hly-s

ensitive

time-re

solve

d flu

ore

scent

imm

unochro

mato

gra

phic a

ssay

80-1

10

5-1

2

0.3

0.1

-200

[28]

UH

PLC

-MS

/MS

79

< 1

5

0.1

8

50-5

00000

[29]

SP

E-U

PLC

-MS

/MS

89-1

20

2-9

0.3

1-3

00

[30]

Liq

uid

-liquid

extra

ctio

n H

PLC

-Flu

ore

scence

73-9

9

2-7

50

- [3

1]

MS

PE

-HP

LC

-FD

91-1

02

5

5

15-1

0000

[32]

Sensors

Optic

al im

munosensor

- -

0.6

U

p to

1.8

× 1

03

[33]

Ele

ctro

chem

ical im

munosensor

- -

12

15-1

000

[34]

Flo

w in

jectio

n in

munoass

ay

80-1

20

8

11

20 - 5

00

[35]

Chapter 3

109

5. References.

[1] J. Bettmer, M. Montes-Bayón, J. Ruiz Encinar, M.L. Fernández, M.

Fernández de la Campa & A. Sanz-Medel. The emerging role of ICP-MS in

proteomic analysis. J. Proteoms. 72 (2009) 989-1005.

[2] G. Grindlay, J. Mora, M.T.C. de Loos-Vollebregt & F. Vanhaecke. A

systematic study on the influence of carbon on the behavior of hard-to-

ionize elements in inductively coupled plasma-mass spectrometry

Spectrochim. Acta Part B 86 (2013) 42-49.

[3] S. Bomke, M. Sperling & U. Karst, Organometallic derivatizing agents in

bioanalysis. Anal. Bioanal. Chem. 397 (2010) 3483–3494.




98-110.



Acta Part B 76 (2012) 27-39.


spectrometry based immunoassay: a review. Mass Spectrom. 33 (2014)

373-393.

[7] C. Zhang, F. Wu & X. Zhang. ICP-MS-based competitive immunoassay for

the determination of total thyroxin in human serum. J. Anal. At. Spectrom.

17 (2002) 1304-1307.

[8] A.R. Montoro Bustos, L. Trapiella-Alfonso, J. R. Encinar, J.M. Costa-

Fernández, R. Pereiro & A. Sanz-Medel. Elemental and molecular detection


110

for Quantum Dots-based immunoassays: A critical appraisal. Biosens.

Bioelectronics 33 (2012) 165-171.

[9] A.P. Deng, H.T. Liu, S.J. Jiang, H.J. Huang & C.W. Ong. Reaction cell

inductively coupled plasma mass spectrometry-based immunoassay using

ferrocene tethered hydroxysuccinimide ester as label for the determination

of 2,4-dichlorophenoxyacetic acid. Anal. Chim. Acta 472 (2002) 55–61.

[10] P. Jarujamrus R. Chawengkirttikul, J. Shiowatana & A. Siripinyanond.

Towards chloramphenicol detection by inductively coupled plasma mass

spectrometry (ICP-MS) linked immunoassay using gold nanoparticles

(AuNPs) as element tags J Anal. At. Spectrom. 27 (2012) 884-890.

[11] C. Giensen, N. Jakubowski, U. Panne, & M.G. Weller. Comparison of

ICPMS and photometric detection of an immunoassay for the determination

of ochratoxin A in wine. J. Anal. At. Spectrom. 25 (2010) 1567-1572.

[12] E. Pérez, P. Martínez-Peinado, F. Marco, L. Gras, J.M. Sempere, J. Mora &

G. Grindlay. Determination of aflatoxin M1 in milk samples by means of an

inductively coupled plasma mass spectrometry-based immunoassay. Food

Chemistry 230 (2017) 721-727.

[13] L. Trapiella-Alfonso, J.M. Costa-Fernández, R. Pereiro & A. Sanz-Medel.

Development of a quantum dot-based fluorescent immunoassay for

progesterone determination in bovine milk. Biosens. Bioelectronics 26

(2011) 4753-4759.

[14] W. Chen, P. Wee & I.D. Brindle. Elimination of the memory effects of gold,

mercury and silver in inductively coupled plasma atomic emission

spectroscopy. J. Anal. At. Spectrom. 15 (2000) 409-413.

Chapter 3

111

[15] F. Vanhaecke, H. Vanhoe, R. Dams & C. Vandecasteele. The use of

internal standards in ICP-MS. Talanta 39 (1992) 737-742.

[16] L.C. Huang, N. Zheng, B.Q. Zheng, F. Wen, J.B. Cheng, R.W. Han, X.M.

Xu, S.L. Li & J.Q. Wang. Simultaneous determination of aflatoxin M1,

ochratoxin A, zearalenone and -zearalenol in milk by UHPLC–MS/MS.

Food Chem. 146 (2014) 242–249.

[17] J. R. Crowter (2001) The ELISA Guidebook. (2nd ed.) New Jersey,

Humana press.

[18] S.J. Gee, T. Miyamoto, M.H. Goodrow, D. Buster & B.D. Hammock.

Development of an enzyme-linked immunosorbent assay for the analysis of

the thiocarbamate herbicide molinate. Journal of Agricultural and Food

Chemistry 36 (1988) 863–870.

[19] W. Jiang, Z. Wang, G. Nolke, J. Zhang, L. Niu & J. Shen. Simultaneous

determination of aflatoxin B1 and aflatoxin M1 in food matrices by enzyme-

linked immunosorbent assay. Food Analytical Methods 6 (2013) 767-774.

[20] Guidance for Industry, Bioanalytical Method Validation, US Department of

Health and Human Services, Food and Drug Administration Centre for Drug

Evaluation and Research (CDER), Centre for Veterinary Medicine (CVM),

May 2001.

[21] International Agency for Research on Cancer (2002). Monographs on the

evaluation of carcinogenic risks to humans. Some traditional herbal

medicines, some mycotoxins, naphthalene and styrene (82, pp. 171–174).

Lyon, France: World Health Organization (WHO).


112

[22] EC (2004). Commission Regulation, Regards Aflatoxins and Ochratoxin a

in Foods for Infants and Young Children, Official Journal of the European

Union L, 106, 5-24.

[23] EC (2006). Commission Regulation (EC) No. 1881/2006 of 19 December

2006 setting maximum levels for certain contaminants in foodstuffs. Official

Journal of the European Union L, 364, 5–24.

[24] FDA, Sec. 527.400 Whole milk, Low fat milk, skim milk – aflatoxin M1 (CPG

7106.10), FDA/ORA Compliancy Guides 2005.

[25] D. Guan, P. Li, W. Zhang, D. Zhang & J. Jiang. An ultra-sensitive

monoclonal antibody-based competitive enzyme immunoassay for aflatoxin

M1 in milk and infant milk products. Food Chemistry 125 (2011) 1359-1364.

[26] M.M. Vdovenko, C.C. Lu, F.Y. Yu & I.Y. Sakharov. Development of

ultrasensitive direct chemiluminescent enzyme immunoassay for

determination of aflatoxin M1 in milk. Food Chemistry 158 (2014) 310-314.

[27] S.C. Pei, Y.Y. Zhang, S. A. Eremin & W.J. Lee. Detection of aflatoxin M1 in

milk products from China by ELISA using monoclonal antibodies. Food

Control 20 (2009) 1080-1085.

[28] X. Tang, Z. Zhang, P. Li, Q. Zhang, J. Jiang, D. Wang & J. Lei. Sample-

pretreatment-free based high sensitive determination of aflatoxin M1 in raw

milk using a time-resolved fluorescent competitive immunochromatographic

assay. RSC Advances 5 (2015) 558-564.

[29] W.L. Chen, T.F. Hsu & C.Y. Chen. Measurement of aflatoxin M1 in milk by

utra-high-performance liquid chromatography/tandem mass spectrometry.

Journal of AOAC International 94 (2011) 872-877.

Chapter 3

113

[30] X. Wang & P. Li. Rapid screening of mycotoxins in liquid milk and milk

powder by automated size-exclusion SPE-UPLC–MS/MS and quantification

of matrix effects over the whole chromatographic run. Food Chemistry 173

(2015) 897-904.

[31] P. Diniz Andrade, J. Laine Gomez da Silva & E. Dutra Caldas.

