Kinetics of Hybridization on Surface Oligonucleotide
Microchips: Theory, Experiment, and Comparison with
Hybridization on Gel-Based Microchips
The optimal design of oligonucleotide microchips and efficient discrimination between
perfect and mismatch duplexes strongly depend on the external transport of target DNA
to the cells with immobilized probes as well as on respective association and disso-
ciation rates at the duplex formation. In this paper we present the relevant theory for
hybridization of DNA fragments with oligonucleotide probes immobilized in the cells
on flat substrate. With minor modifications, our theory also is applicable to reaction-dif-
fusion hybridization kinetics for the probes immobilized on the surface of microbeads
immersed in hybridization solution. The main theoretical predictions are verified with
control experiments. Besides that, we compared the characteristics of the surface and
gel-based oligonucleotide microchips. The comparison was performed for the chips
printed with the same pin robot, for the signals measured with the same devices and
processed by the same technique, and for the same hybridization conditions. The sets
of probe oligonucleotides and the concentrations of probes in respective solutions used
for immobilization on each platform were identical as well. We found that, despite the
slower hybridization kinetics, the fluorescence signals and mutation discrimination ef-
ficiency appeared to be higher for the gel-based microchips with respect to their surface
counterparts even for the relatively short hybridization time about 0.5-1 hour. Both the
divergence between signals for perfects and the difference in mutation discrimination
efficiency for the counterpart platforms rapidly grow with incubation time. In particular,
for hybridization during 3 h the signals for gel-based microchips surpassed their surface
counterparts in 5-20 times, while the ratios of signals for perfect-mismatch pairs for gel
microchips exceeded the corresponding ratios for surface microchips in 2-4 times. These
effects may be attributed to the better immobilization efficiency and to the higher ther-
modynamic association constants for duplex formation within gel pads.
The oligonucleotide microchips provide the high-throughput tool for analysis of
single nucleotide polymorphism, expression profiles, and identification of various
microorganisms [reviewed in Refs. (1-5)]. In this technique the probe oligonucle-
otides immobilized either chemically or photolithographically on a surface of a
substrate or within the gel pads bind specifically with the analyzed target DNA in
solution. There are two major classes of surface sensors, the microarrays with oli-
gonucleotide probes immobilized on the surface of flat glass substrate and the mi-
crofluidic devices with probes immobilized on the surface of microbeads immersed
in hybridization solution. The sensitivity in the quantitative measurements of spe-
cific binding via intensity of fluorescence signals depends strongly on the lengths
and sequences of the immobilized oligonucleotides (6), buffer solution [in particu-
lar, on using denaturing agents (7-9)], hybridization kinetics (10-12), temperature
of hybridization, washing protocol, and the other factors. The role and optimal
choice of the different parameters is still the object of intensive investigations.
Journal of Biomolecular Structure &
Dynamics, ISSN 0739-1102
Volume 24, Issue Number 1, (2006)
©Adenine Press (2006)
N. V. Sorokin
V. R. Chechetkin*
S. V. Pan’kov
O. G. Somova
M. A. Livshits
M. Y. Donnikov
A. Y. Turygin
V. E. Barsky
A. S. Zasedatelev
Engelhardt Institute of Molecular Biology
of Russian Academy of Sciences
Vavilov str., 32
Moscow, Russia 119991
Sorokin et al.
Since the time needed for attainment of the complete thermodynamic equilibrium
is commonly rather long, the most of experiments use relatively short incubation
times during hybridization. In these conditions the discrimination between per-
fect and mismatched duplexes depends significantly on kinetics of hybridization
because mismatched duplexes hybridize faster than their perfect counterparts and
perfect-to-mismatch signal ratio is lower in transient regime (10, 12). Although
the experimental studies of hybridization kinetics on the surface oligonucleotide
microchips are rather numerous (10, 13-15; and references therein), the relevant
theory needed for description of kinetic curves and for the estimation of depen-
dence of hybridization time on the density of immobilized oligonucleotides, rates
of association and dissociation related to duplex formation as well as concentration,
and diffusion of target DNA in solution is yet absent. In their pioneering work Chan
et al. (16) considered the idealized hybridization from a local source to the infinite
surface. The theory in the other papers is mainly based on computer simulations of
reaction-diffusion kinetics (17-19). In this work we present the approximate ana-
lytical description for hybridization on the surface cells and on the microbeads and
compare theoretical predictions with experimental results. To model and to process
experimental kinetic curves, we also developed the special package of programs
ImaGel Kinetics comprising both surface hybridization and reaction-diffusion ki-
netics within gel pads studied previously (20-21).
