Biomimetic optical sensors for environmental endocrine disruptor screening

UTU has successfully built up an Integrated Optical Mach-Zehnder setup, which has been tested with well-investigated antibody reactions. It is possible to produce a multi-analyte interferometric sensor and to evaluate those results with the help of Artificial Neural Networks. Due to the lack of MIPs delivering specific binding signals in water, SPR (Surface Plasmon Resonance) and RIfS (Reflectometric Interference Spectroscopy) were used as alternative optical method to test the MIPs supplied by TUM and LUND in organic solvents. Specific, time-resolved and reproducible binding signals were detected with the MIPs imprinted against atrazine. Within the data evaluation part a chemometric model could be developed which is in principle able to predict endocrine disrupting potential from multi-analyte sensor signals, but could not be tested with MIPs, again because of the lack of MIPs which could be measured in water with the Mach-Zehnder.

A new concept of a high-throughput surface plasmon resonance (SPR) sensor based on the angular spectroscopy of surface plasmons on a two-dimensional array of diffraction gratings was developed and a laboratory prototype (breadboard level) of the sensor was realized. In this sensor, SPR angular spectra from a single row of diffraction gratings are acquired using a two-dimensional detector and the reading of SPR spectra from rows of diffraction gratings is performed sequentially by chip scanning. The special chip with an array of diffraction gratings is integrated with a sensor chip cartridge interfaced with a multi-channel flow-cell. In model refractometric experiments, the sensor sensitivity to refractive index changes was found to be about 160 deg/RIU (refractive index unit). The minimum detectable change in the refractive index was determined to be about 7x10E-7 RIU. This sensor presents a generic platform offering a very high number of independent sensing channels (up to several hundreds). In conjunction with appropriate surface chemistries and biorecognition elements it could be applied to parallelized observation of molecular interactions or detection of large number of analytes or screening of a large number of samples for a single analyte.

The T-47D cell line, a breast carcinoma cell line which expresses receptors to 17- â estradiol, other steroids and calcitonin has been selected for exposure protocols to assess gene expression. The cell line was exposed to a selection of endocrine disrupting compounds (EDCs), compared to negative controls and corresponding RNA-extracts were used to validate the DNA-chip that has been developed by other partners (NIO, ARCS) within the Mendos consortium. Interesting genes have been confirmed by real-time PCR on the available samples. This approach was set up in comparison to another test system with the ZR 75-1 cell line (performed at NIO). By comparing all data, we aimed to identify the most ideal cell line that can be used for validation of other sensors (MIPs) as well as to identify the subset of genes that allow for the best identification of EDCs. Protocols (cell culture, RNA extractions & hybridisation) are available and can be used for future EDC exposure experiments

The first stage of developing the immobilized yeast cells biosensor was to assess and optimized the condition that effect of the luminescence intensity that emitted by the immobilized yeast cells. The aspects of cell density in the hydrogel beads, the time that require for the cells induction and luminescence kinetics were examined. The evaluation was carrying out both for alginate and PVA hydrogel beads. It was found that the best signal to noise ratio was obtained at a volume ratio of 1:2 alginate 3% (w/v): 5X106 [cells/ml]. The luminescence kinetics reflects a decrease of the luminescencentic signal during the incubation time. The low ß-E2 induction factors at higher incubation times could be explained by the increasing basal luminescence background effect of the non-induced cells. As a result, 2.5 h of ß-E2 bead induction was set as the optimal induction time for further applications. The characterization of the Yeast cell immobilization in PVA based hydrogels shows different pattern of induction by ß-E2 in comparing to alginate immobilize beads. The optimum induction time was longer (between 2.5-5.5 h), the luminescence intensity response was higher and also the sensitivity as reflected in the experiments of ß-E2 dose-response calibration curves. The higher luminescence intensity reported by PVA hydrogels than that of alginate beads, could possibly be linked to the difference in viability of immobilized cells and gel porosity. Despite of the higher sensitivity and stability that shows the PVA hydrogel beads, since it is impossible to count the yeast colonies that trapped in the PVA, we chose to work with the alginate hydrogel yeast beads. It has been previously reported that the addition of CoA to the luciferin solution increases the luminescence intensity and lowers the rapid inhibition of light production (25) in invitro conditions. We did not observe in changes of the luminescence intensity as results of the presence of CoA in our experimental system. These results could possibly be explained by the non-specific effect of CoA on the non-ß-E2 induced control bead samples, which increased the basal luminescence background. It is also possible that CoA affects the metabolism of the yeast cells and thus causes changes in the luminescence via indirect ways, which need to be investigated. The option to store the hydrogel yeast biosensor in low temperature for long periods is a remarkable advantage. it is allow us to perform the experiment at shorter time without growing the continually yeast culture, also it provided uniformly of the yeast beads biosensors. The experiments of response to estrogenic EDCs (synthetic, endogenous and the phytoestrogens estrogenic chemicals), which were assessed, observed high sensitive response of the alginate immobilized yeast cells. This result leads us to monitoring of environmental samples. The environmental samples demonstrate the ability of the immobilized yeast cells biosensor to serve as a tool for detection the presence of EDCs contamination in liquid samples - water or extracts of different sources. The luminescent yeast based hydrogel bioassay should be applicable to the analysis of environmental water samples, as a primary screening tool, and as the biological and chemical immobilisation parts in the construction of fibre-optic biosensors. In conclusion, we have demonstrated that it is possible to assess the estrogenic activity of a sample by measuring the reported luminescence with calcium alginate and PVA immobilized yeast cells. The advantages of the developed bioassay are: - Rapidity: the bioassay was characterized by a total period time of 2.5 h. - Shelf life: long term storage of the yeast cells. - Biocompatibility: hydrogels form a protective environment for the entrapped yeast cells protecting from contamination, thus allows the user to work under non-sterile conditions. - Simplicity: the bioassay is simple to perform, the hydrogels are stable and easy to handle. - Cost: inexpensive method in comparing to chemical analysis methods like LC-MS-MS which requires an expensive instrument and well-trained personnel. In addition, we have shown that luminescent yeast based hydrogel bioassay should be applicable to the analysis of environmental water samples, as a primary screening tool, and as the biological and chemical immobilisation parts in the construction of fibre-optic biosensors. The ability of long run storage, easy and cost effective preparation, no need of continuous cell culture cultivation, and only 2.5h of assay time make the immobilized estrogen inducible yeast based hydrogel a promising alternative to the current used recombinant yeast reporter gene assay of cell suspension.

