Академический Документы
Профессиональный Документы
Культура Документы
Please note
Your article will be published Online First approximately one week after receipt of
your corrected proofs. This is the official first publication citable with the DOI.
Further changes are, therefore, not possible.
After online publication, subscribers (personal/institutional) to this journal will have
access to the complete article via the DOI using the URL:
http://dx.doi.org/10.1007/s00216-010-3975-2
If you would like to know when your article has been published online, take advantage
of our free alert service. For registration and further information, go to:
http://www.springerlink.com.
Due to the electronic nature of the procedure, the manuscript and the original figures
will only be returned to you on special request. When you return your corrections,
please inform us, if you would like to have these documents returned.
The printed version will follow in a forthcoming issue.
To: Springer Customer Support 1
E-mail: CorrAdmin1@spi-bpo.com
Fax: +1-703-5621873
SPi
Re: SPi Building, Sacsac Bacong
Oriental Negros 6216
Philippines
Analytical and Bioanalytical Chemistry DOI 10.1007/s00216-010-3975-2
Rapid detection of Aspergillus flavus in rice using biofunctionalized carbon nanotube field effect
transistors
Villamizar · Maroto · Rius
Permission to publish
I have checked the proofs of my article and
file://C:\WMS\Springer\Metadata2PDF\temp\HAB03975.htm 7/6/2010
Springer Metadata to PDF File Page 2 of 2
AUTHOR'S PROOF!
file://C:\WMS\Springer\Metadata2PDF\temp\HAB03975.htm 7/6/2010
AUTHOR'S PROOF! JrnlID 216_ArtID 3975_Proof# 1 - 06/07/2010
F
8 Received: 3 February 2010 / Revised: 8 June 2010 / Accepted: 28 June 2010
O
9 # Springer-Verlag 2010
O
10
11 Abstract In the present study, we have used carbon Introduction 31
PR
12 nanotube field effect transistors (FET) that have been
13 functionalized with protein G and IgG to detect Aspergillus Aspergillus flavus is a mycotoxigenic and filamentous 32
14 flavus in contaminated milled rice. The adsorbed protein G D mould widely distributed in nature on a variety of food 33
15 on the carbon nanotubes walls enables the IgG anti- and agricultural products. It causes a wide range of 34
16 35
TE
Aspergillus antibodies to be well oriented and therefore to diseases, ranging from hypersensitive reactions to invasive
17 display full antigen binding capacity for fungal antigens. A infections associated with angioinvasion. A. flavus is after 36
18 solution of Tween 20 and gelatine was used as an effective Aspergillus fumigatus, the second leading cause of invasive 37
EC
19 blocking agent to prevent the non-specific binding of the and non-invasive aspergillosis [1]. In addition, it is one of 38
20 antibodies and other moulds and also to protect the the most significant fungi in the spoilage of grain during 39
21 transducer against the interferences present in the rice storage and is therefore responsible for the economic 40
R
22 samples. Our FET devices were able to detect at least devaluation of the grain through mycotoxin contamination 41
R
23 10 μg/g of A. flavus in only 30 min. To evaluate the [2]. Rice is one of the most important staple foods for a 42
24 selectivity of our biosensors, Fusarium oxysporum and large part of the world's human population. It can grow in 43
O
25 Penicillium chrysogenum were tested as potential competing different agro-climatic conditions and is vulnerable to 44
C
26 moulds for A. flavus. We have proved that our devices are Aspergillus infection in the field as well as in storage. 45
27 highly selective tools for detecting mycotoxigenic moulds at It is difficult to develop a general method for detecting 46
N
28 low concentrations in real samples. all mycotoxins because they are metabolites that display 47
U
65 the aflatoxin biosynthetic pathway, to monitor and moulds and also to protect the transducer against the 118
66 quantify Aspergillus section Flavi population in peanuts. interferences present in rice samples. The sensor 119
67 Sensitivity tests demonstrated that the test can detect DNA showed high selectivity in the presence of competing 120
68 amounts accounting for a single conidium of Aspergillus moulds. Our devices are label free and able to detect 121
69 parasiticus. It is a rapid, sensitive and specific assay, not 10 μg/g of A. flavus with a time response of 30 min. All 122
70 inhibited by matrix effects. However, it requires the use of these performance parameters are clearly better than those 123
71 a convenient DNA extraction procedure in order to obtain of the methods currently used to detect mycotoxigenic 124
72 reliable results; as a consequence, the method is time fungi. 125
73 consuming.
