Langmuir 2006, 22, 263-268


Chemomechanics of Surface Stresses Induced by DNA Hybridization
Jeanne C. Stachowiak,† Min Yue,†,‡ Kenneth Castelino,† Arup Chakraborty,§,| and Arun Majumdar*,†,|
Departments of Mechanical Engineering, Chemical Engineering, and Chemistry, UniVersity of California, and Materials Sciences DiVision, Lawrence Berkeley National Laboratory, Berkeley, California 94720 ReceiVed August 9, 2005. In Final Form: October 14, 2005
When biomolecular reactions occur on one surface of a microcantilever beam, changes in intermolecular forces create surface stresses that bend the cantilever. While this phenomenon has been exploited to create label-free biosensors and biomolecular actuators, the mechanisms through which chemical free energy is transduced to mechanical work in such hybrid systems are not fully understood. To gain insight into these mechanisms, we use DNA hybridization as a model reaction system. We first show that the surface grafting density of single-stranded probe DNA (sspDNA) on a surface is strongly correlated to its radius of gyration in solution, which in turn depends on its persistence length and the DNA chain length. Since the persistence length depends on ionic strength, the grafting density of sspDNA can be controlled by changing a solution’s ionic strength. The surface stresses produced by the reaction of complementary single-stranded target DNA (sstDNA) to sspDNA depend on the length of DNA, the grafting density, and the hybridization efficiency. We, however, observe a remarkable trend: regardless of the length and grafting density of sspDNA, the surface stress follows an exponential scaling relation with the density of hybridized sspDNA.

Downloaded by READING UNIV on September 11, 2009 | Publication Date (Web): November 17, 2005 | doi: 10.1021/la0521645

Biology is replete with molecules such as molecular motors that convert the free energy of chemical reactions into mechanical work, which is critical in many life processes. Such molecules use the interplay between chemical reactions and thermal fluctuations to produce motion in processes that are relatively well understood,1-4 although the details of the effect of molecular structure are still being resolved. Recently, Fritz et al.5 showed that reaction-induced motion can be generated in synthetic mechanical systems as well. They demonstrated that when biomolecular reactions occur on one surface of a cantilever beam, the cantilever bends due to generation of a mechanical surface stress (see Figure 1). This phenomenon has been observed for DNA hybridization,6-8 antigen-antibody binding,9,5,10 and also many nonbiological reactions in vapor.11-13 While it has formed
* To whom correspondence should be addressed. E-mail: † Department of Mechanical Engineering, University of California. ‡ Current address: Applied Biosystems, Inc., Belmont, CA 94002. § Departments of Chemical Engineering and Chemistry, University of California. | Lawrence Berkeley National Laboratory.
(1) Dimroth, P.; Wang, H.; Grabe, M.; Oster, G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 4924-4929. (2) Wang, H.; Oster, G. Appl. Phys. A 2002, 75, 315-323. (3) Bustamante, C.; Keller, D.; Oster, G. Acc. Chem. Res. 2001, 34, 412-420. (4) Berg, H. C. Annu. ReV. Biochem. 2001, 72, 19-54. (5) Fritz, J.; Baller, M. K.; Lang, H. P.; Rothuizen, H.; Vettiger, P.; Meyer, E.; Guntherodt, H. J.; Gimzewski, J. K. Science 2000, 288, 316-318. ¨ (6) Hansen, K.; Ji, H.; Wu, G. H.; Datar, R.; Cote, R.; Majumdar, A. Anal. Chem. 2001, 73, 1567-1571. (7) Wu, G. H.; Ji, H. F.; Hansen, K.; Thundat, T.; Datar, R.; Cote, R.; Hagan, M. F.; Chakraborty, A. K.; Majumdar, A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 1560-1564. (8) McKendry, R.; Zhang, J. Y.; Arntz, Y.; Strunz, T.; Hegner, M.; Lang, H. P.; Baller, M. K.; Certa, U.; Meyer, E.; Guntherodt, H. J.; Gerber, C. Proc. Nat. ¨ Acad. Sci. U.S.A. 2002, 99, 9783-9788. (9) Thundat, T.; Oden, P. I.; Warmack, R. J. Microscale Thermophys. Eng. 1997, 1, 185-199. (10) Wu, G. H.; Datar, R.; Hansen, K.; Thundat, T.; Cote, R.; Majumdar, A. Nat. Biotechnol. 2001, 19, 856-860. (11) Thundat, T.; Warmack, R. J.; Chen, G. Y.; Allison, D. P. Appl. Phys. Lett. 1994, 64, 2894-2896. (12) Berger, R.; Delmarche, E.; Lang, H.; Gerber, C.; Gimzewski, J. K.; Meyer, E.; Guntherodt, H. J. Science 1997, 276, 2022-2024. ¨

Figure 1. Specific, surface-bound biomolecular interactions between target and probe oligonucleotide strands can produce a sufficiently large surface stress to bend a cantilever beam measurably.