Nano Fountain Pen (NFP) uses hollow glass or quartz capillaries heat-drawn to a tapered tip with an aperture of a few hundred nanometers26. When pre-polymerization solution is loaded in such nano-pipettes, it is drawn to the end of the tapered tip by capillary forces, however the solution will flow out only upon contact with the surface. The nano-pipette is mounted as the probe of an AFM and can be controllably directed to very precise locations on a surface, using the regular control of an AFM probe. When loaded with MIPs pre-polymerization mixture, the NFP printing results in drops as small as ~ 100 nm in diameter, which are subsequently polymerized by UV radiation. This yields arrays of nanometre-sized MIPs.

Fabrication, protocol and hybridization conditions of hydrophobic chip surfaces made of epoxy resin (so-called ARChip Epoxy) and PST-co-VBT (so-called ARChip UV) were optimized for use in an oligonucleotide chip for EDC detection. The lowest oligonucleotide concentration leading to a significant hybridization signal was 2 µM. Fluorescence signals resulting from probes of concentrations higher than 20 µM did not further increase signal intensity. Both, SH- and NH2-modified oligonucleotides led to a significant hybridization signal, whereas almost no signal occurred, when unmodified oligonucleotides were employed. The hybridization fluorescence of SH-and NH2-modified Alf1b50 strongly depends on the pH of the print buffer. Whereas both modifications are readily immobilized at pH 7 generating the same strong signal, NH2-modified oligonucleotides show an improved, optimum binding at pH 8 resulting in a 20-fold increased signal. When adding a 33dT spacer to a 17-bases oligonucleotide, the hybridization efficiency was increased by 50%. The signal-to-noise ratio (S/N) increased by 2 to 3 times, when exposed to UV-light after spotting. Efficient blocking of reactive surface groups after arraying is critical for a reduced fluorescence background. The epoxy chip blocked with ethanolamine tend to provide a more homogeneous background, which is clearly expressed by the coefficient of variance at 10, 1 and 0.1 µM guide dot concentration (with ethanolamine (ETA)): 6.3%, 7.8%, 34.3%; without ETA: 30.5%, 32.8%, 45.0%). The effect of high guide dot concentration on neighbouring spots is most obvious with 10 µM Cy5-labelled oligonucleotide: compared with 1 and 0.1 µM Cy5-fluorescent oligonucleotide the background is double. The increased local background is a result of the background contamination of hybridized probes around the bright guide dots. The choice of print buffer is critical for spot homogeneity and signal strength and thus has direct impact on data analysis and data reliability. 20 different buffers and buffer additives were tested and optimized in oligonucleotide arrays. A series of ARChips Epoxy was produced and distributed among the partners for EDC testing. ARChip UV was optimized with respect to chip fabrication, photoactivation (pre- and post-activation, no activation) and print buffer & additives. The immobilization capacity was effected by the UV energy density used for photoactivation of the vinylbenzylthiocyanate group and was estimated from the signal difference of fluorescent, amino-modified Alf1b after spotting and after hybridization. The decrease in fluorescence signal was taken as a measure for the unbound Alf1b washed off during the washing and blocking steps and was inversely proportional to the immobilization capacity. Increasing the energy density of the 254 nm UV light from 60 to 450 mJ cm-2 resulted in an increase in immobilization capacity from 20 to 80%. Thus, an energy density of 450 mJ cm-2 was chosen for the activation step in all experiments.

