Gross R., Sidorenko A., Tagirov L. Nanoscale Devices

Подождите немного. Документ загружается.

Thermodynamic Principles of Artificial Nose 33

Generally, the structure-energy relationships for the guest vapor – solid

host binding may be much more complicated than shown on Figure 8, when

the host has an essentially different packing in inclusion compounds with

different guests and the guest molecular shape is relevant [7, 10]. While a

lot of supramolecular hosts can be synthesized with different packing

modes, each of them may have its own specific structure-energy

relationship, the bright perspective is open for the applications of

supramolecular hosts in sensor arrays for intelligent recognition of odors.

Acknowledgments

This work was supported by the programs of RFBR (No.03-03-96188 and

05-03-33012), URBR (No. UR.05.01.035) and BRHE (REC007).

References

1. Grate JW (2000). Chem Rev 100:2627

2. Setford SJ (2000). Trends Anal Chem 19:330

3. Edsall JT, Gutfreund H (1983) Biothermodynamics. J Wiley, New York

4. Halling PJ (1994). Enzyme Microb Technol 16 :178

5. Furusho Y, Aida T (1997). Chem Commun 2205

6. Dewa T, Endo K, Aoyama Y (1998). J Am Chem Soc 120:8933

7. Gorbatchuk VV, Tsifarkin AG, Antipin IS, Solomonov BN, Konovalov AI,

Seidel J, Baitalov F (2000). J Chem Soc Perkin Trans 2 11:2287

8. Gorbatchuk VV, Antipin IS, Tsifarkin AG, Solomonov BN, Konovalov AI

(1997). Mendeleev Commun p 215

9. Gorbatchuk VV, Tsifarkin AG, Antipin IS, Solomonov BN, Konovalov AI

(1999). J Inclusion Phenom Macrocyclic Chem 35:389

10. Gorbatchuk VV, Tsifarkin AG, Antipin IS, Solomonov BN, Konovalov AI,

Lhotak P, Stibor I (2002). J Phys Chem B 106:5845

11. Gorbatchuk VV, Savelyeva LS, Ziganshin MA, Antipin IS, Sidorov VA

(2004). Russ Chem Bull 53:60

12. Perutz MF, Wilkinson AJ, Paoli M, Dodson GG (1998). Annu Rev Biophys

Biomol Struct 27:1

13. Bettati S, Mozzarelli A (1997). J Biol Chem 272:32050

14. Coetzee A., Nassimbeni LR, Achleitner K (1997). Thermochim Acta 298:81

15. Barbour LJ, Caira MR, Nassimbeni LR (1993). J Chem Soc Perkin Trans

2 :2321

16. Caira MR, Horne A, Nassimbeni LR, Toda F (1997). J Mater Chem 7:2145

17. Dickert FL, Schuster O (1995). Microchim Acta 119:55

18. Dalcanale E, Hartmann J (1995). Sensors and Actuators B 24-25:39

19. Wang C, Chen F, He X-W, Kang S-Z, You C-C, Liu Y (2001). Analyst

126:1716

34 V. V. Gorbatchuk, M. A. Ziganshin

20. Dickert FL, Tortschanoff M, Weber K, Zenkel M, Fresenius (1998). J Anal

Chem 362:21

21. Ziganshin MA, Yakimov AV, Antipin IS, Konovalov AI, Gorbatchuk VV

(2004). Russ Chem Bull 53: (in press)