Simultaneous analysis of afaltoxins B1, B2, G1, G2, M1 and ochratoxin A in

brest milk by high-performance liquid chromatography/fluorescende after

liquid-liquid extraction with low temperature purification (LLE-LTP). Journal

of Chromatography A 1304 (2013) 61-68.

[32] M. Hashemi & Z. Taherimaslak. Determination of aflatoxin M1 in liquid milk

using high performance liquid chromatography with fluorescence detection

after magnetic solid phase extraction. RSD Advances 4 (2014) 33497-

33506.

[33] Y. Wang, J. Dostálek & W. Knoll. Long range surface plasmon-enhanced

fluorescence spectroscopy for the detection of aflatoxin M1 in milk. Biosens.


[34] A. Vig, X. Muñoz-Berbel, A. Radoi, M. Cortina-Puig & J.L. Marty.

Characterization of the gold-catalyzed deposition of silver on graphite

screen-printed electrodes and their application to the development of

impedimetric immunosensors. Talanta 80 (2009) 942-946.

[35] M. Badea, L. Micheli, M.C. Messia, T. Candigliota, E. Marconi, T. Mottram,

M. Velasco-Garcia, D. Moscone & G. Palleschi. Aflatoxin M1 determination

in raw milk using a flow-injection immunoassay system. Anal. Chim. Acta

520 (2004) 141-148.

Chapter 4

A sensitive size-exclusion inductively

coupled plasma mass spectrometry

multiplexed assay for cancer biomarkers

using antibodies conjugated with a

lanthanide-labelled polymer

Chapter 4

117

1. Introduction.

Recently, a number of immunoassays based on the use of metal-labelled

antibodies and the determination of antibody-antigen complexes by inductively

coupled plasma mass spectrometry (ICP-MS) have been proposed for the

determination of biomolecules, and, in particular, proteins [1,2]. The ICP-MS

quantification offers several advantages over the conventional detection

techniques employed in immunoassays (colorimetry, fluorimetry, etc.), such as,

e.g., (i) specificity to heteroatom detection; (ii) compound-independent detection

sensitivity; (iii) high elemental sensitivity and dynamic range; (iv) limited sample

treatment; (v) stability of the reagents against time, temperature and light (the

isotopic masses do not change, bleach or degrade); (vi) reduction of non-

specific background; (vii) independence of analytical response from incubation

or storage times and (viii) multiplexed detection [2,3].

Antibodies have usually been labelled by either metal nanoparticles [4,5] or

lanthanides [6,7]. The advantage of elemental nanoparticles is the possibility of

the introduction of a significant number of atoms per conjugate which allows the

amplification of the analytical response. This advantage is set off by the high

affinity of nanoparticles to surface of labware and/or ICP-MS sample

introduction system, increasing wash-in and wash-out times, and by the

difficulty to synthetize nanoparticles of uniform size. Lanthanides are introduced

as DOTA or DTPA chelates due to its extraordinary thermodynamic stability [6].

The similar chemical properties make lanthanides well suited for multiplex

assays: different antibodies can be easily and specifically labelled with different

lanthanides in the same experimental workflow. Of particular interest, because

of their simplicity, are homogenous immunoassays in which a mixture of

A sensitive size-exclusion inductively coupled plasma mass spectrometry multiplexed assay for cancer biomarkers using antibodies conjugated with a lanthanide-labelled polymer

118

antibodies, each labelled with a different lanthanide, is made react within a

liquid sample and the antigen-antibody (Ag-Ab) complexes formed are isolated

and specifically determined [8]. Size-exclusion chromatography (SEC) coupled

with ICP-MS first proposed by Terenghi et al. [9] is a convenient method to

isolate the Ag-Ab complex from the excess of reagent offering: (i) multiplexed

capability; (ii) low sample amount consumption; and (iii) virtually no sample

preparation. Hann et al. [10] reported a SEC-ICPMS determination of an In-

DOTA labelled peptide Bβ15–42-antibody complex in cellular extracts. López-

Fernández et al. [11] developed a methodology for monitoring and determining

oligonucleotide sequences by means of DOTA-lanthanide labelled DNA probes

followed by SEC-ICPMS detection. Though the above-mentioned works

demonstrated the benefits of using homogeneous-based immunoassays with

ICP-MS detection, signal amplification was limited since only a single lanthanide

atom was introduced per binding site of the antibody. The number of lanthanide

atoms per antibody can be increased by using metal-loaded polymers [12],

which leads to an increase in sensitivity. This labelling strategy has been

successfully employed for single-cell ICP MS analysis [13,14,15]. Waentig et al.

[16] compared polymer-based lanthanide labelling with other lanthanide-based

labelling strategies for protein quantification in solid phase immunoassays

(Western Blot, SDS-PAGE, etc.). These authors noted that this labelling

strategy improves significantly sensitivity which results in limits of detection in

the low fmol range. However, there has been no attempt so far to investigate the

potential of antibodies conjugated with metal-labelled polymers in homogenous

assays.

Chapter 4

119

The aim of this work was to increase the sensitivity, by at least a factor of

ten, of a direct homogenous multiplexed assay for four cancer biomarkers

(CEA, sErbB2, CA 15.3 and CA 125) in human serum by -labelling the relevant

monoclonal antibodies with a different polymer-based lanthanide group. Size-

exclusion chromatography was used to isolate the antigen-antibody complexes

whereas ICP-MS on-line detection was used for quantification. The method was

benchmarked against the ones using the labelling with DOTA-chelates [9].

2. Experimental.


Carcinoembryonic antigen (CEA) was obtained from Sigma-Aldrich (St.

Quentin-Fallavier, France). The soluble form of human epidermal growth factor

receptor 2 (sErbB2) was purchased from antibodies-online (Aachen, Germany).

Cancer antigen 15.3 (CA 15.3) was obtained from MyBioSource (San Diego,

CA) and CA 125 was from Fitzgerald (MA). Goat polyclonal antimouse IgG

antibody (H&L)was purchased from Abcam (Cambridge, UK).

Mouse Immunoglobulin G subclass 1 (IgG1) antihuman monoclonal

antibody (mAb) for α-CEA (clone 1C11) and mouse IgG1 antihuman mAb for α-

CA 125 (clone X325) were purchased from Gene Tex (Irvine, CA). Mouse IgG1

antihuman mAb for α-sErbB2 (clone 5J297) was obtained from antibodies-

online (Aachen, Germany) and mouse IgG1 antihuman mAb for α-CA 15.3

(clone M002204) was from LifeSpan BioSciences (Seattle, WA). The antibody

(Ab) solutions should not contain additives, such as bovine serum albumin

(BSA) or gelatin, because the latter could be labelled as well and cause


120

interferences. Upon reception, mAb were divided into single working aliquots

and stored at -20ºC.

MAXPARTM- polymer -Ab labelling kits were obtained from Fluidigm (Les

Ulis, France). Human Albumin AlbuteinTM 20% was purchased from Grifols

Biologicals Inc. (Los Angeles, CA). 1,4,7,10 – Tetraazacyclododecane - 1,4,7 -

tris(aceticacid) - 10-maleimido - ethylacetamide (DOTA) was obtained from

Macrocyclics (Dallas, TX). Tris (2-carboxyethyl) - phosphine hydrochloride

(TCEP), TrizmaTM base, lanthanide chlorides (HoCl3, TbCl3, TmCl3, PrCl3) with

natural isotopic abundance, ammonium acetate ( 98%, for molecular biology),

monosodium phosphate, disodium phosphate, ethylenediaminetetraacetic acid

disodium salt (EDTA), dimethyl sulfoxide (DMSO), sodium chloride and

polyethylene glycol sorbitan monolaurate (Tween 20) were from Aldrich

(Schelldorf, Germany). Acetic acid glacial and 69% w w-1 nitric acid were

purchased from Panreac (Barcelona, Spain). Rare earth 100 µg mL-1 Complete

Standard was provided by Inorganic Ventures (Lakewood, Colorado) and DC™

Protein Assay Kit was from Bio-Rad (Marnes-la-Coquette, France).

Ultrapure water 18 M cm from a Milli-Q water purification system

(Millipore, Paris, France) was used throughout the work.

AmiconTM Ultra-0.5 mL centrifugal filters for DNA and protein purification

and concentration (Merck Millipore, Cork, Ireland) with different cutoff limits (3,

30 and 50 kDa) were used throughout the work for washing steps and buffers

exchange during labelling procedure of Abs using a EppenforfTM microcentrifuge

5415R (Eppendorf AG, Hamburg, Germany).