74 Immunological methods like enzyme-linked immuno-
75 sorbent assay (ELISA) rely on the specific binding of an Materials and methods 126
76 antibody to an antigen. This technique is based on the
77 detection of mycelial antigens or extracellular antigens Proteins and biochemicals 127
78 secreted by moulds with detection range of 1–100 μg/mL
F
79 [3]. However, they require the use of specific labels. Anti-Aspergillus (i.e. a polyclonal rabbit anti-Aspergillus; 128
80 Electronic noses are used to detect volatiles and odours 1 mg/mL) was purchased from Oxford Biotechnology Ltd. 129
O
81 produced by the fungi [9–12]. These results can subsequently (Oxford, UK). It was dissolved in PBS to a final 130
O
82 be correlated with certain parameters such as mycelia concentration of 8 μg/mL (pH=7.2) and stored at −20 °C 131
83 development on cereal grain or mycotoxin production [9]. until use. Protein G from Streptococcus sp. recombinant, 132
PR
84 This methodology is able to distinguish between infected expressed in Escherichia coli (1 mg/mL), PBS (pH 7.4), 133
85 and non-infected samples and can sometimes even distinguish Tween 20 and gelatin from cold-water fish skin were 134
86 between non-mycotoxigenic and mycotoxigenic fungi. How- obtained from Sigma-Aldrich. Sabouraud dextrose agar
D 135
87 ever, the methodology is time consuming because the devices (SDA) was provided by oxoid, and brain heart infusion 136
88 137
TE
have to be previously trained to avoid false-positive or false- (BHI) broth and buffered peptone water were obtained from
89 negative results. Moreover, the information obtained is often Scharlau Chemie Microbiology and prepared according to 138
90 qualitative. their specifications. 139
EC
94 microcantilevers to detect in situ the growth process of A mini-orbital shaker Stuart was used to prepare the fungal 141
R
95 Aspergillus niger and Saccharomyces cerevisiae. After 4 h biomass. An environmental scanning electron microscope 142
96 of incubation, the cantilever detected the presence of the (E-SEM), Quanta 600 (FEI, Hillsboro, OR, USA), was used 143
O
97 fungi and was able to distinguish between alive and dead to take images of the as-grown networks of SWCNTs and 144
C
98 cells. Despite the emerging methods, some performance the functionalization process. Electrical measurements were 145
99 parameters such as sensitivity and selectivity must be taken using a 4157A Agilent semiconductor parameter 146
N
F
175 diluted in the same solvent for up to 105-fold. After that, were again thoroughly rinsed with distilled water and 226
176 the resulting samples were artificially contaminated by ready to be used to detect A. flavus. After each 227
O
177 spiking 1 mL of A. flavus at 10, 100 or 1,000 ng/mL in functionalization step, the devices were dried with 228
O
178 9 mL of the diluted rice sample. In this way, the final nitrogen and electrically characterized. Figure 1 shows 229
179 concentration of A. flavus in the diluted rice samples was the experimental process used to functionalize the 230
PR
180 1, 10 or 100 ng/mL. The same procedure was used to CNTFET for the detection of A. flavus. 231
181 contaminate the diluted rice samples with 1, 10 or 100 ng/mL
182 of P. chrysogenum and F. oxysporum [24]. Optimization of the PBSTG and the dilution factor of rice
D 232
183 233
TE
Development and functionalization of the CNTFETs Several CNTFET devices were incubated with protein G.