the basis for label-free detection of a variety of technologically important biomolecular and chemical reactions,14 the mechanism by which chemical reactions produce surface stresses is yet to be fully understood. A fundamental understanding of the chemomechanics of synthetic systems could lead not only to insight regarding intermolecular forces between biomolecules but also to a variety of novel hybrid mechanical devices. Since much is known about the structure of DNA and the chemistry of DNA reactions, DNA hybridization forms a simple model reaction that can be used for understanding the chemomechanical transduction mechanisms. Hansen et al.6 demonstrated the detection of single nucleotide mismatches using cantilever sensors. McKendry et al.8 detected hybridization at nanomolar concentrations and examined the effect of target and ionic strengths on the signal magnitude during hybridization, fitting
(13) Pinnaduwage, L. A.; Gehl, A.; Hedden, D. L.; Muralidharan, G.; Thundat, T.; Lareau, R. T.; Sulchek, T.; Manning, L.; Rogers, B.; Jones, M.; Adams, J. D. Nature 2003, 425, 6957. (14) Thundat, T.; Majumdar, A. In Sensors and Sensing in Biology and Engineering; Barth, F. G., Humphrey, J. A. C., Seecomb, T. W., Eds.; Springer: New York, 2003.

10.1021/la0521645 CCC: $33.50 © 2006 American Chemical Society Published on Web 11/17/2005

Y.23 All reagents were used without further purification. R. H.. Majumdar. Dedrick. (iii) the osmotic pressure of counterions. which contains individually addressable microfluidic wells. complementary (C) sequences used for hybridization on cantilever surfaces (C-n). B. S. Kannan.19 and persistence length21. Sundaram. and HD-n). fluorescently labeled oligonucleotide probes were first attached on the gold substrate using thiol chemistry (GDn) and then displaced into solution by competitive binding of 12 mM β-mercaptoethanol.. J.. Using DNA as a model molecule. Langmuir 2001. 30.19 Briefly. Gold substrates were incubated in the mercaptoethanol solution for 24 h. Syst.. S. R.. (ii) the conformation entropy of DNA.. C. Gadegaard. H. 2 of 12 cantilevers were broken during fabrication. S. V.. (Bottom) An individually addressable well of the cantilever array chip containing up to 12 sensors and an injection port. solution was measured using a fluorescence microscope and a CCD camera. and sequences used to hybridize probes during hybridization density (HD) measurement (HD-n). Macromolecules 1997. Sturm. SS-n. Thiolated oligonucleotides were synthesized with a 6-mercaptohexyl linker to facilitate adhesion to gold surfaces through gold-thiol bonding. G. 72. A. J.. 795-799. E.16 surface densities. A. 17. Majumdar. M. Phys. A. L. (19) Castelino. C. where the 5′ end is thiolated (HS(CH2)6) and the 3′ end is fluorescently labeled (FAM). J. Reynolds. 2004.. Hagan et al. Table 1. A. Anal. Majumdar. Each cantilever in the well has a ridge structure around the square paddle at its end.. R. Publication Date (Web): November 17. B. A. (16) Strey. A. where the 5′ end is fluorescently labeled (FAM) and the 3′ end is unmodified. Chem. Hagan et al. R. (18) Yue. E 1999. Mucic. Mech. Cui.. N.. A. Stachowiak. Chakraborty. and provide further insight regarding the key parameters that govern the chemomechanics of such hybrid structures. 2004. 5763-5765. We utilized a recently developed 2-dimensional microcantilver array.17. 19561961.. Oligonucleotide surface densities were measured using fluorescence-based techniques on gold-coated substrates (not on cantilever arrays). Satyanarayana. B 2002. M. Science 1996. as shown in Figure 2. C. 106. Microelectromech. J. where n is the number of nucleotides (nt’s). and the oligonucleotides were immediately frozen to prevent the disulfide linkages from re-forming. D.20 Results are interpreted according to published studies of molecular interaction potentials. Lin. 290-299. N. based on the empirical potentials developed by Strey et al. K. K.. The concentration of the displaced probes in (15) Hagan. Furthermore.19.18 The array contains 90 (15 × 6) individually addressable fluidic . Surface Density Measurement Procedure. Downloaded by READING UNIV on September 11. G. Thiolated DNA molecules were treated with 100 mM DTT (dithiothreitol) immediately before use to break any disulfide bonds present. (21) Tinland. 