Chapter 4

121

2.2. Buffers.

The buffers used were: (a) ammonium acetate buffer (100 mM, pH 6.8 as

elution buffer and 20 mM, pH 6.0 for metal complexation), (b) phosphate buffer

(100 mM, pH 7.2, 2.5 mM EDTA) for the partial reduction of the antibody with

TCEP and (c) Tris buffer saline (20 mM Tris-HCl, 0,45% NaCl, pH 7.0 for

antibody storage media and 20 mM Tris-HCl, 0,45% NaCl, 10 mM EDTA, pH

7.0 for removing TCEP).

2.3. Serum samples.

Serum samples were provided from Hospital General Universitario of

Alicante (Alicante, Spain).

2.4. Instrumentation.

2.4.1. Size Exclusion Chromatography.

The chromatographic analyses were performed on an Agilent 1200 series

(Agilent Technologies, Santa Clara, CA) equipped with an autosampler.

Separations were carried out isocratically at 0.5 mL min−1 using 100 mM

ammonium acetate (pH 6.8) as mobile phase and sample injection volume of

100 L. Two size exclusion columns of different separation range (GE

Healthcare, Buckinghamshire, UK) were tested: a Superose 6 Increase 10/300

GL (cross-linked agarose composite stationary phase; 10 mm x 300 mm x 8.6

µm average beads size) with the approximate bed volume of 24 mL and an

optimum separation range of 5-5000 kDa for globular proteins and a Superdex

200 HR 10/300 (cross-linked agarose and dextran composite stationary phase;

10 mm x 300 mm x 8.6 µm average beads size) with the approximate bed


122

volume of 24 mL and an optimum separation range of 10-600 kDa for globular

proteins. The performance of the size exclusion column Superose 6 Increase

10/300 GL was verified with a mixture of blue dextran (Mr 2 000 kDa),

thyroglobulin (Mr 669 kDa), ferritin (Mr 440 kDa), aldolase (Mr 158 kDa),

ovalbumin (Mr 44 kDa), ribonuclease A (Mr 13.7 kDa) and ubiquitin from bovine

(Mr 8.6 kDa) using UV-Vis detection at 280 nm with baseline evaluation at 800

nm. Retention times (in minutes) plotted versus the logarithm of molecular mass

(in kDa) does not give a straight line (two straight lines were obtained: y = -

0.049x +4.104; r2 = 1 for proteins which Mr 440 kDa and y = -0.170x +7.754;

r2 = 0.998 for proteins which Mr 440 kDa). Concerning the size exclusion

column Superdex 200 HR 10/300, it was calibrated similarly to the Superose 6

Increase 10/300 GL but without using blue dextran. Retention times (in minutes)

plotted versus the logarithm of molecular mass (in kDa) gave a straight line (y =

-0.103x +4.881; r2 = 0.991).

2.4.2. Inductively Coupled Plasma Mass Spectrometry.

Detection was carried out by means ICPMS 7700x quadrupole – ICPMS

system (Agilent) equipped with a pneumatic concentric nebulizer and a double-

pass spray chamber. The connection between the exit of the column and the

nebulizer was performed directly by means of polyether ether ketone (PEEK)

tubing. The operating conditions and the nuclides measured are listed in Table

4.1. Instrumental conditions for ICP-MS were daily optimized according to the

protocol described in the user's manual. In order to evaluate the plasma

ionization conditions and the matrix load of the plasma, the 138Ba2+/138Ba+ and

156CeO+/140Ce+ signal ratios were also registered. Quantification was based on

peak areas using the Agilent ChemStation software.

Chapter 4

123

Table 4.1. Operating conditions of SEC-ICPMS.

Parameters

RF Power (W) 1500


Plasma gas 15

Auxiliary gas 0.9

Carrier gas 1.15

Carrier

Type 100 mM Ammonium acetate (pH 6.8)

Flow rate (mL min-1) 0.5


Injection volume (µL) 100

Nebulizer Pneumatic concentric

Spray chamber Double-pass Scott

Nuclides 141Pr, 159 Tb, 165Ho, 169Tm

2.5. Antibody labelling procedure.

Antibodies (Ab) have been labelled with either a lanthanide-labelled

polymer or a DOTA-lanthanide chelate complexe. Antibody labelling procedure

was based on a chemical reaction between a maleimide residue employed as a

linker of the different metal labels and free sulfhydryl groups obtained after a

partial reduction of the Ab’s cysteine-based disulfide bridges with TCEP. This

procedure was preferred over other approaches due to its lower complexity [6].


124

2.5.1. Partial reduction of the antibody.

The labelling method started with a pre-rinse of the ultrafiltration

membranes with phosphate buffer. Thereupon, a monoclonal antibody (mAb)

washing by centrifugation (1 x 500 µL, 10000 x g, 15 min, 4ºC) with phosphate

buffer and a partial reduction of the mAb using an excess of TCEP for 30 min at

370C were carried out. According to the polymer Ab labelling kit protocol, a

molar excess of 60 of TCEP relative to Ab molarity has to be used which has

been optimized for a multitude of IgG isotypes. However, it was observed that

the Ab do not show antigen selectivity in the immune reaction and, hence,

TCEP concentration was optimized. The reduction step for DOTA labelling was

carried out using a 6-fold molar excess [9]. It has to be noted that TCEP is not

particularly stable in phosphate buffers, especially at neutral pH; so the working

solutions have to be prepared immediately before use. EDTA was added to

prevent oxidation of the generated sulfhydryl groups by trace metals [16]. The

mAb was quickly washed (1 × 500 L) with Tris buffer saline to remove the

TCEP in solution by centrifugation and resuspended in the same buffer at 1 mg

mL-1. Then, the mAb was labelled following different procedures, namely: (i)

DOTA-chelate complexes or (ii) polymer labelling kit.

2.5.2. Antibody labelling via the polymer labelling kit.

The mAb was labelled following the protocol of the reagent supplied. Briefly,

the polymer was pre-loaded with a lanthanide for 30 - 40 min at 370C. Then, the

mAb was conjugated with the lanthanide - loaded polymer for 1 h at 370C. The

excess of the ligand was removed from the mAb solution by ultracentrifugation.

Chapter 4

125

2.5.3. Antibody labelling via DOTA-chelate complexes.

This labelling procedure was based on that described by Terenghi et al. [9].

Briefly, the mAb was reacted with a 50-fold molar excess of DOTA for 1 h at

370C. Then, the lanthanide (III) ion was made react with DOTA for 30 min at

370C. The excess of the ligand was removed from the mAb solution by

ultracentrifugation.

In both labelling strategies, it is important to avoid moisture condensation;

otherwise the maleimide moiety will hydrolyze and become non-reactive. The

four mAbs towards the four protein molecules chosen: CEA, sErbB2, CA 15.3,

CA 125, were labelled with the lanthanide ions: 165Ho, 159Tb, 169Tm and 141Pr,

respectively, following both labelling methods described above. Element-

labelled mAbs were stored at – 200C in Tris buffer saline until use.

2.6. Determination of the antibody labelling degree.

2.6.1. Protein quantification.

The concentrations of labelled mAbs were measured by a microplate

spectrophotometer (SPECTROstar Nano, BMG LabTech, Champigny s/Marne,

France) at 750nm using a DC™ Protein Assay Kit. The DC™ (detergent

compatible) protein assay is a colorimetric assay, similar to the well-

documented Lowry assay [17], for protein concentration following detergent

solubilization. Bovine serum albumin was used as calibration standard.

2.6.2. ICP-MS analysis of metal content.

A 0.15 µL volume of all mAbs conjugated with the labelled polymer and 3

µL of mAbs labelled with the DOTA-chelate complexes were diluted up to 5 mL

with 3.5% V/V nitric acid for the determination of the labelling degree of the


126

mAbs. An external calibration series from 1 ng mL-1 to 1 µg mL-1 was prepared

using a rare earth multielemental standard solution. Samples were analyzed by

an ICPMS 7700x quadrupole – ICPMS system using the operating conditions

listed in Table 4.1.

2.7. Immunoassay procedure.

A human serum aliquot (120 L) was incubated overnight at 40C with a

mixture of labelled mAbs 2 µg mL-1 or 10 µg mL-1, for the polymer and DOTA-

chelate labels, respectively, and subsequently, analyzed by SEC-ICPMS.