Each device was then incubated with a different 234
184 The SWCNT networks were synthesized on a 500-nm layer concentration of Tween 20 and gelatine for 3 h. After 235
EC
185 of silicon dioxide thermally grown on highly doped n-type this, the devices were immersed in an 8 μg/mL solution 236
186 silicon chips (total area 0.5 cm×0.5 cm) using chemical of anti-Aspergillus antibodies for 30 min at 37 °C. Finally, 237
187 vapour deposition (CVD). Synthesized SWNTs have been the devices were exposed for 1 h to the diluted rice 238
R
188 characterized using both atomic force microscopy and samples: first to the 105-fold diluted rice sample and 239
R
189 electrical techniques. This process has been described finally to the 10-fold diluted rice sample. 240
190 elsewhere [27, 28]. The analysis has shown that a dense
O
191 network of SWCNTs with an average height of 1.5 nm is Determination of the response time 241
C
194 ducting behaviours displayed by the synthesized SWCNTs 2% gelatine) was immersed at 37 °C for 1 h in a 103-fold 243
U
195 are 30% and 70%, respectively, with RON/OFF ~3. To obtain diluted rice solution contaminated with 1 ng/mL of A. 244
196 the field effect transistor configuration, source and drain flavus. Every 15 min, the CNTFET was thoroughly rinsed 245
197 electrodes were screen-printed with silver ink over the with water, dried with nitrogen, electrically characterized 246
198 synthesized SWCNTs. The gap between both electrodes and submerged again in the solution containing 1 ng/mL of 247
199 was 0.5 mm, and the size of the electrodes was 200 μm× A. flavus for another 15 min. 248
200 200 μm. The gate electrode was an aluminium layer on
201 the back side of Si. The CNTFETs were electrically Detection of A. flavus 249
202 characterized by recording the current vs. the gate
203 voltage. To obtain the instrumental variability, we took Another functionalized CNTFET (protected with 1.5% 250
204 all the electrical characterization measurements three Tween and 2% gelatine) was exposed to 103-fold diluted 251
205 times and plotted the mean value and the range value rice samples spiked with increasing concentrations of A. 252
206 of the measurements. flavus (i.e. 1, 10 and 100 ng/mL). For each concentration, 253
207 The functionalization protocol was similar to that the CNTFETs were immersed for 30 min at 37 °C, rinsed 254
208 applied in a previous work where we characterized the thoroughly with distilled water, dried with nitrogen and 255
209 biofunctionalized layer deposited on the SWCNTs [29]. electrically characterized. The presence of the moulds in 256
210 Briefly, CNTFETs were incubated for 30 min at 37 °C in a the recognition layer of the biosensor was confirmed 257
211 5 μg/mL solution of protein G. This is a small globular cell microscopically with E-SEM. 258
AUTHOR'S PROOF!
JrnlID 216_ArtID 3975_Proof# 1 - 06/07/2010
Q2
Fig. 1 Functionalization process to detect A. flavus with CNTFETs. Protein G enables the antibodies to be appropriately orientated while PBSTG
avoids the NSB of biomolecules on the transducer. A. flavus is detected by means of the antigen–antibody interaction
F
259 Selectivity of the CNTFETs the antibodies are able to provide two antigen binding sites 290
O
for A. flavus. Most importantly, preventing the NSB of the 291
O
260 Selectivity was checked in the presence of two mycotoxi- rice compounds onto the CNTs ensures that the electrical 292
261 genic mould P. chrysogenum and F. oxysporum. A characteristics of the devices will only respond to the 293
PR
262 functionalized CNTFET (protected with 1.5% Tween and presence of A. flavus. Therefore, we ensure that the devices 294
263 2% gelatine) was exposed to rice samples contaminated will be selective to A. flavus as long as the antibody does 295