10163-10173.. as described in Castelino et al. thiolated probe oligonucleotides were first immobilized (SS-n) and then hybridized with fluorescently labeled complementary target strands of equal length (HD-n). Table 1 lists the probe and target oligonucleotide sequences used in the experiments (GD-n. 59. Methods Reagent Processing.17. Vol.. 22... No. 1457-1460. All oligonucleotide sequences were purified using HPLC and were obtained from Integrated DNA Technologies Inc. S.acs... 13. 2006 Stachowiak et al. 2009 | http://pubs. 1. Oligonucleotide Nomenclature and Sequencesa name sequence (5′ to 3′) GD-10 HS(CH2)6 5′-TCT CAC CTT C-3′ FAM GD-20 HS(CH2)6 5′-GTG GTA GAT GAA GGT GAG AG-3′ FAM GD-30 HS(CH2)6 5′-GGA GGT GGA CAG ATA ATG GTA GAA GAT AGG-3′ FAM GD-40 HS(CH2)6 5′-CAC TAC GAG TCA AGC TCA CAC ATT CAG GAT TCG GCA TGA T-3′ FAM SS-10 HS(CH2)6 5′-TCT CAC CTT C-3′ SS-20 HS(CH2)6 5′-GTG GTA GAT GAA GGT GAG AG-3′ SS-30 HS(CH2)6 5′-GGA GGT GGA CAG ATA ATG GTA GAA GAT AGG-3′ C-10 5′-GAA GGT GAG A-3′ C-20 5′-CTC TCA CCT TCA TCT ACC AC-3′ C-30 5′-CCT ATC TTC TAC CAT TAT CTG TCC ACC TCC-3′ HD-10 FAM 5′-GAA GGT GAG A-3′ HD-20 FAM 5′-CTC TCA CCT TCA TCT ACC AC-3′ HD-30 FAM 5′-CCT ATC TTC TAC CAT TAT CTG TCC ACC TCC-3′ a Sequences used to measure the grafting density (GD) on gold surfaces (GD-n. where both ends are unmodified.15 provided the first theoretical framework for the chemomechanics in such hybrid structures.. Letsinger. The effects of these parameter variations are correlated to changes in surface density through a previously published fluorescent measurement technique. Viswanadham. DTT was removed by running the solution through commercially available size exclusion columns. Chem. R. 2005 | doi: 10.. Chem. M. Larsen.. The resulting duplexes were displaced from the gold surface using β-mercaptoethanol. A. Bedekar. and (iv) hydration forces between DNA strands.264 Langmuir. Bustamante..20 To find the hybridization density.16 who worked with a hexagonally compressed array of doublestranded DNA. each of which contains multiple microcantilever sensors. (Skokie.. and the concentration of these displaced duplexes was again determined by measuring the fluorescence intensity. after which the displaced fluorescence was seen not to increase further. they highlighted the importance of DNA surface densities. Demers et al. single-stranded (ss) sequences immobilized on cantilever surfaces and used as probes to measure the hybridization density on gold surfaces (SS-n).. have shown this procedure to be highly efficient in displacement of thiolated oligonucleotides from gold surfaces. C-n. A. Biosyst. A. IL). 5535-5541. Figure 2.15 compared the relative importance of various intermolecular forces: (i) electrostatics interactions between neighboring strands.22 and compared to theoretical predictions. This study predicted that hydration forces and osmotic pressure dominate interstrand mechanics. 221-220. Pluen. In this paper. Jenkins.. predicting an exponential increase in cantilever deflection with the density of double-stranded molecules. Persegian. where n is the number of nucleotides (nt’s)). 1. B. Nonspecific attachment of oligonucleotides to the uncoated silicon nitride side of the cantilevers was prevented by self-assembly of a poly(ethylene glycol) (PEG) layer using silane chemistry. W. 22. F. (22) Smith. M.1021/la0521645 the reaction kinetics to a Langmuir isotherm. (20) Demers. where the 5′ end is thiolated HS(CH2)6) and the 3′ end is unmodified. 2000.. 271. Phys. 999-1088. The fabrication process for the microcantilever array chip and the design of optical multiplexing are described in detail in previous papers. (23) Papra. Langmuir 2005. J. (17) Yue. Majumdar. we experimentally examine these predictions. Weill. Elghanian. A. ReV. We utilize a recently developed 2-dimensional microcantilever array. Mirkin. (Top) Cantilever array chip.18 Surface densities are modulated through variations in probe/ target length and ionic strength during probe immobilization. B. H. In the figure.. Microcantilever Array Chip. Podgornik. which functions to create a rigid reflective structure for cantilever tracking.