Molecularly imprinted polymer films by spin-coating. This item constitutes a robust technique for preparing molecularly imprinted polymer (MIP) films, with good control of thickness and porosity. The use of polymer porogens facilitated access to the synthetic receptor sites, and the relationship between the morphology and binding properties was rigorously investigated using AFM and autoradiography. Spin-coating is one of the standard coating techniques used for example in the microelectronics industry. We have recently demonstrated that it is possible to spin-coat the monomer mixture on a flat surface followed by in-situ photopolymerization in the presence of a template, in order to obtain MIP films with thicknesses between 100 nm and several µm. However, with acrylic and vinyl monomers, polymerization of these films is too fast for phase separation to occur, so that non-porous films with very little binding capacity is obtained. However, it is possible to accelerate phase separation by adding a linear polymer such as poly(vinyl acetate) as co-porogen. The porosity of the films can be fine-tuned via the amount and the molecular weight of the poly(vinyl acetate), and even surfaces covered with nanoparticles can be obtained (Figure 5). We found that in contrast to bulk polymers, in these systems phase separation and pore formation is by spinodal decomposition. These MIP films can be used as selective recognition layers in biosensors.

Phthalate MIPs were prepared using only one kind of functional monomer (methacrylic acid - MAA), but different templates (bis(2-ethylhexyl)phthalate - DEHP, benzylbutylphthalate - BBP, and dibutylphthalate - DBP), cross-linkers (ethylene glycol dimethacrylate - EDMA and TRIM), porogens (cyclohexane - CH and acetonitrile - ACN), and polymerization methods (bulk polymerization - BP, precipitation polymerization - PP, miniemulsion polymerization - MEP). While with BBP and DBP only a very weak imprinting effect could be observed with corresponding MIPs (Imprinting Factor ~2.5), the use of DEHP as template molecule in combination with MAA and EDMA led to a polymer, which exhibited most significant selectivity (IF 8.3). The lacking binding of the template (DEHP) in the rather polar porogen (ACN) indicates that the interaction of DEHP with the MIP is not based on hydrogen bonding but moreover on hydrophobic interaction between template and polymer. Using HPLC and ACN-water (60:40, v/v) as mobile phase, almost baseline separation of the three different phthalate esters was obtained. This was not possible with MIPs, which were prepared accordingly but using DBP and BBP as templates and, therefore, is a clear indication that this different binding is not caused by a non-specific interaction of these analytes to the polymer. If the different polymerization methods are compared, significantly higher capacity factors were obtained using PP in comparison to BP. This should be based on the difference in surface area between both polymers. However, the IF were almost the same (8.3 vs. 9.8). Using PP and MEP, nano-sized MIPs were prepared. Significantly smaller particles were obtained by MEP (~250 nm compared to ~500 nm). With both methods, size of NIPs was higher than corresponding MIPs. The effect was most obvious with MEP. Therefore, an influence of the template molecule on the size of the resulting particles seems to be clear. Further, nano-sized MIPs prepared by both methods showed an imprinting effect on pure ACN and ACN-water (60:40, v/v). Compared to the MIPs prepared by PP, the rebinding capacity of MEP-MIPs was significantly higher in both solvents. As revealed from BET experiments, for the latter the highest surface area was calculated (75.83 m2 g-1 compared to 32.71 m2 g-1 for the MEP-NIP), however, this difference is not as big as the observed difference in the amount of bond analyte (DEHP) (factor was about 6). Therefore, it can be concluded that MEP-MIPs contain a considerable number of specific binding sites for DEHP. DEHP-MIPs prepared by BP were used as sorbents for MISPE and selectivity was evaluated using a mixture of 19 EDCs. While a volume of 8 ml of ACN-water (60:40, v/v) removed ~95% of DEHP from the blank column it was almost totally retained on the MIP column. Only a loss of about 5% was observed which could be attributed to non-specific binding of the analyte. However, it could be completely eluted from the MIP with the same volume of pure ACN. Interestingly, beside DEHP also p,p -DDT and Benzo[a]pyrene were almost completely found in the eluate fraction. From the investigated EDCs, these two analytes exhibited by far the highest capacity factors, which should be attributed to their extremely high hydrophobicity. However, the IF were significantly lower compared to DEHP (IF 7.1 and 6.8 for p,p-DDT and BAP vs. 8.3 for DEHP). Based on the findings, a LOD can be calculated as three times the signal-to-noise ratio, i.e. 240 ng of DEHP in 1 l of water.