The incubation was performed at 40C in order to avoid protein degradation.


3.1. Preliminary studies with lanthanide-labelled polymer in SEC-

ICPMS.

Given that polymer-based lanthanide labels have not been tested for the

analysis of biomolecules in homogeneous-based immunoassays so far, a proof

of concept test was initially carried out to evaluate the potential benefits and

drawbacks of this labelling approach. First, following the procedure described in

the experimental section, a goat polyclonal antimouse IgG antibody (pAb) was

labelled with the 165Ho polymer reagents and analyzed by SEC-ICPMS.

Likewise, for the purpose of evaluating the results obtained, this assay was also

carried out using 165Ho DOTA chelate complexes.

Figure 4.1 shows the chromatograms obtained for a solution containing a

nominal concentration of 10 µg mL-1 pAb labelled with 165Ho polymer or 165Ho

DOTA using the Supereose 6 Increase 10/300 GL column. Irrespective of the

Chapter 4

127

labelling approach selected, two 165Ho-related peaks were approximately

obtained at 34 and 40 min. In agreement with the theoretical values expected

from the column calibration curve and UV-Vis measurements at 280 nm, the

first peak corresponds to the 165Ho-labelled pAb; whereas the second one was

identified as metal impurities from the Ab labelling procedure. In fact, the

retention time of the second peak was similar to that obtained from a solution

containing either free 165Ho polymer or free 165Ho DOTA chelate complexes. As

can be seen in Figure 4.1, in the case of using 165Ho polymer, the signal of the

labelled pAb (measured as peak height) was approximately two orders of

magnitude higher than that obtained for the 165Ho DOTA chelate complexes.

These results are totally expected taking into account that there is an average

of 30 chelators per polymer label [18]. Nevertheless, given the signal difference

between both labelling approaches, it could be concluded that the Ab labelling

efficiency achieved with the polymer reagents is at least three times higher than

that afforded by the DOTA chelate complexes.

Figure 4.1. SEC-ICPMS chromatograms of a goat polyclonal antimouse IgG

antibody (pAb) labelled with 165Ho polymer reagents (black line) and 165Ho

DOTA chelate complex (red line). pAb nominal concentration: 10 µg mL-1,

column: Superose 6 Increase 10/300 GL.

0

50

100

150

200

250

300

350

0 5 10 15 20 25 30 35 40 45

165 H

o+

inte

nsi

ty·1

0-4(c

ps)

Time (minutes)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

20 24 28 32 36 40 44

165 H

o+in

tens

ity·

10-4

(cps

)

Time (minutes)


128

Next, with the aim to verify the pAb activity and the immunocomplex

formation, a 10 µg mL-1 mouse IgG1 Ab (antigen) solution in ammonium acetate

was incubated overnight at 40C with pAb labelled with either 165Ho polymer or

165Ho DOTA chelate complex at nominal concentration of 10 µg mL-1 and, then,

the mixture obtained was analyzed by SEC-ICPMS (Figure 4.2). In the case of

using the polymer reagents (Figure 4.2 A), the elution profile shows, in addition

to those shown in Figure 4.1, two new peaks at 17 and 31 min, respectively.

Given that the separation in SEC is based on the size of the molecules as they

pass through the column, these results suggest that two different

immunocomplexes have been formed: the first peak corresponds to a high

molecular weight (HMW) immunocomplex whereas the second one to a low

molecular weight (LMW) immunocomplex. According to the retention time

observed for blue dextran (16.4 min) and thyroglobulin (26.2 min) during column

mass calibration, the size of the HMW immunocomplex might be ranged

between 2000 and 700 kDa. On the other hand, the LMW immunocomplex peak

might be related to small antigen-pAb complex given its proximity to the

unreacted pAb peak. The peak corresponding to the unreacted pAb was still

observed either because of the excess of the pAb used or because of its partial

deactivation during the labelling procedure. Interestingly, the chromatographic

profile registered for the mixture of the antigen with the 165Ho DOTA labelled

pAb (Figure 4.2 B) was different to that obtained using the 165Ho polymer-

labelled pAb. The elution profile just shows one new peak at 17 min that, in

agreement with the literature [9] and previous observations with the 165Ho

polymer-labelled pAb, should be related to a HMW immunocomplex. No peak

corresponding to other type of immunocomplexes was registered. Therefore, it

Chapter 4

129

could be concluded that the HMW immunocomplex formation is favored by the

use of 165Ho DOTA labelled pAb over the use of 165Ho polymer-labelled one.

Figure 4.2. SEC-ICPMS chromatograms obtained after incubation of a mouse

IgG1 antibody solution with a goat polyclonal antimouse IgG antibody (pAb)

labelled with (A) 165Ho polymer reagents and (B) 165Ho DOTA chelate complex.

(1) High molecular weight immunocomplex; (2) low molecular weight

immunocomplex; (3) unreacted labelled pAb; (4) free lanthanide label. pAb

nominal concentration: 10 µg mL-1; antigen concentration: 10 µg mL-1;

incubation medium: 100 mM ammonium acetate; column: Superose 6 Increase

10/300 GL.

0

50

100

150

200

250

300

0 5 10 15 20 25 30 35 40 45

165 H

o+

inte

nsi

ty·1

0-4(c

ps)

Time (minutes)

(1)

(2)

(3)

(4)

A

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 15 20 25 30 35 40 45

165 H

o+

inte

nsi

ty·1

0-4(c

ps)

Time (minutes)

(1)

(3)

(4)

B


130

The origin of this behavior could be related to steric effects caused by the

polymer chains linked to the Ab which make difficult the formation of big

antigen-antibody (Ag-Ab) aggregates. In fact, both HMW and LMW

immunocomplex signals were observed to be strongly dependent on the Ag:Ab

ratio tested. Thus, for a given antigen amount, a reduction of the Ab

concentration favours the HMW immunocomplex formation at the expense of

the LMW immunocomplex. Conversely, the LMW immunocomplex formation

was improved increasing the Ab concentration. So, when the 165Ho polymer-

label is used, either HMW or LMW immunocomplex signals could be

theoretically employed for protein quantitation purposes. Nevertheless, given

the interdependence between both immunocomplexes, the analytical figures of

merit are expected to be strongly dependent on the Ab concentration employed

in the immunoassay. The above mentioned phenomenon was not observed for

the 165Ho DOTA labelled pAb and, hence, protein quantification could only be

performed using the signal of the HMW immunocomplex [9].

Previous works in SEC showed that unwanted interactions between the

sample components and the chromatographic stationary phase could occur,

thus negatively affecting quantitative analysis [9,19,20]. For this reason,

lanthanide content emerging from the Superose 6 Increase 10/300 GL column

for both labelling strategies was compared to that initially present in the sample

before the chromatographic run. Holmium recovery using the polymer Ab

labelling kit was quantitative (106 ± 3%) but not for the DOTA chelate

complexes (70 ± 5%). The origin of the low recoveries obtained with the latter

approach was unclear. The chromatographic recovery was therefore

determined using an alternative SEC column (Superdex 200 HR 10/300) to that

Chapter 4

131

initially employed in this work (Superose 6 Increase 10/300 GL). While the

lanthanide recovery for DOTA labelling with the alternative column was

quantitative and acceptable (113 ± 13%), the peak resolution between the LMW

immunocomplex and the unreacted Ab for lanthanide-labelled polymer was

compromised. No further differences were observed in the chromatograms

between both columns. Therefore, further studies for the mentioned labelling

strategies were carried out using different SEC columns: the Superose 6

Increase 10/300 GL column for lanthanide-labelled polymer and the Superdex

200 HR 10/300 column for DOTA chelate complexes.

3.2. Analysis of cancer biomarkers in human serum by means of

SEC-ICPMS and polymer-labelled antibodies.

Once the feasibility of using the lanthanide-labelled polymer for protein

analysis in homogeneous-based immunoassays was successfully proved, this

labelling approach was applied for the multiplex determination of cancer

biomarkers in human serum samples; namely: CEA, sErbB2, CA 15.3 and CA

125. To this end, mAbs against the above-mentioned biomarkers have been

labelled with 165Ho, 159Tb, 169Tm and 141Pr, respectively.