Woodbury. Surface Density Measurements. it was necessary to construct a ray-optics-based whole field illumination system.5-1 µL) is aspirated just prior to each injection to make room for the new injection. The cantilever tip deflection in response to a temperature change of 1 K was recorded to establish the thermomechanical sensitivity of each cantilever. A. These offsets are averaged for all the cantilevers in the population.24 To simultaneously image the entire cantilever array chip (100-500 cantilevers at a time). 2005. After each injection the cantilevers are allowed to recover a stable position. By measuring the motion of each spot. Langmuir. and the wells were washed three times in a buffer solution of concentration equal to that of the probe solution. each of which can be correlated to a specific grafting . in good agreement with theory. A third injection (not shown) resulted in no further deflection of the cantilevers.18 A micropipet was used to inject 1 µL of 5 µM unlabeled.5 µm thick. the data in this figure are part of the set used to derive the results reported in Figure 5. the cantilever tip deflections were monitored every 15-30 s until the baseline positions of all cantilevers appeared stable. A portion of the buffer within the wells (about 0. and generally 30-40 µm wide. Figure 4B shows the density of hybridized DNA as a function of the ionic strength during attachment of probe ssDNA. the surface stress sensitivity for the devices used in this study has been related to the thermomechanical sensitivity through a mechanical model based on Stoney’s formula and estimated at 24. Each cantilever has a rigid paddle structure at its end to provide a flat reflecting surface for optical diagnostics. A typical set of deflection data from which the results presented in this paper were derived. and show statistical variations. The hybridization solution consisted of 5 µM single-stranded. the ionic strength was fixed at 200 mM. 2009 | http://pubs. Yue et al. Specifically.. All cantilevers were hybridized with fully complementary strands (C20) at 5 µM concentration at a fixed ionic strength of 200 mM. The flat paddles at the end of the cantilevers produced spots on a CCD camera. The response of multiple cantilevers in each well can be averaged to increase the accuracy of the measurements. the offset is recorded for each cantilever. single-stranded DNA do not deflect measurably when exposed to noncomplementary targets that only contain 1-3 consecutive complementary nucleotides over a length of 10-30 nucleotides. Further. 0. and no further injections are needed. Vol. it was possible to detect the deflection of cantilever beams with 1-3 nm resolution. Karnik. 1 M) is apparent. thiolated probe sequences dissolved in sodium phosphate buffer (PB) (10 mM to 1 M) into individual microfluidic wells of the cantilever array. One surface of the cantilevers was coated with a 25 nm thick Au film that was deposited on SiNx with a 5 nm thick Cr adhesion layer. Following the immobilization period. Figure 4C plots the hybridization efficiency. Additional deflection occurs after the second injection. The variation in sensor response for each experiment is accurately reflected in the error bars of Figure 4. J. a laser beam was expanded and reflected off the cantilever array chip. Multiple injections of the hybridization buffer are necessary to ensure that the complementary strand is available to the probe fully concentrated. Cantilevers in each well were incubated in the probe solution (SS-n) (thiolated DNA in sodium phosphate buffer) for at least 3 h. and fresh buffer of the same type and concentration was injected to establish that no significant signal resulted from the injection process alone. A micropipet was used to aspirate the buffer inside each well. fully complementary to and of length equal to that of the immobilized probe DNA (C-n). Yue et al. indicating that the first injection (24) Satyanarayana. No further deflection typically results from a third injection. each of which contains multiple (4-12) cantilever sensors. 