The advance in microchip technology evoked the urgent problem of assessment
and comparison of different platforms. Commonly, the comparison is performed
for the microchips obtained from the different manufacturers and with hybridiza-
tions using particular manufacturers’ protocols (22-24). We compared the kinetic
behavior for the signals and mutation discrimination efficiency for the surface
and gel-based oligonucleotide microchips using the same sets of immobilized
oligonucleotides and the same hybridization conditions for both platforms. It is
also important that the chips of both kinds were made on the same equipment;
the signals were measured with the same devices; and the concentrations of oli-
gonucleotides in solutions used for immobilization were equal to each other. The
lengths of probe oligonucleotides in our work were about 20 nt, while the length
of target fragment was equal to 126 nt. Both lengths are quite typical in the
various applications. For example, Affymetrix GeneChip system uses the probe
oligonucleotides 25 nt in length and in Amersham CodeLink platform the oligo-
nucleotides of length 30 nt are immobilized within gel pads.
Our results show that, despite the slower hybridization kinetics, the fluorescence
signals, and mutation discrimination efficiency appear to be higher for the gel-
based rather than surface oligonucleotide microchips even for the relatively short
hybridization time about 0.5-3 h. The difference rapidly grows with incubation
time. These effects may be explained by the better immobilization of probes
in the gel networks and the higher thermodynamic association constants at the
formation of duplexes within gel pads.
Materials and Methods
Manufacturing of Surface Oligonucleotide Microchips
The microchips with oligonucleotide probes immobilized on the glass surface were
manufactured as described (25-26) with minor modifications.
Microchips were printed on silylated (aldehyde-coated) glass slides (3 in. by 1 in.;
Cell Associate, Inc.). For the surface microchips designated for comparison with
their gel-based counterparts the spotting mixture contained 60 pmole/μl (or 60 μM)
oligonucleotide probe in 50% aqueous glycerol. The printed slides were placed in
a closed chamber contained 50% aqueous glycerol to prevent rapid evaporation of
spotted drops and heated at 45 ºC for 24 h. To make the bonds between oligonucle-
Kinetics of Hybridization
on Surface Microchips
otides and the glass surface irreversible, the slides were incubated for 5 min in a
freshly prepared 0.25% aqueous solution of NaBH4, followed by washing once with
a 0.2% aqueous solution of sodium dodecyl sulfate (SDS) for 1 min and twice with
distilled water (1 min) to remove unbound oligonucleotides. Then slides were dried
for 1 h at 23 ºC and used for hybridization.
The hybridization kinetics was studied with specially designed surface microchips.
Each chip contained four identical sets of five cells with variable density of im-
mobilized probe 5ʹ-TTC TGG TCC ATG AAT TGG-3ʹ. The concentrations of oli-
gonucleotides in solution used for immobilization of probes in different cells, csol,
were equal respectively to 200, 160, 100, 60, and 40 pmole/μl. As shown below, the
surface density of immobilized probes was approximately proportional to csol.
The cells for both the surface and gel-based microchips were printed with a pin ro-
bot QArray (Genetix) delivering approximately 2 nl of a spotting mixture per spot.
The radii of the cells were equal to 150 μm, while the distance between the centers
of the cells was 550 μm. The work was completed on a batch of five identical oli-
gonucleotide microchips of each kind.
Manufacturing of Gel-Based Oligonucleotide Microchips
The manufacturing of microchips with gel-immobilized oligonucleotides followed
copolymerization method as described (27) with several modifications. Glass slides
(Corning 2947 Micro Slides, Corning Glass Works, Corning, NY) were treated with
Bind Silane (Amersham Biosciences), washed in ethanol and deionized water, and
air-dried under dust-free conditions.