264 with 100 ng/mL of P. chrysogenum, for 30 min at 37 °C, not have cross reactions with other fungi. 296
265 thoroughly rinsed with distilled water, dried with nitrogen
D Sánchez-Acevedo et al. [30] prevented the NSB of small 297
266 298
TE
and electrically characterized. It was subsequently exposed molecules by protecting the CNTs with a Tween 20 at
267 to 100 ng/mL of A. flavus under the same conditions 0.05% and gelatine at 0.8% diluted in phosphate buffer 299
268 mentioned above. The same procedure was followed for F. solution (PBSTG). Figure 2 shows the results obtained for a 300
EC
269 oxysporum. The presence of the A. flavus and the CNTFET device protected with this solution and after being 301
270 interference moulds was also confirmed microscopically exposed to a solution of anti-Aspergillus antibodies. It can 302
271 with E-SEM. be seen that in fact the PBSTG effectively protects the 303
R
273 Electrical characterization of the functionalized CNTFETs shows that increasing the percentage of gelatine to 2% and 308
Tween to 1.5% enables A. flavus to be detected in 103-fold 309
N
274 The electrical behaviour of the CNTFETs was monitored diluted rice samples. The change in the electrical current 310
U
275 after each functionalization step in dry conditions and room when the devices are exposed to higher dilution is 311
276 temperature. We measured three times the dependence of
277 the source-drain current, I, on the back gate voltage, Vg, in
278 the range +5 V to −5 V. The bias voltage, Vsd, was fixed at
279 250 mV. Each electrical current plotted corresponds to the
280 mean value of three replicates. The change in the electrical
281 current as a consequence of the attached proteins has been
282 already studied [29].
F
functionalized CNTFET coated with different concentrations of for 30 min at 37 °C, rinsed thoroughly with distilled 335
PBSTG. CNTFET1: 0.5% Tween 20 and 0.8% gelatine; CNTFET 2:
water, dried with nitrogen and electrically characterized 336
O
1% Tween 20 and 1.5% gelatine; CNTFET2: 1.5% Tween 20 and 2%
gelatine. Each device was exposed to different dilutions of milled rice by measuring the I–V characteristics three times and 337
O
(from 10,000-fold to 100-fold). Each electrical current plotted plotting the mean value and range of the measurement 338
corresponds to the mean value of three replicates. The error bars values. 339
PR
correspond to the range of the electrical current measured for the three
Figure 5 shows the I–V characteristics (for a Vsd=0.25 V) 340
replicates
of a functionalized CNTFET before and after exposure to 1, 341
10 and 100 ng/mL of the mould. It can be seen that the
D 342
312 attributed to the nature of the sample (i.e. the presence of adsorption of A. flavus decreases the conductance of the 343
313 polysaccharides such as starch). As a result, a 103-fold 344
TE
devices (i.e. from a 9% decrease for 1 ng/mL to a 30%
314 dilution of milled rice was chosen as the medium to be decrease for 100 ng/mL). This may be because the antigen– 345
315 artificially contaminated with A. flavus, assuring that the antibody interaction causes a deformation on the SWCNTs 346
EC
316 matrix sample would not affect the electrical behaviour of and thus a scattering effect. The anti-Aspergillus antibody is 347
317 devices protected with 2% of gelatine and 1.5% of Tween able to recognize the antigen galactomannan (GM). It is the 348
318 20. major cell wall component present in Aspergillus species and 349
R
320 The time response was then optimized by immersing SWCNTs. Figure 6 shows an image of a functionalized 353
C
Q2
F
species [33]. Therefore, because these two fungi present 378