2006 265 Downloaded by READING UNIV on September 11. Microelectromech. the ionic strength of the phosphate buffer was varied only during immobilizing or grafting of the probe single-stranded DNA (ssDNA) using thiolAu chemistry. which required temperature control during biological experiments.18 demonstrated that cantilever sensors functionalized with end-thiolated. which represent the first standard deviation.Surface Stresses Induced by DNA Hybridization reaction wells. S. The probe ssDNA (SS-n) of known grafting density was then hybridized using fluorescently labeled complementary target ssDNA (C-n) at 200 mM ionic strength. Multiple injections of the hybridization buffer are used to ensure that the complementary strand is available to the cantilever sensor at its full concentration.17 reported an experimentally observed thermomechanical sensitivity of 208 ( 14 nm/K for the 200 µm long cantilevers. mitigate false readings. Majumdar.25 Cantilever-Based Surface Stress Measurements. and linear drift trends are removed (Figure 3). Thermally induced deflections are calibrated in nanometer units using a white light interferometer (Veeco.19 showed using fluorescence measurements that the density of thiol-attached ssDNA could be controlled by changing the ion concentration from 0.05 to 1 M (Figure 4A). defined as the fraction of probe ssDNA hybridized by complementary target ssDNA as a function of the grafting Publication Date (Web): November 17. Syst. 22. Castelino et al. Once the cantilever array was mounted on the chip holder and aligned with the optical system. After the deflection signals have reached stable values following the injection. NY) to establish the sensitivity per unit of temperature change. inset.acs. 14. No. Results In all the experiments discussed below. The Au-SiNx cantilever formed a thermal bimorph. We measured the change in surface stress resulting from DNA hybridization under various combinations of chain length and ionic strength. After injection of the hybridization buffer (about 1 µL). low-stress silicon nitride and were 200-400 µm long. Greater than 50% of the cantilever response to hybridization was typically observed in the first 5-10 min following injection. the cantilever response to DNA hybridization was optically monitored for at least 1 h following injection. was diluted by liquid remaining in the well after aspiration (just prior to the first injection). with the remainder of the response observed within 1/2 h following injection. dissolved in 200 mM sodium phosphate buffer. unthiolated DNA. Cantilever deflections are set to zero at the time of the hybridization buffer injection. R. During hybridization of the complementary strands.18 Briefly. Here the difference in cantilever deflection for hybridization of 20-nt probes (SS-20) at a 5 µM concentration immobilized at various ionic strengths (10 mM. N. 392-399. Each reaction well was physically separated from its neighboring wells by adhesive bonding of a Pyrex wafer containing an array of wells wet-etched to the cantilever array. the probe solution was aspirated from each well. 1. which is about 2 cm2 in area. The cantilevers were made of silicon-rich. indicating that the sensors have responded to the fully concentrated hybridization buffer.5 mJ/(m2 K). 500 mM.1021/la0521645 Figure 3. The entire cantilever array was then immersed in 200 mM sodium phosphate buffer for at least 10 min prior to being mounted on the temperature-controlled chip holder. Here two injections are visible. 2005 | doi: 10. Most of the deflection occurs after the first injection. Three injections are typically necessary. All biologically induced cantilever deflections are normalized against the cantilever deflection resulting from a unit temperature change.