Solutions for the preparation of the copolymerization chips contained methacryl-
amide (4.75%, w/v), N,Nʹ-methylenebisacrylamide (Bis) (0.25%, w/v), glycerol
(50%, v/v), N,N,Nʹ,Nʹ-tetramethylenediamine (TEMED) (5%, v/v), and oligonucle-
otides with terminal aliphatic amino group.
The polymerization of gel pads was induced by illumination with UV light with
maximal wavelength 350 nm and irradiation intensity 0.06 μW/cm2 (GTE lamp
F15T8/350 BL; Sylvania, Danvers, MA) for 40 min at 20 ºC. After polymeriza-
tion, the microchips were washed in 0.1 M PBS buffer containing 0.1% Tween 20,
then deionized water, and stored dry.
Oligonucleotide Probes and Target DNA
The oligonucleotides for immobilization were synthesized in a 0.2- to 1-μmol scale
using a 394 DNA/RNA synthesizer (Applied Biosystems, Foster City, CA) by stan-
dard phosphoramidite chemistry (27). To immobilize oligonucleotides in gel pads,
an amino group was introduced during synthesis using 3ʹ-Amino-Modifier C7 CPG
500. Oligonucleotides were purified by reverse-phase HPLC on a Supelco LC-18
column (5 μm, 4.6 × 250 mm2) eluted with either 0.05 M triethylammonium acetate
(TEAA), pH 7.0, or 0.05 M TEAA with gradient of CH3CN from 0 to 50%.
The oligonucleotides for comparison of characteristics for surface and gel-based
microchips are presented in Table I. They were chosen from a set of probes used
earlier for the detection of rifampin-resistant mutations in Mycobacterium tuber-
culosis genome (8). The sequences of the oligonucleotides were complementary
to the different segments of the rpoB gene. The resulting perfect duplexes cor-
respond to the probes denoted by “P” in Table I, while the counterpart 1-base
mismatches are denoted by “M” with the same number as for “P” (variable bases
are marked by bold letters in Table I). Unlike the other pairs designated for mu-
tation identification, the pair P3-M3 serves for identification of deletion (marked
by bold letters in sequence P3).
Sorokin et al.
Target samples of DNA from M. tuberculosis were prepared by two-stage PCR as
described (8, 12). The target DNA sequence of 126 nt long corresponded to the
wild-type fragment of the rpoB gene (nucleotides 2336-2461, GenBank Acc. No.
L27989). The target DNA was labeled by Texas Red analogously to (8, 27). The
concentration of hybridized target DNA in buffer solution was about 10-7 M.
Hybridization and Measurement of Fluorescence Signals
All experiments were carried out in 35 μl hybridization chamber. Hybridization
mixtures were prepared by adding 12 μl of the second-stage PCR mixtures to 24
μl of 1.5 M guanidine thiocyanate (GuSCN), 0.075 M HEPES, pH 7.5, 7.5 mM
EDTA. The hybridization chamber of the microchip was filled with this mixture
and incubated at 37 ºC.
The fluorescence signals during hybridization were measured in real time each 300
seconds using an automatic experimental setup consisting of a fluorescent micro-
scope (LOMO, Russia), CCD-camera (SenSys 1602E, 1024 × 1536 pixels, Roper
Scientific Inc., Tucson, AZ), Peltier thermotable, temperature controller, and a com-
puter equipped with a data acquisition board. The microscope had Texas Red exci-
tation and emission filters (28). Data processing and hybridization curve analysis
were performed as described earlier (21).
The fluorescence signal was defined according to the formula
where C(t) is the median value for fluorescence at moment t calculated for the im-
age area occupied by a cell, while C(0) is the counterpart fluorescence at the initial
moment t = 0. To take into account the possible spatial inhomogeneity of illumina-
tion source, the microchip slide was replaced by the slide of red glass of identical
sizes and the corresponding fluorescence within a position occupied by a cell, Br.g.,
was measured. This value was corrected by the noise signal Bd.c. produced by dark
current at zero illumination intensity.
