22060
J. Phys. Chem. B 2006, 110, 22060-22063
ARTICLES Bending and Radial Deformation of Lipid Tubules on Self-Assembled Thiol Monolayers† Yue Zhao, Nidhi Mahajan, and Jiyu Fang* AdVanced Materials Processing and Analysis Center and Department of Mechanical, Materials, and Aerospace Engineering, UniVersity of Central Florida, Orlando, Florida 32816 ReceiVed: September 5, 2005; In Final Form: NoVember 29, 2005
Lipid tubules represent a hollow, cylindrical supramolecular structure formed by rolled-up lipid bilayers. We find that the lipid tubules of 1,2-bis(tricosa-10,12-diynoyl)-sn-glycero-3-phosphocholine can be bent into a loopike shape by the shrinking contact line of droplets on self-assembled monolayers (SAMs) of 1-dodecanethiol. The persistence length of individual lipid tubules is estimated to be ∼41 µm. The radial deformation of the lipid tubules on SAMs is studied under applied load using atomic force microscope. The stiffness of the tubules in the radial direction is found to increase when the number of the lipid bilayers in the tubule wall increases.
1. Introduction Self-assembled lipid tubules represent interesting hollow cylindrical supramolecular structures.1,2 A number of synthetic lipids with modified headgroups or acyl chains have been found to self-assemble into tubule structures in solutions.3-14 The diameter of these lipid tubules spans over a range from 10 nm to 1.0 µm, depending on the nature of the lipids and the conditions under which molecular self-assembly occurs. The hollow, cylindrical shape and the crystalline molecular order of the bilayer walls make lipid tubules attractive as a template for the synthesis of one-dimensional (1D) inorganic materials with high aspect ratio,15-21 a substrate for the helical crystallization of proteins,22,23 and a controlled release system for drug delivery.24-26 An understanding of the mechanical properties of lipid tubules is critical in developing their applications in nanotechnology. So far, little is known about their mechanical properties with a few exceptions. For example, Shimizu and co-workers27 reported the bending of a lipid nanotube of synthetic cardanyl-β-D-glucopyranoside with optical tweezers. They converted the bending stiffness K of the lipid nanotube into the Young’s (bending) modulus, E ) K/I, where I is the geometrical moment in inertia and then estimated that Young’s modulus of the nanotube is about 720 MPa. In a previous publication,28 we used microfluidic networks (µFNs) to align and bend lipid tubules of 1,2-bis(tricosa-10,12-diynoyl)-snglycero-3-phosphocholine (DC8,9PC) on glass substrates. It was found that the lipid tubules could be bent into a well-defined shape at the entrance of the µFN by the capillary force-induced channel flow. The shape of the bent tubules reflects the structure of the capillary entrance. In this paper, we find that the DC8,9PC tubules can be bent into a loopike shape by a shrinking contact line of droplets on self-assembled monolayers (SAMs) of 1-dodecanethiol. The †
Part of the special issue “Charles M. Knobler Festschrift”. * To whom correspondence should be addressed. E-mail: jfang@ mail.ucf.edu.
persistence of the lipid tubules is estimated to be ∼41 µm. We also study the radial deformation of the DC8,9PC tubules on SAMs under applied load using an atomic force microscope (AFM). The stiffness of the tubules in the radial direction is found to increase when the number of lipid bilayers in the tubule wall increases. 2. Experimental Section Lipid tubules used in our experiments were prepared by cooling a 5 mg/mL suspension of 1,2-bis(tricosa-10,12-diynoyl)sn-glycero-3-phosphocholine (DC8,9PC) (Avanti Polar Lipids, Alabaster, AL) in ethanol/water (70:30 v/v) from 60 °C to room temperature at a rate of ∼0.5 °C/min. The polymerization of the DC8,9PC tubule suspension was performed with UV irradiation (254 nm) for 20 min at room temperature. 1-Dodecanethiol (DDT) (Aldrich) was dissolved in ethanol. SAMs were prepared on Au-coated mica substrates with the (111) surface (Molecular Imaging Inc.) through adsorption from 1 mM DDT ethanol solution. A drop of tubule solution was dried on the hydrophobic DDT SAMs in air at room temperature. An optical microscope (BX 40 Olympus) with a digital camera (Olympus C2020 Zoom) and an AFM (Dimension 3100, Digital Instruments) were used to image the DC8,9PC tubules dried on the SAMs. Silicon nitride cantilevers with a normal spring constant of 0.2 N/m and 30 N/m were used in contact and tapping modes, respectively. The size of the cantilever tips (radius of curvature) is about 20 nm according to the manufacturer. All AFM measurements were performed at a scan rate of 0.5 Hz in air at room temperature. 3. Results and Discussion Water contact angle on DDT SAMs is measured to be ∼105° at room temperature. When droplets of DC8,9PC tubule solution are dried on the hydrophobic DDT SAMs in air, we find that the tubules are bent into a loopike shape (parts a and b of Figure 1). The orientation of the tubule loops appears to be random on the SAMs. The tubules bent in a single droplet can form a
10.1021/jp0550199 CCC: $33.50 © 2006 American Chemical Society Published on Web 01/10/2006
Deformation of Lipid Tubules on Thiol Monolayers
J. Phys. Chem. B, Vol. 110, No. 44, 2006 22061
Figure 1. (a-c) Optical microscopy images and (d) AFM image of the bent tubules with a loop shape on the DDT SAM.