Thin molecularly imprinted polymer (MIP) films based on self-assembly of nanoparticles MIP nanoparticles were coated on a gold-coated glass substrate for use in SPR and other evanescent-wave-based transducers. The gold surface is first modified with a self-assembled monolayer (SAM) of 11-mercaptoundecanoic acid (MUA). This layer is then overlayed with a thin layer of polyethylene imine (PEI) or polyallylamine forming ion pairs with the MUA. MIP nanoparticles based on methacrylic acid as the functional monomer can be directly adsorbed onto this film in form of a monolayer, via ionic interactions. If the functional monomer contains a basic function, a polysulfonic acid underlayer is required. The coating procedure was followed in an SPR assembly. These MIP films can be used as the selective recognition layers in biosensors.

Hydrogel surfaces based on polyurethane (PU) and poly(vinyl alcohol) (PVA) modified with anionic, cationic and neutral additives (e.g. RL100 (Meth1), RS100 (Meth2), E100 (Meth3), Praestol 187, Praestol 185, Praestol 190, Polystabil KWS, chitosan) and crosslinkers (e.g. polyethylenimine (PEI), polyallylamine (PAH), glutaraldehyde, phenylendiamine (HAD), trianzinyl-ß-cyclodextrin (MCT), cyanurchloride (TsT) have been developed and tested for protein immobilization. Best-of surfaces have been selected with respect to signal-to-noise ratio and data reproducibility (CV <40%). The surfaces have been characterized with regard to biologically relevant parameters (signal-to-noise ratio S/N, % background, immobilization capacity) and by physical methods (AFM done at BGU and ellipsometry at IREE) and compared with ARChip Epoxy and the poly(acrylamide) slide from Perkin Elmer. Data from contact angle measurements and swellability studies, furthermore, topographic and roughness data from AFM were correlated with S/N values to design a high quality protein surface. PU1 was used as platform in an on-chip immunoassay for 17ß-estradiol. The dynamic range was from 0.001 to 0.014 µg/ml 17ß-estradiol.

A laboratory prototype of an eight-channel mobile surface plasmon resonance (SPR) sensor was developed at IREE. The sensor exploits the spectral modulation and the wavelength division multiplexing (WDM). The wavelength division multiplexing is realized in a special multireflection sensing element coated with a thin film of SPR-active metal. The mobile SPR sensor encompassed the following key subsystems: an SPR coupling optics, a miniature broadband light source, a multichannel spectrograph module, a microfluidic flow-cell and supporting electronics. Performed characterization experiments confirmed that the wavelength division multiplexed sensing channels exhibit minimum cross-sensitivity and that this approach is indeed a viable approach for the development of multichannel SPR sensing devices. In model refractometric experiments it was demonstrated that the laboratory prototype of the mobile SPR sensor can detect variations in the refractive index as small as 10E-6 RIU (refractive index unit). This sensor presents a generic sensing platform and can be combined with specific biorecognition elements and applied for detection of numerous chemical or biological analytes.

Six kinds of molecularly imprinted polymers (MIPs) or microspheres (MIMs) were firstly synthesized by bulk polymerization and aqueous micro-suspension polymerization using benzo[a]pyrene (BAP) as imprinting molecule. Three monomers (MAA, TFMAA, 4-VP) and two cross-linkers (EDMA, DVB) were investigated. The MIPs or MIMs obtained were subsequently characterized in detail with rebinding investigation, solid-phase extraction (SPE) and liquid chromatography (HPLC). All the results indicate that the polarity of both monomer and cross-linker play an important role for affinity and retention behaviour. The affinities of the polymers to BAP increased with the decreasing polarity of monomer and cross-linker. The rebinding results demonstrate that the affinity order is 4-VP-co-DVB > MAA-co-DVB > TFMAA-co-DVB > 4-VP-co-EDMA. The HPLC results also indicate that the retention time of BAP on 4-VP-co-DVB and 4-VP-co-EDMA columns is much different (18.02 min versus 10.89 min) and that it needs a more polar mobile phase for the former (ACN-DCM, 80:20, v/v versus ACN-water, 80:20, v/v). The selectivity of the imprinted 4-VP-co-DVB column to BAP and its homologous is much better than that of the 4-VP-co-EDMA column (IF-BAP 8.44 versus 1.34). In this case, it is illustrated that the cross-linker is much more important than the monomer, and that the hydrophobic interaction and Pi-Pi interaction between analytes and the MIP-HPLC column play a key role in the retention process. The 4-VP-co-DVB column was employed to enrich BAP from tap water, lake water and instant coffee samples which were spiked at low ppb levels without any sample pre-treatment apart from filtration. The recoveries were 96.5%, 84.7%, and 72.5%, respectively.

Deliverables

Share this page

Download