22. Manakov AYu, Soldatov DV, Ripmeester JA, Lipkowski J (2000). J Phys

Chem B 104:12111

23. Brihi TE, Jaubert J-N, Barth D, Perrin L (2002). J Chem Eng Data 47:1553

24. Nassimbeni LR (1998) Inclusion compounds: kinetics and selectivity.

Molecular Recognition and Inclusion. Dordrecht: Kluwer

25. Atwood JL, Barbour LJ, Jerga A (2002). Chem Comm 2952

26. Brouwer EB, Enright GD, Udachin KA, Lang S, Ooms KJ, Halchuk PA,

Ripmeester JA (2003). Chem Comm 1416

27. Mosbach K, Haupt K (2000). Chem Rev 100:2495

28. Gorbatchuk VV, Mironov NA, Solomonov BN, Habicher WD (2004).

Biomacromolecules 5:1615

29. Halling PJ (1990). Biochim Biophys Acta 1040:225

30. Sirotkin VA, Borisover MD, Solomonov BN (1997). Biophys Chem 69:239

31. Gorbatchuk VV, Ziganshin MA, Solomonov BN (1999). Biophys Chem

81:107

32. Mironov NA, Breus VV, Gorbatchuk VV, Solomonov BN, Haertle T (2003). J

Agric Food Chem 51:2665

33. Gregory RB (1995) In Protein-Solvent Interactions: Protein Hydration and

Glass Transition Behavior. Marcel Dekker Inc, New York

34. Kramer RZ, Vitagliano L, Bella J, Berisio R, Mazzarella L, Brodsky B, Zagari

A, Berman HM (1998). J Mol Biol 280:623

35. Yennawar NH, Yennawar HP, Farber GK (1994). Biochemistry 33:7326

36. Ladbury JE (1996). Chemistry & Biology 3:973

(2004). Macromol Symp 210:263

38. Klibanov AM (2001). Nature 409:241

39. Grate JW, Patrash SJ, Abraham MH, Du CM (1996). Anal Chem 68:913

40. Grate JW, Wise BM, Abraham MH (1999). Anal Chem 71:4544

37. Gorbatchuk VV, Ziganshin MA, Savelyeva LS, Mironov NA, Habicher WD

Molecular Detection with Magnetic Labels and

Magnetoresistive Sensors

J. Schotter

1,2

, M. Panhorst

, M. Brzeska

, P. B. Kamp

, A. Becker

A. Pühler

, G. Reiss

, H. Brueckl

1,2

ARC Seibersdorf research GmbH, Nano-Systemtechnologien, 1220 Wien,

Austria

Department of Physics, University of Bielefeld, 33615 Bielefeld, Germany

Department of Genetics, University of Bielefeld, 33615 Bielefeld,

Germany

Abstract: For future lab-on-a-chip devices, compact and inexpensive detection

units are required that directly translate the abundance of certain

biomolecules into an electronic signal. By detecting specifically

bound magnetic labels with magnetoresistive sensors, a versatile

platform can be designed that fulfils those requirements and even

enables on-chip manipulation of biomolecules by suitable magnetic

gradient fields. Here, we present sensitive recognition of different

types of magnetic labels by magnetoresistive sensors based both on

giant magnetoresistance (GMR) and tunneling magnetoresistance

(TMR). Hybridization experiments show that our prototype

magnetoresistive biosensor can detect complex DNA with a length of

one thousand base pairs down to a concentration of 24 pM. A direct

comparison of our magnetoresistive and a standard fluorescent

detection method clearly shows the advantage and competitiveness of

our approach.

Keywords: biosensor, DNA, lab-on-a-chip, microbead, magnetoresistance, GMR,

TMR

R. Gross et al. (eds.), Nanoscale Devices - Fundamentals and Applications, 35–46.

36 J. Schotter, M. Panhorst, M. Brzeska, et al.

Introduction

Selective and quantitative detection of small amounts of biomolecules plays

an important role in the biosciences, in clinical diagnostics or medical

research. So far, it is standard procedure to collect the samples on site and

send them to specialized laboratories for analysis, which is rather cost

intensive and time consuming. Faster and also more reproducible results

could be obtained by so-called ‘lab-on-a-chip’ devices, which have received

great attention recently [1, 2]. Ideally, these chip-based systems integrate all

the necessary steps to detect certain biomolecules from an originally

unprocessed specimen (e.g. separation, amplification, chemical

modification and detection) into a single easy-to-use portable device.

Recent progress in the field of microfluidics [3] suggests that the

preparation of biological samples within an integrated microfluidic device

is evolving and will become commercially available within the next few

years.

Concerning the molecular detection unit of future lab-on-a-chip devices,

different techniques are currently employed or actively researched. With

respect to DNA analysis, they all rely on the principle of detection by

hybridization, which allows a highly parallel analysis of thousands of

sequences at a time, each of them within a separate specifically

functionalized spot of the sensor. As the sample solution is spread across

the entire sensor area, target DNA that is complementary to the

immobilized probes hybridizes. As the sequence and position of the

preassembled probe DNA is known, the type and concentration of the

unknown sample is mapped by measuring the abundance of hybridized

target DNA at each spot location. The majority of the established detection

methods accomplish this task by specifically adding labels to the hybridized

target DNA only and collecting the signals from those labels. Concerning

today’s semi-integrated micro-arrays, optical methods that employ

fluorescent markers are most common [4]. However, due to spatial

restrictions for lab-on-a-chip devices, it seems unlikely that they are going

to be used extensively for this new market segment. More fit for that

purpose are electrochemical detection methods via suitable electroactive

labels, which can easily be integrated into chip-based formats and show

sensitivity values comparable to optical techniques [5].

Another concept for DNA detection was introduced in 1998 by Baselt et

al. [6, 7]. It is based on detecting the stray field of magnetic labels by

embedded magnetoresistive sensors. Such a technique, apart from being

sensitive,

flexible and compatible with standard CMOS fabrication, also

has the advantage of directly providing an electronic signal that is suitable

Molecular Detection with Magnetic Labels and Magnetoresistive Sensors 37

for automated on-chip analysis. Furthermore, the possibility to attract the

magnetic labels by suitable gradient fields also opens up a completely new

prospect that is not possible in such a way for any of the competing

detection schemes, i.e. the selective on-chip manipulation of desired

biomolecules [8]. This option could be used, for example, to reduce

incubation times during hybridization, or to test binding forces of adhered

molecules [9]. Due to these opportunities, magnetoresistive biosensors

represent a very promising alternative detection unit for lab-on-a-chip

devices, and since the first publication in 1998, a number of research groups

initiated work on that subject [10]. Here, we present some of our key results

on magnetoresistive biosensors [11].