and mounted on the optical detection system. regardless of chain length. Hydration forces . For a given ionic strength during probe immobilization. To better resolve the relationship between surface stress change during hybridization and ionic strength during immobilization. Cantilever response to hybridization of 10-. Following incubation in the probe solution. as is discussed in the next section. 20.. inset). suggesting. Osmotic forces arise due to the entropic motion of the counterions that are localized around the DNA.266 Langmuir. resulting in a greater cantilever response. to the largest surface stress change (immobilization in 1 M PB). 50 mM. Vol. that surface densities significantly influence the cantilever response. which are expected to yield densities lower than those previously measured by Castelino et al. Hybridization took place at 200 mM ionic strength for all targets. we expect slight decreases in hybridization density and slight increases in efficiency below 50 mM. 100 mM. the cantilever response was observed to increase with increasing ionic strength during immobilization.acs. 500 mM. end-thiolated probe DNA of three lengths (10. but counts them equally with fully hybridized targets.. 200 mM. is expected to increase with increasing ionic strength. (Inset) Cantilever response to hybridization of 10-nucleotide DNA in 200 mM sodium phosphate buffer. The density of partially hybridized strands. Owing to these Publication Date (Web): November 17. 2005 | doi: 10. (C) Hybridization efficiency. the transition from a strong dependence of the grafting density on ionic strengths (at low ionic strengths) to a weaker dependence (at higher ionic strengths) suggests a transition from the dominance of osmotic forces to dominance of hydration forces. and 30 nucleotides) was immobilized on cantilever surfaces at three ionic strengths (without saltsin deionized water.19 but found that difficulties in the experimental procedure prevented the collection of reliable data. as discussed below. The hybridization buffer contained 5 µM single-stranded DNA of length equal to that of and fully complementary to the immobilized probe in 200 mM PB. Discussion The most important observation of this work is that. Downloaded by READING UNIV on September 11. any of the fluorescently labeled target strands that remain attached to displaced probes at the time of the fluorescence measurement will be “counted” by the measurement technique whether they are fully hybridized to the probes or not. thiolated probe DNA as a function of the probe length and ionic strength. The results from this experiment are presented relative (25) We attempted to measure the hybridization density of probe surfaces immobilized at ionic strengths below 50 mM. 100 mM sodium phosphate buffer. as demonstrated by the discovery of a single-exponential scaling relation between surface stress and the density of hybridized DNA strands. Figure 4. 50 mM. and 1 M phosphate buffer. and 1 M sodium phosphate buffer. with moderate salts100 mM PB. if present. (B) Hybridization density of fully complementary DNA as a function of the chain length and ionic strength at which the probe molecules were immobilized. The 10-nucleotide DNA was chosen because it grafts at a higher density. The effectiveness of the washing procedures toward removal of incompletely hybridized targets is not known. these forces are repulsive in nature and are active on length scales comparable to the Debye screening length. This conclusion arises from correlating the surface density and surface stress as a function of ionic strength and DNA chain length. each performed in an individually addressable microfluidic well containing multiple cantilevers. These trends appear to follow those in the surface density measurements. On the basis of the results of Castelino et al. Immobilization took place in 10 mM. Single-stranded. the surface stress measurements for hybridization of probes immobilized in DI water are in all cases correlated to the measurements of the hybridization density and efficiency of probes immobilized in 50 mM PB. 500 mM.19 and to a hybridization density and efficiency in Figure 4. 1. the cantilever response was observed to decrease with increasing probe/target length.1021/la0521645 Figure 5. the surface stress is dominated by the density of hybridization events. No. density in Castelino et al. Figure 5 plots the steady-state surface stress changes for different probe/ target lengths and ionic strengths used during grafting of probe ssDNA. sensors were washed. defined as the fraction of probe ssDNA hybridized by complementary target ssDNA. (A) Grafting density of single-stranded. 200 mM. We note that this measurement technique does not exclude partially hybridized targets or targets that partially hybridize more than one probe. In all figures error bars represent the first standard deviation from measurements on multiple gold substrates and are generally about 5%. 100 mM. permitting the resolution of a greater number of response levels over the range of ionic strengths considered. We do not expect these small changes to significantly affect the trends observed when the two data sets are compared in the Discussion. That is. and 30-nucleotide strands in 200 mM sodium phosphate buffer. 2006 Stachowiak et al. and 1 M) and monitored the surface stress response resulting from their hybridization in 200 mM PB (Figure 5. Immobilization of probe strands occurred in DI water. at various ionic strengths. For monovalent ions. In Figure 4A. For a given probe/target length. and with excess salts1 M PB) for a total of nine experiments. We present these data to show qualitatively the trends observed in the previously described experiment in greater detail. transferred to 200 mM PB. Numbers in parentheses are the numbers of cantilever measurements contributing to the 1 standard deviation error bar. 2009 | http://pubs. we immobilized 10-nucleotide probes at 6 PB concentrations (10 mM. 22. for the range of ionic strengths and DNA chain lengths considered. 20-.