Figure 2. Schematic of the suggested mechanism for the bending of lipid tubules by the shrinking contact line of a droplet of tubule solution on the DDT SAM.
near circular structure (Figure 1c). The AFM image shown in Figure 1d reveals that the gradually bent tubules have a considerably smooth external surface. There are no wrinkles observed, suggesting a uniform bending stress distributed over the entire length of the tubules. It is clear that the circular structure is formed by several bent tubules. The gaps between the ends of the bent tubules and the overlaps of the bent tubules are visible in Figure 1d. When droplets of DC8,9PC tubule solution are placed on hydrophilic Au surfaces with a water contact angle of about 29°, we note that they quickly wet the surfaces. There are no tubule loops observed. On the basis of the observed features of the bent tubules on the hydrophobic DDT SAMs, we suggest that the shrinking contact line of the drying droplet is responsible for the formation of the tubule loops. It has been demonstrated that the convective flow occurring inside the droplets can carry virtually all the dispersed particles to the edge of the droplets.29 The tubules, which are driven by the convective flow to the edge of the droplets, are expected to be bent to follow the droplet circumference by shrinking the contact line to form a loopike shape (Figure 2). Recently, Tsukruk and co-workers30 studied the bending of carbon nanotubes on patterned surfaces. They found that the carbon nanotubes could be bent into opened and closed loops with a curvature radius of 200-300 nm on the amine-terminated SAMs of silanes by the shrinking contact line of the drying droplet. Bensimon et al.31 showed that DNA grafted at both ends on the SAMs with the vinyl terminal groups could be bent into a loopike shape with a curvature radius of 5-10 µm by the moving contact line. The measured radius of the curvature of the tubule loops is in the range from 14 to 25 µm, which is larger than that of the carbon nanotube loops and DNA loops. The curvature radius of the semi-flexible DNA, lipid tubules, and carbon nanotube is related to their stiffness.
Figure 3. (a) Optical microscopy image of a tubule loop with rotating vectors (l). (b) A plot of 〈θ2(l)〉 as a function of vector length (l). The persistence length is determined from the inverse of the slope of the plot. (c) A plot of 〈θ4(l)〉/〈θ2(l)〉2 as a function of vector length (l).
The persistence length (Lp), a parameter related to the stiffness of lipid tubules, is calculated with the technique developed by Frontali et al.,32 who quantify the persistence length of DNA from scanning electron microscopy images of two-dimensional (2D) DNA conformation. The persistence length can be calculated based on the following assumptions: (1) The tubule consists of n rotating vectors of length l joined in succession, where each vector is oriented at an angle θ with respect to the previous vector (Figure 3a). The probability density P(θ(l)) of the angle θ(l) between consecutive vectors is Gaussian and can be represented by
p(θ(l)) )
(
x
)
Lp Lpθ(l)2 exp 2πl 2l
(1)
(2) The observed 2D conformation is obtained by permitted deformation and is not a projection of the 3D structure. (3) The SAM surface does not change the local stiffness of the tubule. The odd moments of the distribution are all zero, whereas, the even moments are
l Lp
(2)
)3
(3)
〈θ2(l)〉 ) and
〈θ4(l)〉 〈θ2(l)〉2
22062 J. Phys. Chem. B, Vol. 110, No. 44, 2006
Zhao et al.
Figure 4. (a) Contact mode AFM image of a straight portion of the tubule loop on the DDT SAM. The zoom-up images of the bilayer edges are inserted. (b) Corresponding cross-sections of region A at different set-point forces.