Figure 1 displays the different steps involved in DNA detection by our

magnetoresistive biosensor. First, samples of probe DNA are spotted onto

the sensor surface and are immobilized via epoxy groups embedded into the

top polymer layer (part a). Second, the biotin-labeled analyte DNA is added

and hybridizes to complementary probe DNA (part b). In the final step,

streptavidin-coated magnetic markers are introduced and bind specifically

to the biotin of the hybridized analyte DNA (part c). After each step,

washing removes unbound DNA or markers. The magnetic stray field of

the markers is detected as a resistance change in a magnetoresistive (MR)

sensor embedded underneath the probe DNA spot.

Fig. 1. Sketch of the DNA detection process by the magnetoresistive biosensor.

A magnetoresistive sensor based on the giant magnetoresistance (GMR)

effect [12] is developed and optimized for spotter-based microarray

applications. It consists of multilayers in the second antiferromagnetic

coupling

maximum of the following composition: (Si/(Ni

)

1.6nm-

/[Cu

1.9nm

/(Ni

)

1.6nm

]

/Ta

3nm

GMR Type Biosensor

38 J. Schotter, M. Panhorst, M. Brzeska, et al.

The film stack is sputter-deposited and subsequently patterned by

electron beam lithography and Argon ion etching. A thin Ta film on top of

the layer stack serves as an adhesion promoter for the final SiO

protection

layer.

The inset of Fig. 2 shows an electron micrograph of a typical sensor

element. The spiral-shaped line has a width and separation of 1 µm, and it

covers a circular area with a diameter of 70 µm. An entire probe DNA spot

can thus be covered by a single sensor element. The total resistance of a

sensor element depends on layer thickness and multilayer number and

amounts to about 12 kΩ in this case. The response characteristic to an in-

plane magnetic field is also displayed in Fig. 2. Due to the isotropic

geometry, the response is the same for every angle of the in-plane field. The

relative resistance change is about 7% at a saturation field of roughly

15 kA/m, resulting in a sensitivity to in-plane magnetic fields of 0.5% per

kA/m. The prototype sensor consists of 206 individual spiral-shaped

elements on a total area of 5×12 mm

. The contact pads and interconnect

lines are made of a Ta

10nm

50nm

10nm

sandwich and are patterned using

positive photo lithography and lift-off.

Fig. 2. Layout and magnetoresistance response of a GMR-type sensor element.

The entire sensor with the exception of the contact pads is passivated by

a 160 nm thick sputter-deposited SiO

-layer. Additionally, a 60 nm thick

methacryl-based polymer layer is spin-coated onto the surface from a

dioxane solution. It contains epoxy side-groups and allows covalent

Molecular Detection with Magnetic Labels and Magnetoresistive Sensors 39

bonding of the probe DNA amino groups to the surface. Furthermore, the

polymer acts as an additional protection layer for the underlying sensor.

To test the dependence of the sensor signals on the marker coverage,

different marker types in varying concentrations are directly spotted on top

of separate elements of the same sensor. In a differential setup, the sensor

element of interest is measured relative to a reference element, which is not

covered by any markers. In order to magnetize the superparamagnetic

labels, a magnetic field is applied perpendicular to the plane, which

minimizes its direct influence onto the sensor while assuring high magnetic

moments of the markers. The in-plane components of the labels’ stray

fields are radially symmetric around their center positions and cause local

distortions of the magnetization configuration of the underlying

magnetoresistive sensor.

Figure 3 displays typical data for such a measurement. If the measured

element is also free of any markers, almost no change of resistance occurs

in any of the two measurement branches while increasing the magnetizing

field, and the output signal of the differential amplifier stays constant (gray

line). The black line shows the data for a measured element in which 5% of

the surface area is covered by Bangs 0.86 µm magnetic markers. With

increasing magnetizing field, the induced dipole moment of the

microspheres becomes stronger, and the resistance in the upper branch

drops. The output signal is symmetric and displays a nearly linear

dependence on the magnetizing field, which is expected as a direct

consequence of the linear rise of the marker moment at small fields and the

linearity of the GMR characteristics.

Fig. 3. Sensor signals with (black line) and without (gray line) magnetic markers in

dependence on the perpendicular magnetizing field.

40 J. Schotter, M. Panhorst, M. Brzeska, et al.

Fig. 4. Dependence of the sensor output signals on the surface coverage of three

different marker types. The lines are linear regressions to the data.