similar antigenic epitopes, cross-reactivity would occur with 379
O
the antibody. 380
O
A functionalized CNTFET was first immersed in a rice 381
Fig. 6 An E-SEM image of a functionalized CNTFET after exposure solution containing 100 ng/mL of P. chrysogenum for 382
PR
to A. flavus. The mould is linked to the SWCNT network through 30 min at 37 °C, thoroughly rinsed with distilled water, 383
antigen–antibody interactions
dried with nitrogen and electrically characterized. The 384
device was then exposed to 100 ng/mL of A. flavus under
D 385
354 device after being exposed to A. flavus. We observed that the same conditions as mentioned above. We follow the 386
355 387
TE
regardless of the mycelia filtration process, some spores same procedure with another functionalized CNTFET for
356 remained attached. A. flavus has hyphae that varies in length 100 ng/mL of F. oxysporum. Figure 7 shows that the 388
357 and that is rough with conidia globose to subglobose varying electrical current changes slightly after the CNTFET was 389
EC
358 from 3.5 to 4.5 μm in diameter. GM antigen is produced at exposed to F. oxysporum whereas it decreases about 16% 390
359 an early stage in the metabolic activation of resting conidia, after exposure to A. flavus. The slight changes of the 391
360 but it is increasingly expressed during the swollen conidium electrical current after exposing the devices to F. oxysporum 392
R
361 and hypha stages. Since moulds are primarily found as are due to the variability of the electrical current. By 393
R
362 asexual spores or dried mycelia on food, the anti-Aspergillus contrast, when the devices are exposed to P. chrysogenum, 394
363 antibody used in our assays could recognize the GM in both the electrical current decreases about 3.5% whereas it 395
O
364 conidia and hypha structures [25]. decreases by about 20% after exposure to A. flavus. 396
C
N
U
Q2/Q4
Fig. 7 a Gate voltage dependence of the source-drain current of a of P. chrysogenum (solid line with x marks); and 100 ng/mL of A.
functionalized CNTFET before exposure to the mould (solid line) and flavus (dashed line). Each electrical current plotted corresponds to the
Q3
after exposure to 100 ng/mL of F. oxysporum (solid line with x marks); mean value of three replicates. The inset shows the behaviour of
and 100 ng/mL of A. flavus (dashed line). b Gate voltage dependence the source-drain current at Vg=−5 V. The error bars correspond to
of the source-drain current of a functionalized CNTFET before the electrical current obtained for the three replicates
exposure to the mould (solid line) and after exposure to 100 ng/mL
AUTHOR'S PROOF! JrnlID 216_ArtID 3975_Proof# 1 - 06/07/2010
F
411 and dried with nitrogen. For each concentration, we scanned 461
412 the area between the electrodes (500 μm2) with E-SEM. We in milled rice. This is probably because the use of protein G 462
O
413 did not observe either conidia or mycelia when the device permits the IgG antibodies to be properly oriented. The 463
O
414 was exposed to the different mould concentrations. The same transduction power of the carbon nanotubes makes a special 464
415 procedure was followed for P. chrysogenum. In this case, we labelling process unnecessary; as a result, no additional 465
PR
416 did not observe any conidia or mycelia when the device had reagent is required, thereby reducing costs. The sensor 466
417 been exposed to 1 and 10 ng/mL, whereas three conidia were devices only measure the change caused by the antigen– 467
418 observed when it was exposed to 100 ng/mL P. chrysogenum antibody interaction; therefore, the system displays high
D 468
419 (thus showing a slight cross reaction between the antibody selectivity and, since it is well protected, it is not affected 469