Where eq 2 is used to calculate the persistence length Lp, while eq 3 is used to check whether the first assumption is valid. Figure 3b is a plot of 〈θ2(l)〉 as a function of l. We note that the intercept of the fit line is not zero. The offset is likely due to the experimental error. The persistence length Lp of the lipid tubule is calculated from the reciprocal slop of the plot to be ∼41 µm, which is much larger than that of a DNA double helix (∼50 nm) measured from AFM images of 2D DNA conformation.33 This suggests that the lipid tubules are much stiffer than DNA. To check whether the distribution is Gaussian, we plot the 〈θ4(l)〉/〈θ2(l)〉2 as a function of l (Figure 3c). As can be seen, there are some deviations at large l values, but for small l values, the 〈θ4(l)〉/〈θ2(l)〉2 approaches 3. The deviation from Gaussian in the analysis of the AFM images of 2D conformations of DNA33 and xanthan34 was also reported. Figure 4a shows a contact mode AFM image of a portion of the straight region near the end of the tubule loop on the DDT SAMs. It is known that the tubule is formed by rolled-up lipid bilayer ribbons. Here, the edges of helical bilayer ribbons forming the tubules are visible in the zoom-up images inserted in Figure 4a. The measured thickness of the helical ribbons is 27.5 ( 1.0 nm, corresponding to a stack of four lipid bilayers, because the thickness of a single DC8,9PC bilayer in the tubule wall has been measured to be about 6.6 nm.4 It is clear that there are changes in the external diameter along the tubule. But, due to the hollow nature of the lipid tubules, the thickness of the tubule wall is difficult to be measured by AFM. The measured height at a set-point force of 2.60 nN is ∼205 nm in
region A, ∼176 nm in region B, and ∼145 nm in region C, respectively. These heights are found to gradually decrease with the increase of the set-point forces. For region A, the initial height of ∼205 nm at a set-point force of 2.60 nN reduces to a height of ∼180 nm at 42.25 nN, a deformation of ∼12%. The deformation is partially recovered as the set-point force is returned to the 2.60 nN (Figure 4b). Here, the tubule deforms through gradual flattening with the increase of the set-point force. The flattening-induced increase in the apparent width is not detected due to the effect of the tip size. For region B, the deformation is ∼20% in the force range from 2.60 to 42.3 nN, which is larger than that of region A. The deformation of region C is ∼24% in the force range from 2.60 to 24.5 nN. If the force is larger than the 24.5 nN, then region C is gradually moved by the multiscans. Regions A and B are stable on the SAMs until the set-point force is larger than 42.25 nN. We plot the strain measured from regions A, B, and C of the tubule as a function of set-point forces (Figure 5a). The strain is defined as ) (H0 - H/H0), where H0 and H are the initial height and the deformed height of a tubule, respectively. The force-strain plots reveal nonlinear dependences of deformation on the setpoint force. As can be seen, region A is more rigid than regions B and C, suggesting that the increase of the number of bilayers in the tubule wall makes the tubule less deformable when subjected to applied load. The stiffness of the tubule in the radial direction is measured from the slope of force-strain plots and then is plotted as a function of the set-point force for regions A, B, and C, respectively (Figure 5b). Under the small set-point
Deformation of Lipid Tubules on Thiol Monolayers
J. Phys. Chem. B, Vol. 110, No. 44, 2006 22063 References and Notes
Figure 5. Plots of set-point forces vs strains obtained from regions A, B, and C of the lipid tubule shown in Figure 4. (b) Plots of radial stiffness vs set-point forces obtained from regions A, B, and C of the lipid tubule.
forces from 2.6 to 13.5 nN, the radial stiffness is the same for regions A, B, and C. As the set-point force is above 13.5 nN, the differences in the radial stiffness are evident between these regions. For example, at the set-point force of 24.5 nN, the radial stiffness is 1.50 N/m for region A, 1.24 N/m for region B, and 0.85 N/m for region C. These measured stiffnesses of the polymerized lipid tubule are comparable with the stiffness of polymeric materials (0.5-1 N/m).35 In conclusion, we report that the self-assembled DC8,9PC lipid tubules can be bent into a loopike shape on the hydrophobic SAMs by the shrinking contact line of droplets of tubule solution. The persistence length of the lipid tubules is estimated to be ∼41 µm. The radial deformation of the lipid tubules on SAMs is studied under applied load using AFM. The stiffness of the tubule in the radial direction increases as the number of lipid bilayers in the tubule wall. Acknowledgment. This work was supported by University of Central Florida’s nano-initiative and the National Science Foundation.