Figure 1 illustrates the counterpart signals for the gel-based and surface oligonucle-
otide microchips after hybridization with target DNA.
Results and Discussion
In this section we present the approximate theory for reaction-diffusion hybrid-
ization kinetics on the surface oligonucleotide microchips and on the microbeads.
Immobilized Oligonucleotide Probes.
G/C content, %
J(t) =C(t) – C(0)
Br.g. – Bd.c.
Figure 1: The counterpart fragments of gel-based (A)
and surface (B) oligonucleotide microchips after hy-
bridization with target DNA.
Kinetics of Hybridization
on Surface Microchips
The evolution of the volume concentration of the target DNA in solution, h(r
governed by the diffusion equation
where Dsol is diffusion coefficient for target DNA in solution. At a boundary of a
cell with the immobilized probes the balance related to formation and dissociation
of duplexes is determined by diffusive flow,
while at a surface beyond a cell the diffusive flow should be equal to zero
from solution to substrate, m
d ~ is the surface density of the formed duplexes, while kass and kdiss are the associa-
tion and dissociation rates respectively. Equations  and  should be solved
with the initial conditions
b is the unit vector normal to a boundary of a substrate surface and directed
~ is the density of the immobilized probes per unit area,
away from a cell.
→,t) should tend to the volume concentration of target DNA hsol far
If R2/Dsol << τ(H)
teristic hybridization time defined below in Equation ), then at t >> R2/Dsol the
asymptotic distribution of concentration of the target DNA in solution beyond cell
region can be approximated with a reasonable accuracy by quasistationary profile,
diff (where R is the radius of cell or microbead and τ(H)
diff is the charac-
where hsol is the homogeneous concentration of target oligonucleotides in solution
far away a cell with immobilized probes and h(t) is the concentration of target oli-
gonucleotides at the cell surface.
The substitution of concentration  into Equation  yields the relationship
where the numerical factor β is equal to unity for hybridization on the microbeads
and on the semispherical surface cells, while in the case of hybridization on the
flat surface cells it introduces the necessary geometric correction. After resolving
h(t) from Equation  and substituting it into the reaction part of Equation , the
explicit integration provides the dependence,
Here J(t) is current fluorescence signal evolving to its saturation value Jmax at ther-
Jmax = Amηeq
∂t= -Dsolnb · ∇h|r∈cell = kass(m – d)h|r∈cell – kdissd
Dsolnb · ∇h|r∈surface; r∉cell = 0
d(r,t)|r∈cell, t=0 = 0
h(r,t) = (hsol – h(t))(1 –
) + h(t)
(hsol – h(t))
(1 – ηeq)(1 +
) 1n(1 –
) – ηeq = -
1 + Kahsol
βDsol(1 + Kahsol)
Sorokin et al.
where A is apparatus constant and Ka = kass/kdiss is thermodynamic association con-
stant. The factor ηeq characterizes the equilibrium fraction of probes participating
in duplex formation. Note that the solution  is a bit more general than the par-
ticular approximation  used at its derivation and is applicable for a broader class
of approximations with separable variables.
If a product τ(H)
hybridization, which is determined by exponent exp(-kasshsolt – kdisst) similarly to
hybridization in solution. In such a limit the kinetics does not depend either on
diffusion coefficient Dsol or on the surface density of immobilized oligonucle-
otides m ~. Alternatively, if τ(H)
be limited by diffusion and the characteristic hybridization time is determined by
Equation . This time characterizes the intermediate stage of hybridization,
while the asymptotic evolution to saturation is governed by exponential law. The
values R ~ 10-2 cm, m
~ ~ 1012 molecule/cm2, Dsol ~ 10-6 cm2/s, and Ka ~ 108 ÷ 109
M-1 give the estimate τ(H)
diff kdiss << 1, then evolution is reduced to the spatially homogeneous
diff kdiss >> 1, then hybridization kinetics becomes to
diff ~ 103 - 104 s.