3.2.1. Optimization of polymer-labelled antibodies synthesis.

The labelling degree of the polymer-labelled mAbs depends on the number

of sulfhydryl groups obtained after reducing the Ab’s cysteine-based disulfide

bridges with TCEP. To achieve the highest labelling efficiency, it is necessary to

reduce as many disulfide bridges of the mAb as possible. However, the

experimental conditions should not be too harsh so the labelled Ab still shows

antigen selectivity in the immune reaction. In other words, the conditions have


132

to be as mild as possible, so that the Ab is not separated into its heavy and light

chain by the breaking of too many disulfide bridges. Preferably, the disulfide

bridges of the hinge region can be cleaved resulting in two identical and still

binding Ab fragments. To evaluate both labelling efficiency and mAb’s activity, it

has been proceeded as follows. First, aliquots of the different mAbs were

reduced with a given molar excess of TCEP, respectively. Next, after labelling

the different mAb with the polymer reagent, a solution containing a nominal

concentration of 1 µg mL-1 of labelled mAb is made to react with different

amounts of their corresponding antigen (0-50 ng mL-1) in human serum. Finally,

the mixture was analyzed by SEC-ICPMS.

Initially, a molar excess rate of 60 of TCEP (concentration recommended by

the reagent supplier) relative to mAb molarity was tested but no

immunocomplexes were registered for all the mAb tested. Similar findings were

observed for 20-fold molar excess, suggesting that the reduction of the disulfide

bridges was too harsh leading to a denaturation of the mAbs. These results

were totally unexpected taking into account polymer manufacturer

recommendations and previous data reported by Waentig et al. [16]. As in the

IgG1 subclass the 2 heavy chains are connected in the hinge region by 2

disulfide bonds [21] and each disulfide bridge needs at least to be reduced 2

protons from the TCEP, the molar excess of TCEP was further reduced ranging

from 2 to 8-fold. For a molar excess of TCEP lower than or equal to 8-fold, the

chromatographs obtained showed the four peaks previously mentioned (free

lanthanide-labelled polymer, unreacted labelled mAb, LMW and HMW

immunocomplexes) thus showing that the mAbs conserved their binding

properties. The elution time for all the peaks was similar to that previously has

Chapter 4

133

pointed out. In general, with a decreasing amount of TCEP, the intensities of the

different immunocomplexes also decreased. Because of signal differences

between 4-fold and 8-fold molar excess of TCEP were lower than 5% and the

reduction step is critical in keeping mAb binding properties, the former was

finally chosen for further studies.

Finally, the labelling degree of the labelled mAbs was evaluated as

described before (section 2.6.). To this end, the metal content of the labelled

mAbs was measured by ICP-MS. In advance, the total amount of the mAb after

labelling was measured with the DC™ Protein Assay Kit in a microtiter plate

because during sample preparation and in particular during the purification step

losses can occur. The labelling degrees (n(Ln):n(Ab)), which correspond to the

number of lanthanide atoms labelled to the mAb, are shown in Table 4.2.

Table 4.2. Labelling degree of the different mAbs using polymer-reagents and

DOTA-chelate complexes.

CEA (165Ho) sErbB2 (159Tb) CA 15.3 (169Tm) CA 125 (141Pr)

Polymer reagents 25 31 36 28

DOTA-chelate

complexes 0.2 0.11 0.3 0.09

On average, there were 29 lanthanides per mAb and, considering that the

lanthanide-labelled polymer contains an average of 30 chelators per label [18],

it points out that almost one polymer label is attached to each Ab. These values

are about 6 times lower than those reported elsewhere [16] but it should be

taking into account that the molar excess of TCEP employed for the partial


134

reduction of the mAb in this work was 15 times lower. For the sake of

comparison, the mAbs were also labelled with DOTA-chelate complexes. The

experimental conditions selected were those previously described by Terenghi

et al. [9] where a 6-fold molar excess of TCEP (with regard to the Ab) was used

for the partial reduction of the mAbs. Data gathered in Table 4.2 clearly show

that the use of DOTA-chelate complexes was a less efficient approach for mAb

labelling. In agreement with previous works [16], approximately every thirtieth

Ab was modified with SCN-DOTA which covalently bound to amino groups.

From these experiments, and considering the differences in the lanthanide

content, better analytical figures of merit should be expected for the lanthanide-

labelled polymer.

3.2.2. Influence of the incubation medium on immunocomplex

formation.

Thereupon different solution media were evaluated for incubating the

polymer-labelled mAbs with the biomarkers. Previous works [22] have shown

that nonspecific proteins may assist the formation and stabilization of Ag-Ab

complexes by maintaining the correct conformation of the Ab and antigen for

optimum binding. For this purpose, a solution containing a nominal

concentration of 1 µg mL-1 of the polymer-labelled mAbs was incubated

overnight at 40C with the maximum concentrations of wished-to be determined

antigens (namely: 50 ng mL-1 of CEA, 100 ng mL-1 of sErbB2, 100 IU mL-1 of

CA 15.3 and 100 IU mL-1 CA 125) in the pertinent incubation medium. The

resulting mixture was subsequently analyzed by SEC-ICPMS. The incubation

media tested were: (i) 100 mM ammonium acetate (SEC carrier); (ii) 0.1 % w w-

1 Tween 20; (iii) 6% w w-1 human serum albumin; and (iv) human serum. In this

Chapter 4

135

experiment, the antigen and the mAb concentration was modified regarding

previous sections. The antigen concentration was selected according to the

concentration range of interest in clinical sample analyses whereas the mAb

nominal concentration was decreased 10-fold due to the high signals afforded

by the polymer-labelled Abs and the low biomarker concentration tested.

As expected, regardless of the biomarker, HMW and LMW

immunocomplexes were observed using 0.1 % w w-1 Tween 20, 6% w w-1

human serum albumin or human serum as incubation medium. No detectable

immunocomplex signal was obtained for ammonium acetate despite this

medium was successfully employed in the preliminary studies. Table 4.3

summarizes the experimental data obtained for CEA with the different

incubation media tested. From these data, it was concluded that, given the low

levels of the biomarkers expected in human serum samples, the incubation

medium should contain surfactants and/or proteins to favor immunocomplex

formation [9,22]. In fact, the absence of both HMW and LMW

immunocomplexes signals with ammonium acetate could be probably attributed

to the low levels of the biomarkers tested and the incubation medium

inefficiency to stabilize the Ab and the Ag-Ab complexes.

Human serum from a healthy person contains significant levels of all the

biomarkers studied (CEA, sErbB2, CA 15.3 and CA 125) and, hence, the

concentration values obtained for unknown human serum samples have been

relative to their content in the control human serum employed in the incubation

step. While this situation is not the ideal from an analytical point of view, it

should not be especially troublesome for clinical sample analyses since its main

interest is focused on status changes from reference range concentrations.

A sen

sitive size-exclusio

n in

du

ctively cou

pled

plasm

a mass sp

ectrom

etry mu

ltiplexed

assay for can

cer bio

markers u

sing

antib

od

ies con

jug

ated

with

a lanth

anid

e-labelled

po

lymer

Tab

le 4.3. Influ

en

ce

of th

e in

cub

atio

n m

ed

ium

on

the H

MW

an

d L

MW

imm

un

oco

mp

lexe

s in

teg

rate

d s

ign

als

ob

tain

ed

for C

EA

.

An

tibod

y n

om

ina

l co

nce

ntra

tion

: 1 µ

g m

L-1; C

EA

con

ce

ntra

tion: 5

0 n

g m

L-1 (m

ea

n ±

t·s·n

1/2, n

= 3

, P =

95

%).

C

EA

con

centratio

n (n

g m

L-1)

Imm

un

oco

mp

lex A

mm

on

ium

acetate T

ween

20

(0.1% w

w-1)

Hu

man

serum

albu

min

(6% w

w-1)

Hu

man

serum

HM

W

No

t dete

cte

d

35

00

± 6

00

33

00

± 4

00

91

00

± 1

300

LM

W

No

t dete

cte

d

11

00

00

± 8

00

0

98

00

0 ±

900

0

24

50

00

± 1

00

00

Chapter 4

137

Obviously, this makes imperative to use a control human serum with a known

concentration of all the biomarkers. In this work, a pooled serum, prepared from

15 healthy patients with a declared amount of tumor biomarkers determined

with the conventional heterogeneous immunoassays usually employed in the

clinical analytical laboratories, was used. The concentration levels for all the

biomarkers studied in the control human serum were: 1.7 ng mL-1 CEA, 7 ng

mL-1 sErbB2, 13 IU mL-1 15 IU mL-1 CA 15.3 and CA 125.