The response of three different types of magnetic markers is investigated

at different surface coverages (Fig. 4). For each single measurement, the

maximum difference in the output signal is taken and plotted versus the

marker coverage of the specific sensor element. The shaded area represents

the maximum signal obtained from reference elements with zero marker

coverage, which is always less than 40 mV. At coverages not too close to

saturation, a linear increase of the sensor signal on the marker coverage is

observed, as more and more of the sensor area gets affected by the induced

stray fields. Deviations can be attributed to conglomerations of markers in

specific regions.

In summary, the GMR-based magnetoresistive sensors are able and

appropriate to directly determine the density of magnetic markers on the

sensor surface. Therefore, just by measuring an electrical resistance, it is

possible to conclude the abundance of hybridized DNA strands on the

surface of each sensor element, which is demonstrated in the next

paragraph.

In order to test the biological sensitivity of the magnetoresistive biosensor,

a comparison experiment with standard fluorescent DNA detection is

carried out. Different concentrations of double stranded PCR-amplified

DNA sequences with a length of 1 kb are used as specific probe, while the

unspecific probe consists of double stranded salmon sperm DNA of the

same length in a single but much larger concentration (see Table 1).

DNA-detection by the GMR Type Biosensor

Molecular Detection with Magnetic Labels and Magnetoresistive Sensors 41

Table 1. Overview of the different probe DNA spots used in the comparison

experiment.

In the case of the magnetoresistive biosensor, the probe DNA spots are

positioned at the rows of the sensor elements, while a polymer coated glass

slide (TeleChem SuperClean substrate) is used in the case of fluorescent

detection. The optimized spotting solution contains 29% DMSO and the pH

of 10 is adjusted by the addition of 1% TEMED. It denaturates the double-

stranded DNA under test. Afterwards, the probe DNA amino groups are

covalently coupled to the epoxy sites embedded into the polymer. Non-

bound probe DNA is removed in a subsequent washing step, and the

remaining epoxy-groups are inactivated by incubation in a high molar

acetate buffer of pH 5.0 at 55°C.

In the next step, single stranded biotin-labeled (5´ and internal) analyte

DNA complementary to the specific probe DNA with a concentration of

15 nM is hybridized by incubation in a 35% formamide solution at 42°C for

12 hours. Subsequently, non-hybridized analyte DNA is removed by

washing.

Only at this stage, as the markers are being added, the magnetoresistive

biosensor and the fluorescent chip are treated differently. In the case of the

magnetoresistive biosensor, streptavidin-coated Bangs 0.35 µm

microspheres are bound to the biotin-labeled analyte DNA in a neutral

solution at a mass concentration of 1% for one hour at 37°C. Afterwards,

unspecifically bound magnetic markers are washed away. For the

fluorescence sample, Cy3 streptavidin markers are coupled to the biotin-

labeled analyte DNA, and the fluorescence is measured with a laser

scanner. The sensitivity of the laser scanner is adjusted to almost saturation

for the highest specific DNA concentration. The average value of 7-8 spots

for each specific probe DNA concentration is taken relative to the average

fluorescence of the unspecific DNA spots.

42 J. Schotter, M. Panhorst, M. Brzeska, et al.

Fig. 5. Sensitivity comparison of the magnetoresistive biosensor and a comparable

fluorescent detection experiment.

In case of the magnetoresistive biosensor, sensor elements covered by

probe DNA are contacted via Au-wire-bonding. For each sensor element,

the maximum output signal at a magnetizing field of 40 kA/m is taken

relative to the averaged signal of the unspecific sensor elements. These

sensitivity values are plotted together with the fluorescent sensitivities over

the probe DNA concentration in Fig. 5. The standard deviations of the

signals taken from different probe DNA spots of the same concentration are

represented by error bars. Both methods are sensitive to the whole range of

probe DNA concentration (i.e. almost over a range of three orders of

magnitude). At the high concentration region, both sensor types are

saturated, whereas the sensitivity at the lower end is limited by unspecific

signals. However, in the case of the magnetic biosensor, the density of

unspecifically bound markers within unspecific probe DNA spots is the

same as in regions outside of the DNA spots. Contrary to that, there is some

additional background signal within the probe DNA spots in the case of

fluorescent detection, which decreases the relative sensitivity. Therefore, in

this experiment the sensitivity of the magnetic biosensor is superior to the

fluorescent detection at low probe DNA concentrations, for example by a

factor of 2.7 at 600 pM. Due to the good comparability of the parallel

experiments, our magnetoresistive biosensor has proven to be compatible to

standard fluorescence in terms of sensitivity.

Gross R., Sidorenko A., Tagirov L. Nanoscale Devices - Fundamentals and Applications

Подождите немного. Документ загружается.