The counterpart evolution equation for washing is given by
Here J(0) is the initial fluorescence signal before washing,
and h0 corresponds to the concentration of target DNA in solution for preced-
At fitting experimental kinetic curves, Equations  and  depending on the
parameters η, τdiff, and τdiff kdiss permit to determine only two parameters from three.
Thus, either the parameter η should be determined beforehand by varying the con-
centration of hybridized DNA in solution hsol and measuring the signals at satura-
tion or the limit of large values of product τdiff kdiss should be assumed.
Hybridization Kinetics on the Surface Oligonucleotide Microchips
The diffusion-limited character of hybridization kinetics manifests itself in the de-
pendence of hybridization kinetics on the surface density of probes m
perimental conditions the proportionality of fluorescence signals at saturation Jmax
to the concentration of probes in solution for immobilization csol (see Figure 2)
proves that the values of m
~ also should be proportional to csol. Therefore, we will
use csol instead of m
~. In our ex-
~ in subsequent analysis.
The experimental kinetic curves are presented in Figure 3. Remind that in our
experiments each chip designed for the study of concentration dependence of hy-
bridization kinetics contained four identical clusters of five cells with gradated con-
centrations of probes m
~. Thus, the data in Figure 3 correspond to the median curves
from four clusters. The comparison of experimental (solid curves) and theoretical
(broken curves) results for the normalized signals J(t)/Jmax are shown in Figure 4.
Finally, Figure 5 provides the dependence on csol for time τ0.9 corresponding to 0.9
level of saturation and the counterpart dependence for time τ(H)
fitting of parameters in Equation . The other parameters for the fitted curves
were as follows, ηeq = 0.05; τ(H)
60, and 40 pmole/μl. The partial divergence between experimental and theoretical
curves in Figure 4 may be explained by the relative crudeness of approximation
used at the derivation of Equation . The agreement between experimental results
diff obtained by the best
diff kdiss = 25, 20, 15, 7.5, and 5 for csol = 200, 160, 100,
) + η0(1 –
) = -
1 + Kah0
Figure 2: Dependence of signals at saturation on the
concentration of oligonucleotide probes in solution used
for immobilization, csol. The straight line corresponds to
the best linear fit of experimental results.
Kinetics of Hybridization
on Surface Microchips
and theoretical predictions for the characteristic time τ(H)
5B). The concentration dependence for time τ0.9 shown in Figure 5A also appears
to be in a good correspondence with theory. Thus, the theory may be applied to the
assessment of hybridization time for the surface oligonucleotide microchips and its
dependence on various parameters.
diff is much better (cf. Figure
Comparison of Signals and Mutation Discrimination Efficiency for Surface and
Gel-Based Oligonucleotide Microchips
The molecular interactions between oligonucleotide probes immobilized within gel
pads and target DNA remain very similar to the hybridization in solution (29). In
contrast, the interference of glass substrate seriously affects both thermodynamic
and kinetic characteristics at the duplex formation on the microbeads (15) and on
the surface microchips (30-31). Besides that, there is difference in the efficiency
of probe immobilization on the glass substrate and within gel networks. The latter
effect remains practically uninvestigated. For this reason, we held the same con-
centration of probes in respective solutions used for immobilization of oligonucle-
otides on the surface and gel-based microchips and printed the cells with the same
pin robot. The sequences of probes (presented in Table I) as well as hybridization
conditions also were identical. The fluorescence signals were measured with the
same equipment and processed by the same scheme. Therefore, if the immobiliza-
tion efficiency and the molecular interactions would be the same for surface and
gel-based microchips, then the intensities of fluorescence signals and the sensitivity
in mutation discrimination should coincide for both platforms.
We found, however, that both signals and mutation discrimination efficiency are
strongly different for surface and gel-based microchips. As shown in Figure 6,
these effects are pronounced even for hybridization during relatively short inter-
val 0.5-3 h and rapidly increase with incubation time. Although the hybridization
kinetics is slower for the gel-based microchips, the fluorescence signals appear to
be stronger and the discriminant ratios Jperfect/Jmismatch for the perfect/mismatch pairs
turn out higher than their counterparts on the surface oligonucleotide microchips.