3.3.3. Optimization of the concentration of the polymer-labelled

antibody.

As it has been pointed out (section 3.1. and elsewhere [23]), the Ag:Ab ratio

employed in the immune reaction determines which types of immunocomplexes

are formed. To investigate this effect in detail, two types of experiments were

carried out. First, a human serum sample containing a fixed amount of each

biomarker was incubated with variable amounts of the corresponding polymer-

labelled mAb. Alternatively, the concentration of the polymer-labelled mAb was

fixed and the biomarker concentration was modified.

Figure 4.3 shows the chromatograms obtained after incubation overnight at

40C of a human serum sample spiked with 50 ng mL-1 CEA and with the

corresponding 165Ho polymer-labelled mAb at a nominal concentration of 6 ng

mL-1 or 2 µg mL-1. As expected, the Ag:Ab ratio employed was critical on

immunocomplex formation. Thus, incubating the antigen with the polymer-

labelled mAb at a nominal concentration of 6 ng mL-1, just the HMW

immunocomplex was formed and no LMW immunocomplex signal was


138

detectable. The opposite behavior was observed for the polymer-labelled mAb

at a nominal concentration of 2 µg mL-1.


serum sample spiked with 50 ng mL-1 CEA and with its corresponding 165Ho

polymer labelled mAb at a nominal concentration of: (A) 6 ng mL-1 or (B) 2 µg

mL-1. Column: Superose 6 Increase 10/300 GL.

Alternatively, human serum samples containing concentrations from 5 to 50

ng mL-1 of CEA were incubated with the corresponding 165Ho polymer-labelled

mAb at the nominal concentrations of 6 ng mL-1 or 2 µg mL-1 (Table 4.4).

0

0.5

1

1.5

2

2.5

0 10 20 30 40

165 H

o+

inte

nsi

ty·1

0-3(c

ps)

Time (minutes)

A

0

100

200

300

400

500

600

0 10 20 30 40

165 H

o+

inte

nsi

ty·1

0-3(c

ps)

Time (minutes)

B

Chapter 4

139

Table 4.4. Influence of the CEA concentration on the immunocomplexes

integrated signals after incubation with 165Ho polymer-labelled mAb at nominal

concentrations of 6 ng mL-1 or 2 µg mL-1. Incubation medium: human serum.

(mean ± t·s·n1/2, n = 3, P = 95%).

CEA concentration (ng mL-1)

Immunocomplex 5 15 30 50

HMW$ 34100±200 36000±600 34100±200 32000±700

LMW& 4391000±2000 4624000±5000 5089000±2000 5589000±3000

$ 6 ng mL

-1 polymer-labelled mAb;

& 2 µg mL

-1 polymer-labelled mAb.

Interestingly, the HMW immunocomplex signal did not increase at increasing

antigen concentration when the polymer-labelled mAb nominal concentration

was 6 ng mL-1. Nevertheless, the LMW immunocomplex signal did show a

linearly increased response for a polymer-labelled mAb nominal concentration

of 2 µg mL-1. The fact that, in the former case, the assay dose response had a

maximum is related to the Hook effect [24] and it is caused by excessively high

concentrations of antigen saturating all of the available binding sites of the Ab

without forming complexes. Consequently, the immunocomplex formation is not

favored and the SEC-ICPMS signal decreases instead of increasing. This

phenomenon is common in one-step immunometric assays, as the one

developed in this work, affecting negatively the dynamic linear range. The Hook

effect can be mitigated by either decreasing the amount of antigen or increasing

the concentration of the Ab. From a practical point of view, the only feasible

approach to deal with this problem is to modify the concentration of the

polymer-labelled mAb. However, as indicated above, when the concentration of


140

the polymer-labelled mAb was increased, the LMW immunocomplex was clearly

favored over the HMW one. As a result, the use of the HMW immunocomplex

signal for quantitative purposes must be discarded in favor of the LMW

immunocomplex signal. No Hook effect was observed when the LMW

immunocomplex signal was used for quantification since the Ag-Ab reaction did

not go into antigen excess. These findings were similar for all the biomarkers

studied (Figure 4.4) and, hence, the mAb nominal concentration was set at 2 µg

mL-1 for further studies.

At this point, it is interesting to compare the above-mentioned findings with

experimental data obtained for DOTA labelled mAbs. In agreement with

Terenghi et al. [9] observations, no Hook effect was observed for biomarker

quantification using the HMW immunocomplex signal. This behavior is

explained considering that optimum mAb nominal concentration used (10 µg

mL-1) was 5-fold higher than that using reagents due to the lower signal

amplification afforded by DOTA-chelate complexes.

Chapter 4

141


serum sample spiked with 50 ng mL-1 of sErbB2, CA 15.3 or CA 125 antigen

with its corresponding polymer-labelled antibody at a nominal concentration of 2

µg mL-1. Column: Superose 6 Increase 10/300 GL.

0

100

200

300

400

500

600

0 10 20 30 40

141 T

b+

inte

nsi

ty·1

0-3(c

ps)

Time (minutes)

A

0

100

200

300

400

500

600

0 10 20 30 40

169 T

m+

inte

nsi

ty·1

0-3(c

ps)

Time (minutes)

B

0

100

200

300

400

500

600

0 10 20 30 40

141 P

r+in

ten

sity

·10-3

(cp

s)

Time (minutes)

C


142

3.3.4. Figures of merit.

Due to the lack of a certified biomarker reference material for CEA, sErbB2,

CA 15.3 and CA 125 antigen, method accuracy was evaluated by means of a

recovery test. To this end, a control human serum was spiked with the four

tumor biomarkers at three different known concentration levels. This assay was

performed using the optimum operating conditions described in previous

sections. Table 4.5 shows the average of the recovery values of the multiplexed

method developed for the 4 biomarkers.

Table 4.5. Recovery values for the CEA, sErbB2, CA 15.3 and CA 125

biomakers of interest using polymer labelling kit (mean ± t·s·n1/2, n = 3, P =

95%).

Antigen [Antigen]spiked [Antigen]calc Recovery (%)

CEA

3.5 ng mL-1 3.4 ± 0.4 ng mL-1 103 ± 11

16 ng mL-1 16.6 ± 0.3 ng mL-1 96 ± 2

32 ng mL-1 33 ± 2 ng mL-1 98 ± 5

sErbB2

4 ng mL-1 4.0 ± 0.2 ng mL-1 101 ± 6

20 ng mL-1 19 ± 1.2 ng mL-1 105 ± 7

60 ng mL-1 57 ± 3 ng mL-1 106 ± 6

CA 15.3

11 IU mL-1 11.6 ± 0.4 IU mL-1 95 ± 3

40 IU mL-1 41 ± 5 IU mL-1 98 ± 3

68 IU mL-1 69 ± 4 IU mL-1 98 ± 5

CA 125

10 IU mL-1 10.2 ± 0.8 IU mL-1 98 ± 8

35 IU mL-1 34.8 ± 0.5 IU mL-1 100 ± 2

65 IU mL-1 67 ± 2 IU mL-1 97 ± 3

For all the biomarkers studied, the recovery values were quantitative ranging

from 90% to 114%. The repeatability (intra-assay precision) of the method was

Chapter 4

143

determined by analyzing four replicates of each sample on the same day. The

RSD of the biomarkers concentration levels was in the 0.6 - 4% range.

Reproducibility (inter-assay precision) was also verified by analyzing the spiked

human serum samples in four different days with RSDs ranging from 4 to 8%.

The limit of detection (LoD), sensitivity (defined as the slope of the

calibration curve) and linear dynamic range are given in Table 4.6. Given that

the control human serum employed as incubation medium (blank solution) is not

antigen free, a theoretical LoD was roughly estimated dividing 3 times the

standard deviation of the instrument response for blank matrix by the slope of

the calibration curve, from the linear regression analysis [25]. For the sake of

comparison, these parameters were also calculated for the DOTA labelling

approach. In this case, since no LMW immunocomplex signal was observed,

the calibration was carried out using the signal of the HMW immunocomplex. In

general, the sensitivity and the LoDs obtained using polymer reagents were

improved 10-fold (on average) regarding the DOTA labelling. These results

were lower than expected according to the differences in the labelling degree

between both approaches. It should be considered that both unreacted labelled

mAb and LMW immunocomplex were not well-resolved in the chromatogram

and, hence, signal reproducibility for low biomarker concentrations was partially

compromised. As regards linear dynamic range, this parameter was also

improved by 10-fold using polymer reagents. From data shown in Table 4.6,

and despite the low chromatographic resolution, there is no doubt that polymer–

labelling significantly improves the analytical figures of merit of the previous

labelling approach employed in ICPMS homogeneous-based immunoassays for

biomolecules analysis. Moreover, it is worth to mention that the concentration of


144

the polymer-labelled mAb required in the immunoreaction is decreased 5-fold to

that required using DOTA-labelled mAbs. As a consequence, a considerable

reduction of the cost per analysis is obtained using polymer reagents.