420 470
TE
and P. chrysogenum). With F. oxysporum, the sensor by possible interferences from the raw food. Using suitable
421 displayed high selectivity; however, due to the antigenic molecular receptors, this type of device would be highly 471
422 similarity between P. chrysogenum and A. flavus, there was a sensitive and could detect any mycotoxigenic mould in grains 472
EC
423 slight cross-reactivity of the commercially available antibody. and foods in a short time. 473
424 Therefore, a monoclonal antibody against A. flavus would be 474
425 able to overcome this problem. Acknowledgments We thank the Spanish Ministry of Education and 475
R
Science, MEC, for supporting this study with the project grant 476
426 Our CNTFETs are then quite fast. Although the
CTQ2007-67570. RAV acknowledges the University of Rovira i 477
R
427 construction of the sensor takes about 4 h, the measuring Virgili University for providing economic support. RAV would also 478
428 time required to detect the mycotoxigenic mould takes only 479
O
430 interpret the results [4]. The devices can be built in batch,
431 obtaining hundreds of them simultaneously. Moreover,
N
432 CNTFETs are selective and the lowest amount of mould References 481
U
502 11. de Lacy Costello BPJ, Ewen RJ, Gunson H, Ratcliffe NM, 22. Piao MH, Noh MH, Rahman MA, Won MS, Shim YB (2008) 525
503 Sivanand PS, Spencer PTN (2003) Meas Sci Tech 14:397–409 Electroanalysis 20:30–37 526
504 12. Falasconi M, Gobbi E, Pardo M, Della Torre M, Bresciani A, 23. Hao W, Pan YX, Ding YQ, Xiao S, Yin K, Wang YD, Qiu LW, 527
505 Sberveglieri G (2005) Sens Actuator B Chem 108:250–257 Zhang QL, Woo PCY, Lau SKP, Yuen KY, Che XY (2008) Clin 528
506 13. Muramatsu H, Kajiwara K, Tamiya E, Karube I (1986) Anal Chim Vaccine Immunol 15:194–202 529
507 Acta 188:257–261 24. Tsai GJ, Yu SC (1999) Int J Food Microbiol 50:181–189 530
508 14. Nugaeva N, Gfeller KY, Backmann N, Lang HP, Duggelin M, 25. Hetherington SV, Henwick S, Parham DM, Patrick CC (1994) 531
509 Hegner M (2005) Biosens Bioelectron 21:849–856 Clin Diagn Lab Immunol 1:63–67 532
510 15. Chen H, Heng CK, Puiu PD, Zhou XD, Lee AC, Lim TM, Tan 26. De Vos MM, Nelis HJ (2003) J Microbiol Meth 55:557–564 533
511 SN (2005) Anal Chim Acta 554:52–59 27. Villamizar RA, Maroto A, Rius FX, Inza I, Figueras MJ (2008) 534
512 16. Gui EL, Li LJ, Zhang K, Xu Y, Dong X, Ho X, Lee PS, Kasim J, Biosens Bioelectron 24:279–283 535
513 Shen ZX, Rogers JA, Mhaisalkar SG (2007) JACS 129:14427– 28. Villamizar RA, Maroto A, Rius FX (2009) Sens Actuator B Chem 536
514 14432 136:451–457 537
515 17. Star A, Tu E, Niemann J, Gabriel JCP, Joiner CS, Valcke C (2006) 29. Villamizar RA, Braun J, Gompf B, Dressel M, Rius FX (2009) 538
516 PNAS 103:921–926 Biosens Bioelectron 25:161–166 539
517 18. So HM, Won K, Kim YH, Kim BK, Ryu BH, Na PS, Kim H, Lee 30. Sánchez-Acevedo ZC, Riu J, Rius FX (2009) Biosens Bioelectron 540
518 JO (2005) JACS 127:11906–11907 24:2842–2846 541
519 19. Maehashi K, Katsura T, Kerman K, Takamura Y, Matsumoto K, 31. Bernard M, Latgé JP (2001) Med Mycol 39:9–17 542
F
520 Tamiya E (2007) Anal Chem 79:782–787 32. Schoffelmeer EAM, Klis FM, Sietsma JH, Cornelissen BJC 543
521 544
O
20. Dastagir T, Forzani ES, Zhang R, Amlani I, Nagahara LA, Tsui R, (1999) Fungal Genet Biol 27:275–282
522 Tao N (2007) Analyst 132:738–740 33. Latgé JP, Kobayashi H, Debeaupuis JP, Diaquin M, Sarfati J, 545
523 546
O
21. So HM, Park DW, Jeon EK, Kim YH, Kim SK, Lee CK, Choi SY, Wieruszeski JM, Parra E, Bouchara JP, Fournet B (1994) Infect
524 Kim SC, Chang H, Lee JO (2008) Small 4:197–201 Immun 62:5424–5433 547
PR
548
D
TE
EC
R
R
O
C
N
U
AUTHOR'S PROOF!
AUTHOR QUERIES
F
O
O
PR
D
TE
EC
R
R
O
C
N
U