The stronger signals for gel-based microchips are partially due to the better immo-
bilization efficiency for this platform (at least, for the protocols used in our work).
The behavior of discriminant ratios shows that the role of molecular interactions
and the higher thermodynamic association constants Ka for gel chips are also sig-
nificant. It is worth noting that the discrimination between perfect and mismatch
pairs fails in about 15-35% of cases for GeneChip data collection (32-33).
The posterior application of washing can improve perfect/mismatch ratios for the
surface microchips, but would cause the subsequent drop in the relatively weak
fluorescence signals. If necessary, the procedure of washing might be applied to the
gel-based microchips as well (12).
Our experiments on the comparison of surface and gel microchips support the pre-
vious observations (10, 12) that hybridization kinetics for mismatch duplexes is
Figure 3: Dependence of hybridization ki-
netics on incubation time for the different
concentrations of probes in solution for im-
mobilization, csol = 200 (n), 160 (l), 100
(®), 60 (°), and 40 (t) pmole/μl.
Figure 4: Evolution of fluorescence signals J(t)/Jmax
normalized with respect to their equilibrium values for
the different concentrations of probes in solution for im-
mobilization, csol = 200 (n), 160 (l), 100 (®), 60 (°),
and 40 (t) pmole/μl. The solid curves correspond to
the experimental results, while the broken curves cor-
respond to the best fitting obtained with Equation .
Figure 5: The dependence on csol for time τ0.9 cor-
responding to 0.9 level of saturation (A) and for time
tion  (B).
diff obtained by the best fitting of parameters in Equa-
Sorokin et al.
faster than for perfect duplexes (data not shown). These observations are in a good
accordance with Equation , since the hybridization time grows with the increase
of association constants, which are higher for the perfect duplexes.
Three-dimensional gel networks possess the higher capacity for immobilization
of molecules with respect to immobilization on a two-dimensional surface. As is
seen from our results, the immobilization efficiency within gel pads appears to be
higher as well. Besides that, the substrate surface introduces the significant steric
restrictions to the molecular interactions between hybridized strands. The addi-
tion of special linkers for suppression of steric hindrance imposed by a surface
diminishes the permissible density of immobilized probes (14, 34), because the
“forest effects” hamper the accessibility of probe molecules at high densities (13-
14). Since the association constants for the longer DNA strands are higher and
the impact of secondary and tertiary structures on the molecular interactions be-
comes stronger, the lengthening of probes will simultaneously lead to the longer
hybridization times. In contrast with surface microarrays, the addition of linkers
is not needed in gel-based microchips, because the molecular interactions in gel
pads and in solution remain similar. Note also that the output of synthesis de-
creases, while the variation in lengths of synthesized oligonucleotides increases
with lengthening of probe oligonucleotides (13).
Some principal differences between surface and gel microchips might arise at the
analysis of very long DNA strands or large protein molecules. It is yet known that
DNA strands of more than 500 base pairs in length can penetrate into gels via rep-
tation mechanism (35-36). The hybridization of DNA up to 500 nt in length was
actually demonstrated for gel-based oligonucleotide microchips (27).
The study of hybridization kinetics on the surface oligonucleotide microchips re-
tains the previous conclusions about the competing roles of thermodynamic and
kinetic effects for the choice of optimal parameters at the chip design (12). Equa-
tions - show that such factors, as thermodynamic association constant Ka and
the surface density of immobilized probes m
from the cells and determine the sensitivity of measurements, yield at the same time
the slower saturation kinetics and the lower perfect-to-mismatch signal ratio in the
transient regime. These factors slow down the washing kinetics as well (see Equa-
tions -). Thus, their optimal choice depends on the sensitivity of recording
apparatus and limiting time of analysis.
~, which enhance fluorescence signals
To shorten the time of analysis, the reaction-diffusion kinetics on the surface or
gel microchips may be accelerated by applying electric field (37) or by enhanc-
ing external transport with acoustic waves (38) or hydromechanical circulation
(39). The previous study of the reaction-diffusion kinetics on the gel-based pro-
tein microchips with external transport enforced by mixing with peristaltic pump
(40) makes plausible the suggestion that the enhancement of transport may be
described by the replacement of diffusion coefficient Dsol in Equation  by some
effective value dependent on mixing conditions.