3.3.5. Comparison with other methodologies.

Analytical figures of merit of CEA, sErbB2, CA 15.3 and CA 125 analysis

using polymer labelled mAbs and SEC-ICPMS have been compared with those

previously reported in the literature for other immunoassay-based

methodologies (Table 4.7).

In general, the analysis of these biomarkers is usually carried out by

heterogeneous-based immunoassays which have wider dynamic range but are

time consuming and costly. The major advantage of the proposed methodology

is the possibility to analyze simultaneously 4 biomarkers. In fact, comparing to

commercial sandwich ELISA spectrophotometric kits, detection limits are

improved approximately 10-fold.

4. Conclusions.

This work shows that lanthanide-labelled polymers conjugated with

antibodies can be successfully employed for multiplexed biomarkers analysis

using a homogeneous-based immunoassay and SEC-ICPMS detection. This

new approach improves detection limits 10-fold regarding the lanthanide-DOTA

complex traditionally employed for antibody conjugation.

The established method affords lower LODs than those obtained by commercial

sandwich ELISA spectrophotometric kits for CEA, s ErbB2, CA 15.3 and CA

125 determination as well as higher sample throughput and lower operative

cost.

Ch

apte

r 4

Tab

le 4

.6.

Se

nsitiv

ity,

Lo

Ds a

nd

lin

ea

r d

yn

am

ic r

an

ge

of

CE

A,

sE

rbB

2,

CA

15

.3 a

nd

CA

12

5 a

na

lysis

usin

g b

oth

po

lym

er

lab

elli

ng

kit a

nd D

OT

A-c

he

late

co

mp

lexe

s *

(me

an

± t

·s·n

1/2

, n

= 3

, P

= 9

5%

).

P

oly

mer

lab

ellin

g k

it

DO

TA

-ch

elat

e co

mp

lexe

s

Bio

mar

kers

S

ensi

tivi

ty*

LO

D

Lin

ear

dyn

amic

ran

ge

Sen

siti

vity

* L

OD

L

inea

r d

ynam

ic r

ang

e

CE

A

(4.3

±0.3

)·1

05

0.1

2 n

g m

L-1

0

.4 - 5

0 n

g m

L-1

(4

.2 ±

0.6

)·1

04

3 n

g m

L-1

1

1 -

50

ng m

L-1

sErb

B2

(5.8

± 0

.4)·

10

4

0.5

ng m

L-1

2

- 1

00

ng m

L-1

(0

.8 ±

0.2

)·1

04

8 n

g m

L-1

2

7 -

100

ng m

L-1

CA

15.

3 (1

.0 ±

0.4

)·1

05

0.6

IU

mL

-1

2 -

10

0 IU

mL

-1

(0.3

2 ±

0.0

2)·

10

5

5 I

U m

L-1

1

8 -1

00 I

U m

L-1

CA

125

(1

.13

± 0

.08

)·1

06

0.5

IU

mL

-1

2 -

10

0 IU

mL

-1

(0.4

4 ±

0.0

6)·

10

6

7 I

U m

L-1

2

4 -

100

IU

mL

-1

A sen

sitive size-exclusio

n in

du

ctively cou

pled

plasm

a mass sp

ectrom

etry mu

ltiplexed

assay for can

cer bio

markers u

sing

antib

od

ies con

jug

ated

with

a lanth

anid

e-labelled

po

lymer

Tab

le 4.7. Co

mp

aris

on

of d

iffere

nt m

eth

od

s fo

r CE

A, s

Erb

B2

, CA

15

.3 a

nd

CA

12

5 a

na

lysis

.

An

alytical meth

od

T

arget p

rotein

C

on

centratio

n ran

ge

LO

D

Referen

ce

Com

mercial E

LISA

kit C

EA

0.343 - 250 ng m

L-1

0.2 ng mL

-1 A

bcam (C

ambridge, U

.K.)

Com

mercial E

LISA

kit sE

rbB2

0.819 - 200 ng mL

-1 0.8 ng m

L-1

Abcam

(Cam

bridge, U.K

.)

Com

mercial E

LISA

kit C

A 15.3

5 - 240 UI m

L-1

5 UI m

L-1

Abcam

(Cam

bridge, U.K

.)

Com

mercial E

LISA

kit C

A 125

5 - 400 UI m

L-1

5 UI m

L-1

Abcam

(Cam

bridge, U.K

.)

Chem

iluminescent im

munoassay

CE

A

0.5 - 100 ng mL

-1 0.12 ng m

L-1

[26]

ICP

-MS

based

magnetic

imm

unoassay C

EA

0.2 - 50 ng m

L−

1 0.05 ng m

L-1

[27]

Chip-based

magnetic

imm

unoassay-

ET

V-IC

P-M

S

CE

A

0.2 - 50 ng mL

-1 0.06 ng m

L-1

[28]

Am

perometric m

agnetoimm

unosensor sE

rbB2

0.1 - 32.0 ng mL

-1 0.03 ng m

L-1

[29]

Gold

nanorod-based plasm

onic

sensor C

A 15.3

0.0249 - 0.2387 UI m

L-1

- [30]

Optical m

icroresonators C

A 125

Limit of linearity of 10 U

I mL

-1 ~ 1.8 U

I mL

-1 [31]

Fluorescence spectroscopy

CA

125 Lim

it of linearity of 500 UI m

L-1

0.26 UI m

L-1

[32]

Chapter 4

147

5. References.

[1] J. Bettmer, M. Montes-Bayón, J.R. Encinar, M.L. Fernández Sánchez, M.R.

Fernández de la Campa & A. Sanz-Medel. The emerging role of ICP-MS in

proteomic analysis. J. Proteoms. 72 (2009) 989-1005.



Acta Part B 76 (2012) 27-39.


spectrometry based immunoassay: a review. Mass Spectrom. 33 (2014)

373-393.

[4] Z. Liu, B. Yang, B. Chen, M. He & B. Hu. Upconversion nanoparticle as

elemental tag for the determination of alpha-fetoprotein in human serum by

inductively coupled plasma mass spectrometry. Analyst 142 (2017) 197-

205.

[5] E. Pérez, P. Martínez-Peinado, F. Marco, L. Gras, J.M. Sempere, J. Mora &

G. Grindlay. Determination of aflatoxin M1 in milk samples by means of an

inductively coupled plasma mass spectrometry-based immunoassay. Food

Chemistry 230 (2017) 721-727.

[6] G. Schwarz, L. Mueller, S. Beck & M.W. Linscheid. DOTA based metal

labels for protein quantification: a review. J. Anal. At. Spectrom. 29 (2014)

221-233.

[7] L. Waentig, P.H. Roos & N. Jakubowski. Labelling of antibodies and

detection by laser ablation inductively coupled plasma mass spectrometry.

PART III. Optimization of antibody labeling for application in a Western blot

procedure. J. Anal. At. Spectrom. 24 (2009) 924-933.


148




98-110.

[9] M. Terenghi, L. Elviri, M. Careri, A. Mangia & R. Lobinski. Multiplexed

determination of protein biomarkers using metal-tagged antibodies and size

exclusion chromatography−inductively coupled plasma mass spectrometry.

Anal. Chem. 81 (2009) 9440-9448.

[10] S. Hann, T. Falta, K. Boeck, M. Sulyok & G. Koellensperger. On-line fast

column switching SEC × IC separation combined with ICP-MS detection for

mapping metallodrug-biomolecule interaction. J. Anal. At. Spectrom. 25

(2010) 861-866.

[11] L. López-Fernández, E. Blanco-González & J. Bettmer. Determination of

specific DNA sequences and their hybridisation processes by elemental

labelling followed by SEC-ICP-MS detection. Analyst 139 (2014) 3423-

3428.