it is well-known that when alkanethiols bind to one surface of a cantilever Publication Date (Web): November 17. When plotted against the grafting density of probe ssDNA (Figure 7B. as seen in our experiments.and 30-nucleotide singlestranded probes. On the basis of this hypothesis. J.16 showed that hydration forces dominate interstrand dynamics at separations below 3. (28) Pincus. and followed an exponential fit reasonably well (Figure 7).B and shown in Figure 7. P. as seen for all the lengths of probes considered. are caused by perturbation of the hydrogen-bonding network of water surrounding DNA molecules. because the length of the 10-nucleotide probes is comparable to the minimum persistence lengths. F. IV 2000. the actual number density of hybridized probes on the surface ends up being much larger as compared to that of the low-concentration case. Strey et al. No. Vol. Schorr.acs. or about 0. 2005 | doi: 10. Hence. the mean spacing between probes observed in this work is significantly larger than the Debye length and thus does not fit the definition of a brush. 1163-1169. which is influenced by the molecular length and ionic strength (through the persistence length). D. all the data points collapsed to yield a single relationship between the surface stress and hybridization density for all lengths of DNA and various ionic strengths. which have a much longer range. the Debye length is much smaller than the hydration length and. Argillier. which reduces interprobe repulsion and leads to a higher grafting density in this intermediate regime. p is the persistence length in angstroms. 2006 267 Downloaded by READING UNIV on September 11. We apply this analysis only to 20.26-28 However. the density of which is controlled by the radius of gyration. the hybridization efficiency is lowered due to steric and electrostatic hindrances experienced by target molecules. we conclude that the key parameter controlling surface stress is the hybridization density of DNA molecules.2 nm. This finding is in agreement with the model by Hagan et al.Surface Stresses Induced by DNA Hybridization Langmuir. Figure 7A shows that. Under these conditions. the immobilization density of probe ssDNA is high and. P.. for 10 bp DNA hybridization. arising from intrastrand sterics at high ionic strengths. K.5 Å. and follows an exponential relation.28 That is.3 nm. However. Macromolecules 1991. since the original number density of immobilized probes is low. Johnson. Sci.. if the Tinland model is used and the molecules are presumed to pack in a dense cubic array on the sensor surface. it would be intriguing to see if the principles learned from DNAsthat surface stress depends on the number of binding events per unit areashold true for other molecular reactions as well. are dominant and determine the spacing between probes. . hence. the resulting predictions of mean center-to-center probe spacing are in good agreement with those arising from grafting density measurements by Castelino et al. However. (27) Guenoun. For example. Acad. The lowest ionic strengths produce the minimum grafting density. the experimentally measured mean center-to-center molecular spacing of singlestranded DNA probes is roughly equal to their radii of gyration in solution. C. they produce a self-assembled monolayer (SAM). D. which creates surface stress to bend the cantilever beam. the distribution of counterions is not uniform in the coordinate direction parallel to the surface. hence.21 have measured the effect of ionic strength on the persistence length of ssDNA and suggested that the radius of gyration of the ssDNA polymer strand can be correlated to the persistence length according to the relation Rg ) (b0Np/3)1/2 Å. Comparison of center-to-center separation distances of surface-grafted probes predicted by grafting density data and by persistence length data21 with the assumption that probe radii are in direct contact. the surface stress increases monotonically with increasing grafting density (decreasing chain length).3 Å for single-stranded DNA. in agreement with Smith et al. the surface stress data in Figure 5 are plotted against the grafting (probe) and hybridization (target) densities of Figure 4A.. 2009 | http://pubs. Ser. Using this result and imposing an intrinsic minimum persistence length of 7. the hybridization densities achieved are still quite low in this regime.. beyond a certain ionic strength. Macromolecules 1998. hydration forces determine the probe spacing. which has been used as a model molecular system. The above discussion has focused on DNA hybridization. Under high ionic strength conditions. P. T. Tinland et al. Frisbie. then depending on the binding affinity adsorption of gas molecules also produces further surface stress.1 chains/nm2. However.. the surface stress (a linear function of the cantilever deflection) is a strong function of the hybridization density.. At low ionic strength conditions.. or both.15 which predicted an exponential increase in cantilever deflection with the density of DNA duplexes. the hybridization efficiency in this regime is high. M. W. a separate trend exists for each ionic strength during immobilization. which suggests that the surface stress is in some way related to either the grafting (probe) or hybridization (target) densities. However. What we found remarkable was that when the surface stress was plotted against the hybridization density. However. and b0 is the proportionality constant between the strand contour length and N ) 4. Tirrell. 4297-4300. The large screening length consequently results in lower probe densities at low ionic strengths. We note that the theory of planar electrolyte brushes and the impact of ionic strength on their formation have been well studied. 31.. 22. M. Tirrell. Antigen-antibody binding and various other protein-protein interactions have been observed to produce (26) Kelley. The results in Figure 5 show that the cantilever mechanical response increases with ionic strength. the decrease in hybridization efficiency is not large enough to account for the higher immobilization density of the probe ssDNA. there are few steric constraints for the target molecules to hybridize with the probe ssDNA. A. the grafting density is relatively constant in this regime. 9. R. 24.27 The results for the hybridization density and efficiency are shown in Figure 4. In Figure 6 we observe that.1021/la0521645 Figure 6. Thus. we may consider the surface of the cantilever not as a set of isolated molecules but as a semicontinuous film. To examine this relationship further. 1. the Debye length is large and the osmotic forces. inset). where the decay length of the interactions is about 0. C. Since the hydration forces do not depend strongly on the ionic strength. (Inset) Ratio of separations predicted by a model and measurement.. Hence.22 we examine the radii of gyration for the molecules of interest in this work.19 That is. Increasing the ionic strength reduces the Debye length. where N is the number of nucleotides.13 When this SAM is exposed to gas molecules. 2912-2919.