The stronger fluorescence signals on the gel-based microchips allow more easily
performing kinetic measurements under buffer with the labeled molecules, while
the higher discrimination efficiency between perfect and mismatch duplexes per-
mits to avoid the washing procedure. As was proved earlier, the surface oligonucle-
otide microchips prepared according to protocol described in Refs. (25-26), can be
stored for two months at room temperature, while the gel-based microchips can be
stored for one year at the same conditions (27). All these advantages overcome
likely the slower kinetics within gel pads.
Figure 6: The comparison of characteristics for the gel-
based and surface oligonucleotide microchips. (A) The
dependence on hybridization time for the ratio of sig-
nals corresponding to the counterpart perfect duplexes
formed on the gel-based and surface microchips. (B)
Evolution of discriminant ratio, Jperfect/Jmismatch, for the
perfect/mismatch pairs on the gel-based (solid curves
and upright numbers) and surface (broken curves and
italic numbers) microchips. The nomenclature of probes
and perfect/mismatch pairs corresponds to Table I.
Kinetics of Hybridization
on Surface Microchips
The authors are very thankful to V. Chizhikov (Center for Biologics Evaluation
and Research, Food and Drug Administration, Maryland) for delivering the pro-
tocols for the manufacturing of surface oligonucleotide microchips and useful ad-
vises. We are also very indebted to V. Mikhailovich, D. Gryadunov, K. Evseev,
and R. Yurasov for the help during this work and valuable discussions. This work
is supported by CRDF grant assistance program (project RUC2-11036-MO-04)
and ISTC grants Nos. 2508 and 2906.
References and Footnotes
Jain, K. K. Pharmacogenomics 1, 289-307 (2000).
Chipping Forecast II. Nature Genetics 32, Supplement (2002).
Kolchinsky, A., Mirzabekov, A. D. Human Mutation 19, 343-360 (2002).
Stears, R. L., Martinsky, T., Schena, M. Nature Medicine 9, 140-145 (2003).
Ochs, M. F., Godwin, A. K. BioTechniques 34, S4-S15 (2003).
Kane, M. D., Jatkoe, T. A., Stumpf, C. R., Lu, J., Thomas, J. D., Madore, S. J. Nucleic Acids
Res. 28, 4552-4557 (2000).
Blake, R. D., Delcourt, S. G. Nucleic Acids Res. 24, 2095-2103 (1996).
Mikhailovich, V., Lapa, S., Gryadunov, D., Sobolev, A., Strizhkov, B., Chernyh, N.,
Skotnikova, O., Irtuganava, O., Moroz, A., Litvinov, V., Vladimirskii, M., Perelman, M.,
Chernousova, L., Erokhin, V., Zasedatelev, A., Mirzabekov, A. J. Clinical Microbiology 39,
Drobyshev, A. L., Machka, C., Horsch, M., Velmann, M., Liebscher, V., De Angelis, M. H.,
Beckers, J. Nucleic Acids Res. 31, e1 (2003).
Dai, H., Meyer, M., Stepaniants, S., Ziman, M., Stoughton, R. Nucleic Acids Res. 30,
Dorris, D. R., Nguen, A., Gieser, L., Lockner, R., Lublinsky, A., Patterson, M., Touma, E.,
Sendera T. J., Elghanian, R., Mazumder, A. BMC Biotechnology 3, 6-16 (2003).
Sorokin, N. V., Chechetkin, V. R., Livshits, M. A., Pan’kov, S. V., Donnikov, M. Y., Gryadunov,
D. A., Lapa, S. A., Zasedatelev, A. S. J. Biomol. Struct. Dyn. 22, 725-734 (2005).
Forman, J. E., Walton, I. D., Stern, D., Rava, R. P., Trulson, M. O. ACS Symposium Series
682, 206-228 (1998).
Peterson, A. W., Heaton, R. J., Georgiadis, R. M. Nucleic Acids Res. 29, 5163-5168 (2001).