(1) Schnur, J. M. Science 1993, 262, 1669. (2) Shimizu, T.; Masuda, M.; Minamikawa, H. Chem. ReV. 2005, 105, 1401. (3) Thomas, B. N.; Safinya, C. R.; Plano, R. J.; Clark, N. A. Science 1995, 267, 1635. (4) Thomas, B. N.; Corcoran, R. C.; Cotant, C. L.; Lindemann, C. M.; Kirsch, J. E.; Persichini, P. J. Phosphonate Lipid Tubules. J. Am. Chem. Soc. 1998, 120, 12178. (5) Oda, R.; Huc, I.; Schmutz, M.; Candau, S. J.; MacKintosh, F. C. Nature 1999, 399, 566. (6) Song, J.; Cheng, Q.; Kopta, S.; Stevens, R. C. J. Am. Chem. Soc. 2001, 123, 3205. (7) John, G.; Masuda, M.; Okada, Y.; Yase, K.; Shimizu, T. AdV. Mater. 2001, 13, 715. (8) Spector, M. S.; Singh, A.; Messersmith, P. B.; Schnur, J. M. Nano Lett. 2001, 7, 375. (9) Thomas, B. N.; Lindemann, C. M.; Corcoran, R. C.; Cotant, C. L.; Kirsch, J. E.; Persichini, P. J. J. Am. Chem. Soc. 2002, 124, 1227. (10) Singh, A.; Wong, E. M.; Schnur, J. M. Langmuir 2003, 19, 1888. (11) Lee, S. B.; Koepsel, R.; Stolz, D. B.; Warriner, H. E.; Russell, A. J. Am. Chem. Soc. 2004, 126, 13400. (12) Lauf, U.; Fahr, A.; Westesent, K.; Ulrich, A. S. ChemPhysChem 2004, 5, 1246. (13) John, G.; Mason, M.; Ajayan, P. M.; Dordick, J. S. J. Am. Chem. Soc. 2004, 126, 15012. (14) Zhao, Y.; Mahajan, N.; Lu, R.; Fang, J. Y. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 7438. (15) Archibald, D. D.; Mann. S. Nature 1993, 364, 430. (16) Baral, S.; Schoen, P. Chem. Mater. 1993, 5, 145. (17) Seddon, A. M.; Patel, H. M.; Burkett, S. L.; Mann, S. Angew. Chem., Int. Ed. 2002, 41, 2988. (18) Price, R. R.; Dressick, W. J.; Singh, A. J. Am. Chem. Soc. 2003, 125, 11259. (19) Patil, A. J.; Muthusamy, E.; Seddon, A. M.; Mann, S. AdV. Mater. 2003, 15, 1816. (20) Jung, J. H.; Lee, S. H.; Yoo, J. S.; Yoshida, K.; Shimizu, T.; Shinkai, S. Chem.-Eur. J. 2003, 9, 5307. (21) Yang, B.; Kamiya, S.; Shimizu, Y.; Koshizaki, N.; Shimizu, T. Chem. Mater. 2004, 16, 2826. (22) Wilson-Kubalek, E. M.; Brown, R. E.; Celia, H.; Milligan, R. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 8040. (23) Melia, T. J.; Sowa, M. E.; Schutze, L.; Wensel, T. G. J. Struct. Biol. 1999, 128, 119. (24) Schnur, J. M.; Price, R. R.; Rudolph, A. S. J. Controlled Release 1994, 28, 3. (25) Meilander, N. J.; Yu, X.; Ziats, N. P.; Bellamkonda, R. V. J. Controlled Release 2001, 71, 141. (26) Zarif, L. J. Controlled Release 2002, 81, 7. (27) Frusawa, H.; Fukagawa, A.; Ikeda, Y.; Araki, J.; Ito, K.; John, G.; Shimizu, T. Angew. Chem., Int. Ed. 2003, 42, 72. (28) Mahajan, N.; Fang, J. Y. Langmuir 2005, 21, 3153. (29) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Negel, S. R.; Witten, T. A. Nature 1997, 389, 827. (30) Ko, H.; Peleshanko, S.; Tsukruk, V. V. J. Phys. Chem. B 2004, 108, 3485. (31) Bensimon, D.; Simon, A. J.; Croquette, V.; Bensimon, A. Phys. ReV. Lett. 1995, 74, 4754. (32) Frontali, C.; Bore, E. Ferranto, A.; Gratton, E. Biopolymers 1979, 18, 1353. (33) Rivetti, C.; Guthold, M.; Bustamante, C. J. Mol. Biol. 1996, 264, 919. (34) Camesano, T. A.; Wilkinson, K. J. Biomacromolecules 2001, 2, 1184. (35) Dieter, E. G. Mechanical Metallurgy SI Metric Edition; McGrawHill Book Company: London, 1988.