[12] X. Lou, G. Zhang, I. Herrera, R. Kinach, O. Ornatsky, V. Baranov, M. Nitz &

M.A. Winnik. Polymer-based elemental tags for sensitive bioassays. Angew.

Chem. 119 (2007) 6111-6114.

[13] E. Razumienko, O. Ornatsky, R. Kinach, M. Milyavsky, E. Lechman, V.

Baranov, M.A. Winnik & S.D. Tanner. Element-tagged immunoassay with

ICP-MS detection: Evaluation and comparison to conventional

immunoassays. Journal of Immunological Methods 336 (2008) 56-63.

[14] D.R. Bandura, V.I. Baranov, O.I. Ornatsky, A. Antonov, R. Kinach, X.D.

Lou, S. Pavlov, S. Vorobiev, J.E. Dick & S.D. Tanner. Mass cytometry:

Chapter 4

149

Technique for real time single cell multitarget immunoassay based on

inductively coupled plasma time-of-flight mass spectrometry. Anal. Chem.

81 (2009) 6813-6822.

[15] L. Mueller, A.J. Herrmann, S. Techritz, U. Panne & N. Jakuwbowski.

Quantitative characterization of single cells by use of immunocytochemistry

combined with multiplex LA-ICP-MS. Anal. Bioanal. Chem. 409 (2017)

3667-3676.

[16] L. Waentig, N. Jakubowski, S. Hardt, C. Scheler, P.H. Roos & M.W.

Linscheid. Comparison of different chelates for lanthanide labeling of

antibodies and application in a Western blot immunoassay combined with

detection by laser ablation (LA-)ICP-MS. J. Anal. At. Spectrom. 27 (2012)

1311-132.

[17] G.L. Peterson. Review of the folin phenol protein quantitation method of

lowry, rosebrough, farr and randall. Analytical Biochemistry 100 (1979) 201-

220.

[18] S.D. Tanner, D.R. Bandura, O. Ornatsky, V.I. Baranov, M. Nitz & M.A.

Winnik. Flow cytometer with mass spectrometer detection for massively

multiplexed single-cell biomarker assay. Pure Appl. Chem. 80 (2008) 2627-

2641.

[19] S. Mounicou, J. Szpunar, R. Lobinski, D. Andrey & C.J. Blake.

Bioavailability of cadmium and lead in cocoa: Comparison of extraction

procedures prior to size-exclusion fast-flow liquid chromatography with

inductively coupled plasma mass spectrometric detection (SEC-ICP-MS). J.

Anal. At. Spectrom. 17 (2002) 880-886.


150

[20] V. Vacchina, K. Polec & J. Szpunar. Speciation of cadmium in plant tissues

by size-exclusion chromatography with ICP-MS detection. J. Anal. At.

Spectrom. 14 (1999) 1557-1566.

[21] H. Liu & K. May. Disulfide bond structures of IgG molecules: Structural

variations, chemical modifications and possible impacts to stability and

biological function. mAbs 4 (2012) 17-23.

[22] H. Wang, M. Lu, M. Weinfeld & X.C. Le. Enhancement of immunocomplex

detection and application to assays for DNA adduct of benzo[α]pyrene.

Anal. Chem. 75 (2003) 247-254.

[23] A.K. Abbas, A.H. Lichtman & S. Pillai. Cellular and Molecular Immunology,

eighth ed., Elsevier Saunders, 2015.

[24] C. Selby. Interference in immunoassay. Ann Clin Biochem. 36 (1999) 704-

721.

[25] A. Shrivastava & V.B. Gupta. Methods for the determination of limit of

detection and limit of quantitation of the analytical methods. Chronicles of

Young Scientists 2 (2011) 21-25.

[26] X. Pei, B. Chen, L. Li, F. Gao & Z. Jiang. Multiplex tumor marker detection

with new chemiluminescent immunoassay based on silica colloidal crystal

beads. Analyst 135 (2010) 177-181.

[27] X. Zhang, B. Chen, M. He, Y. Zhang, G. Xiao & B. Hu. Magnetic

immunoassay coupled with inductively coupled plasma mass spectrometry

for simultaneous quantification of alpha-fetoprotein and carcinoembryonic

antigen in human serum. Spectrochimica Acta Part B 106 (2015) 20-27.

[28] B. Chen, B. Hu, P. Jiang, M. He, H. Peng & X. Zhang. Nanoparticle

labelling-based magnetic immunoassay on chip combined with

Chapter 4

151

electrothermal vaporization - Inductively coupled plasma mass spectrometry

for the determination of carcinoembryonic antigen in human serum. Analyst

136 (2011) 3934-3942.

[29] U. Eletxigerra, J. Martinez-Perdiguero, S. Merino, R. Barderas, R.M.

Torrente-Rodríguez, R. Villalonga, J.M. Pingarrón & S. Campuzano.

Amperometric magnetoimmunosensor for ErbB2 breast cancer biomarker

determination in human serum, cell lysates and intact breast cancer cells.

Biosensors and Bioelectronics 7 (2015) 34-41.

[30] S. Chen, Q. Zhao, L. Zhang, L. Wang, Y. Zeng & H. Huang. Combined

detection of breast cancer biomarkers based on plasmonicsensor of gold

nanorods. Sensors and Actuators B 221 (2015) 1391-1397.

[31] M.K. Pal, M. Rashid & M. Bisht. Multiplexed magnetic nanoparticle-antibody

conjugates (MNPs-ABS) based prognostic detection of ovarian cancer

biomarkers, CA-125, -2M and ApoA1 using fluorescence spectroscopy

with comparison of surface plasmon resonance (SPR) analysis. Biosensors

and Bioelectronics 73 (2015) 146-152.

[32] H.A. Huckabay, S.M. Wildgen & R.C. Dunn. Label-free detection of ovarian

cancer biomarkers using whispering gallery mode imaging. Biosensors and


Capítulo 5

Conclusiones generales

Capítulo 5

155

Los resultados obtenidos en la presente Tesis Doctoral demuestran el

enorme potencial que presenta la Espectrometría de Masas mediante

ionización en Plasma Acoplado Inductivamente como técnica de análisis de

biomoléculas. La selección adecuada del formato de inmunoensayo y la

correcta optimización de las condiciones de trabajo son el punto crítico que

determinará los parámetros analíticos del método desarrollado. Así:

- Es posible determinar haptenos a niveles de ultratraza (sub-ppt)

empleando inmunoensayos competitivos.

- En general, para los haptenos, se obtienen mejores resultados analíticos

cuando el inmunoensayo competitivo emplea como especie trazadora

anticuerpos en lugar del analito conjugado a proteínas.

- El empleo de polímeros de lantánidos, como estrategia de marcaje en

inmunoensayos en fase homogénea, permite llevar a cabo (mediante la

adecuada separación cromatográfica) el análisis multiparamétrico de

biomoléculas mediante ICP-MS con elevada sensibilidad, precisión y

exactitud.

Capítulo 6

Futuros estudios

Capítulo 6

159

La gran mayoría de trabajos referentes a la determinación de biomoléculas

mediante inmunoensayos con detección por medio de ICP-MS, llevan a cabo la

cuantificación de las biomoléculas mediante el registro continuo de la señal que

proporciona el heteroátomo. Cuando se trabaja con nanopartículas, también se

puede utilizar el modo single particle para relacionar la concentración de analito

con la frecuencia de pulsos que provocan las nanopartículas. Los resultados

muestran que los límites de detección entre ambos métodos son similares

cuando se trabajan con inmunoensayos. No obstante, una revisión de la

(escasa) bibliografía en este campo pone de manifiesto que los límites de

detección mediante single particle podrían ser mejorados respecto al modo

convencional mediante la selección adecuada del inmunoensayo, el disolvente

con el que se introducen las nanopartículas en el ICP-MS y el sistema de

introducción de muestras. Otro aspecto a tener en cuenta es que, a pesar de la

capacidad de análisis multicomponente del ICP-MS, en muy contadas

ocasiones se aprovecha dicho potencial para el análisis de haptenos. Otro de

los problemas a resolver en cuanto a la cuantificación es la variabilidad en el

grado de funcionalización del anticuerpo con el heteroátomo. Ello dificulta la

cuantificación del analito de forma exacta y reproducible e impide el empleo de

algunas de las técnicas de calibración que se pueden emplear en ICP-MS (e.g.

dilución isotópica).

development of new methodologies for plasma mass...

Documents