Sekar, M. M. A., Bloch, W., St. John, P. M. Nucleic Acids Res. 33, 366-375 (2005).
Chan, V., Graves, D. J., McKenzie, S. E. Biophys. J. 69, 2243-2255 (1995).
Erickson, D., Li, D., Krull, U. J. Anal. Biochem. 317, 186-200 (2003).
Bhanot, G., Louzoun, Y., Zhu, J., DeLisi, C. Biophys. J. 84, 124-135 (2003).
Gadgil, C., Yeckel, A., Derby, J. J., Hu, W.-S. Journal of Biotechnology 114, 31-45 (2004).
Livshits, M. A., Mirzabekov, A. D. Biophys. J. 71, 2795-2801 (1996).
Sorokin, N. V., Chechetkin, V. R., Livshits, M. A., Vasiliskov, V. A., Turygin, A. Y.,
Mirzabekov, A. D. J. Biomol. Struct. Dynamics 21, 279-288 (2003).
Tan, P. K., Downey, T. J., Spitznagel Jr., E. L., Xu, P., Fu, D., Dimitrov, D. S., Lempicki, R.
A., Raaka, B. M., Cam, M. C. Nucleic Acids Res. 31, 5676-5684 (2003).
Yauk, C. L., Berndt, M. L., Williams, A., Douglas, G. R. Nucleic Acids Res. 32, e124 (2004).
Shippy, R., Sendera, T. J., Lockner, R., Palaniappan, C., Kaysser-Kranich, T., Watts, G.,
Alsobrook, J. BMC Genomics 5, 61 (2004).
Chizhikov, V., Wagner, M., Ivshina, A., Hoshino, Y., Kapikian, Z., Chumakov, K. J. Clin.
Microbiol. 40, 2398-2407 (2002).
Volokhov, D., Rasooly, A., Chumakov, K., Chizhikov, V. J. Clin. Microbiol. 40, 4720-
Rubina, A. Y., Pan’kov, S. V., Dementieva, E. I., Pen’kov, D. N., Butygin, A. V., Vasiliskov,
V. A., Chudinov, A. V., Mikheikin, A. L., Mikhailovich, V. M., Mirzabekov, A. D. Anal.
Biochem. 325, 92-106 (2004).
Barsky, Y., Grammatin, A., Ivanov, A., Kreindlin, E., Kotova, E., Barskii, V., Mirzabekov, A.
J. Opt. Technol. 65, 938-941 (1998).
Vasiliskov, V., Prokopenkо, D., Mirzabekov, A. Nucleic Acids Res. 29, 2303-2313 (2001).
Shchepinov, M. S., Case-Green, S. C., Southern, E. M. Nucleic Acids Res. 25, 1155-
Peterson, A. W., Heaton, R. J., Georgiadis, R. J. Am. Chem. Soc. 122, 7837-7838 (2000).
Naef, F., Lim, D. A., Patil, N., Magnasco, M. Phys. Rev. E 65, R040902-1−R040902-4 (2002).
Binder, H., Preibisch, S. Biophys. J. 89, 337-352 (2005).
Steel, A. B., Levicky, R. L., Herne, T. M., Tarlov, M. J. Biophys. J. 79, 975-981 (2000).
De Gennes, P. G. Scaling Concepts in Polymer Physics. Cornell University Press, Ithaca and
Pluen, A., Netti, P. A., Jain, R. K., Berk, D. A. Biophys. J. 77, 542-552 (1999).
Heaton, R. J., Peterson, A. V., Georgiadis, R. M. Proc. Natl. Acad. Sci. USA 98, 3701-
66 Download full-text
Sorokin et al.
Toegl, A., Kirchner, R., Gauer, C., Wixforth, A. J. Biomol. Techn. 14, 197-204 (2003).
Meldrum, D. Genome Res. 10, 1288-1303 (2002).
Zubtsov, D. A., Ivanov, S. M., Rubina, A. Yu., Dementieva, E. I., Chechetkin, V. R.,
Zasedatelev, A. S. Journal of Biotechnology 122, 16-27 (2006).
Date Received: February 15, 2006
Communicated by the Editor Valery Ivanov