Vol. (A) Cantilever response to hybridization of 10-nucleotide DNA in 200 mM sodium phosphate buffer as a function of the hybridization density. which in turn depends on the ionic strength. the grafting density is independent of the ionic strength and hydration interactions dominate. The surface stress response to the grafting density (B. Jerry Hu. which depends on the chain length. 500 mM.S. and 1 M sodium phosphate buffer. inset) is scaled by the hybridization efficiency to produce a single trend that is less influenced by the DNA chain length. We acknowledge the funding through the DARPA Simbiosys program. main) Cantilever response during DNA hybridization as a function of the hybridization density. 2005 | doi: 10. 2009 | http://pubs.9. The facilities provided by the Berkeley Microfabrication Laboratory are much appreciated. We acknowledge helpful suggestions and discussions with Richard J. 100 mM. The surface stress response to the grafting density appears to be influenced by the chain length. 22.C. 2006 Stachowiak et al. and Karolyn Hansen of Oak Ridge National Laboratories. The lessons learned from this work could perhaps be useful in providing insight into the chemomechanics of cell membranes. the implications of our observations on the mechanics of cell membranes are also worth investigation. We show Publication Date (Web): November 17. Immobilization took place in 50 mM. The cantilever response increased monotonically with an increase in the ionic strength used during immobilization. at low ionic strength. we observe that the surface stresses collapse onto a single curve and can be described with a single-exponentially increasing trend. during which much of this work was performed. Downloaded by READING UNIV on September 11. The hybridization . the NCI IMAT program. (B. the osmotic pressure of counterions dominates the intermolecular forces while. the density is governed by the radius of gyration of ssDNA. and it would be worth investigating how the surface stress depends on the density of the binding events.10 Finally. J. No. produces surface stresses on micromechanical structures. LA0521645 Conclusions In this paper. 1. we investigate how DNA hybridization. This finding suggests that the hybridization density brings together the impact of the ionic strength and chain length as a controlling parameter in the creation of surface stress by DNA hybridization in the medium to high ionic strength regime. density and efficiency are also found to depend on the grafting density and the length of the DNA. inset) Cantilever response during DNA hybridization as a function of the grafting density. Life Sciences Division. Measurement of surface stresses generated from DNA hybridization reactions suggests that surface stress is a strong function of the grafting density of probe molecules. Our measurements of the grafting density of probe single-stranded DNA suggest that. acknowledges funding from the ARCS Foundation and the NSF through a graduate fellowship. when the surface stress is plotted against the hybridization density.acs. at low ionic strength. and small molecules that bind to membrane-bound proteins and receptors on a cell surface. (B. Acknowledgment. there are many proteins. at higher ionic strength. Mechanical interactions between these receptor molecules and their interplay with a membrane’s mechanical compliance play a critical role in many cell functions and responses. peptides. Ram Datar. and the DOE Basic Energy Sciences. though it must be taken into account that the dynamics of flexible cell membranes are very different from those of rigid surfaces. A. Cote. surface stress. Two aspects of this investigation include (i) identifying the role of various intermolecular forces and (ii) finding a key parameter that controls mechanical response. However. 200 mM.1021/la0521645 Figure 7. a model biomolecular reaction. which depends on the persistence length. where hydration forces dominate the response. In particular. and Henry Lin of the Department of Pathology at the Keck School of Medicine of the University of Southern California and with Thomas Thundat. acknowledges the support of the Miller Institute through a professorship.268 Langmuir.M.