Directed Growth of Pure Phosphatidylcholine Nanotubes in

10814

Langmuir 2005, 21, 10814-10817

Directed Growth of Pure Phosphatidylcholine Nanotubes in Microfluidic Channels Ksenia P. Brazhnik,† Wyatt N. Vreeland,† J. Brian Hutchison,† Rani Kishore,‡ Jeffrey Wells,‡ Kristian Helmerson,‡ and Laurie E. Locascio*,† Analytical Chemistry Division and Atomic Physics Division, National Institute of Standards and Technology, Gaithersburg, Maryland 20899-8394 Received June 6, 2004. In Final Form: May 19, 2005

The morphology of self-assembled phospholipid membranes (e.g., micelles, vesicles, rods, tubes, etc.) depends on the method of formation, secondary manipulation, temperature, and storage conditions. In this contribution, microfluidic systems are used to create pure phosphatidylcholine (PC) micro- and nanotubes with unprecedented lengths. Tubes up to several centimeters in length and aligned with the long axis of the microchannel were created from spots of dry films of 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC), 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC). These high aspect ratio structures, which, to our knowledge, represent the first examples of extended tubes formed from pure PC lipids, were examined by fluorescence microscopy, electron and optical microscopy, and optical manipulation tools (i.e., a laser trap and laser scalpel) to characterize structure and stability. In particular, the tubular structure was confirmed by observation of fluorescent dyes that were sequestered within the aqueous cavity or within the phospholipid tube. Compared to other phospholipid tubes, the tubes formed from PC lipids in microfluidic channels show high mechanical stability and rigidity that depend on tube size, age, and storage conditions.

Introduction First observations of synthetic lipid nanotubes occurred more than two decades ago, yet their application remains largely unexplored due, in part, to their instability and a lack of tools with which to characterize and manipulate nanoscale objects.1,2 As new tools for nanotechnology have become available, methods for development of lipid nanotubes have gained momentum in the biological and analytical chemistry communities for both fundamental and applied studies.3-8 For example, recent advances made by Orwar and co-workers in the design and manipulation of phospholipid tube networks have demonstrated the applicability of lipid self-assembled systems to fluid and sample manipulation at the submicron level.9,10 Lipid nanotubes have unique advantages for single molecule separations, including biological compatibility and ease of fabrication via manipulation of self-assembled structures. * Corresponding author. E-mail: [email protected]. † Analytical Chemistry Division. ‡ Atomic Physics Division. (1) Papahadjopoulos, D.; Vail, W. J.; Jacobson, K.; Poste, G. Biochim. Biophys. Acta 1975, 394, 483-491. (2) Lasic, D. D. Biochem. J. 1988, 256, 1-11. (3) Frusawa, H.; Fukagawa, A.; Ikeda, Y.; Araki, J. A.; Ito, K.; John, G.; Shimizu, T. Angew. Chem., Int. Ed. 2003, 42, 72-74. (4) Sott, K.; Karlsson, M.; Pihl, J.; Hurtig, J.; Lobovkina, T.; Orwar, O. Langmuir 2003, 19, 3904-3910. (5) Evans, E.; Bowman, H.; Leung, A.; Needham, D.; Tirrell, D. Science 1996, 273, 933-935. (6) John, G.; Masuda, M.; Okada, Y.; Yase, K.; Shimizu, T. Adv. Mater. 2001, 13, 715-718. (7) Karlsson, A.; Karlsson, R.; Karlsson, M.; Cans, A. S.; Stromberg, A.; Ryttsen, F.; Orwar, O. Nature 2001, 409, 150-152. (8) Wilson-Kubalek, E. M.; Brown, R. E.; Celia, H.; Milligan, R. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 8040-8045. (9) Karlsson, M.; Sott, K.; Davidson, M.; Cans, A. S.; Linderholm, P.; Chiu, D.; Orwar, O. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 1157311578. (10) Karlsson, A.; Karlsson, M.; Karlsson, R.; Sott, K.; Lundqvist, A.; Tokarz, M.; Orwar, O. Anal. Chem. 2003, 75, 2529-2537.

10.1021/la047151q

Although spherical phospholipid structures (liposomes or vesicles) are common, elongated, phase-dependent11 structures, such as myelin figures12,13 and tubes,3,6 which vary in lamellarity, surface properties, and size distribution, have been observed to self-assemble11,14,15 from a dried lipid film or from preformed liposomes. Lipid tubes have been prepared previously from various phospholipids and mixtures including sphingolipids,16 phosphatidylcholines,4,17 glycolipids,3,8 and phosphatidylserines1 by techniques that include mechanical manipulation of liposomes with a micropipet,4,17 self-assembly in aqueous dispersions,1,3,6,11 and high-pressure shearing.16 Tubes prepared by each of these methods ranged in length from a few micrometers to approximately 500 micrometers. Interestingly, mechanically drawn tubular structures exist in a thermodynamically metastable state but collapse back into a spherical shape after several hours.4,17 This contribution introduces a new technique for preparing phosphatidylcholine (PC) tubes with nanometer to micrometer diameters and lengths of up to several centimeters. To our knowledge, this is the first reported observation of phospholipid tubes of this length scale. The technique is based on hydration of a lipid film in a microfluidic channel where lipid tube growth is induced by fluid flow under vacuum and guided by a microfluidic channel. The lipid tubes are further manipulated in the microfluidic environment using optical tweezers and an optical scalpel to observe membrane properties. In this (11) Rudolph, A. S. J. Cell. Biochem. 1994, 33, 183-187. (12) Haran, M.; Chowdhury, A.; Manohar, C.; Bellare, J. Colloids Surf., A 2002, 205, 21-30. (13) Mishima, K.; Satoh, K.; Ogihara, T. Chem. Phys. Lett. 1984, 106, 513-516. (14) Monnard, P. A.; Deamer, D. W. Anat. Rec. 2002, 268, 196-207. (15) Lasic, D. D.; Papahadjopoulos, D. Science 1995, 267, 12751276. (16) Kulkarni, V.; Wong, J.; Aust, D.; Wilmott, J.; Hayward, J. J. Cosmet. Sci. 2001, 52, 344-345. (17) Karlsson, M.; Sott, K.; Cans, A. S.; Karlsson, A.; Karlsson, R.; Orwar, O. Langmuir 2001, 17, 6754-6758.

This article not subject to U.S. Copyright. Published 2005 by the American Chemical Society Published on Web 09/17/2005

Growth of PC Nanotubes in Microfluidic Channels

Langmuir, Vol. 21, No. 23, 2005 10815

study, the growth, stability, and rigidity of lipid tubes are characterized. Materials and Methods18 1,2-Dilauroyl-sn-glycero-3-phosphocholine (DLPC), 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), and 1,2-dipalmitoylsn-glycero-3-phosphocholine (DPPC) were obtained from Avanti Polar Lipids (Alabaster, AL). Chloroform (99+%) and sodium azide were obtained from Aldrich (Milwaukee, WI). N-[2Hydroxyethyl]piperazine-N’-[2-ethanesulfonic acid] (HEPES, >99.5%) and bovine serum albumin (98-99%) were obtained from Sigma Chemical Company (St. Louis, MO). The fluorescent membrane dye, 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI(C18)), and sulforhodamine B were obtained from Molecular Probes (Eugene, OR). Calcium chloride (CaCl2‚2H2O) was purchased from Acros Organics (Fairlawn, NJ). Ultrapure deionized 18 MΩ‚cm water was used in all experiments. Poly(dimethylsiloxane) (PDMS) elastomer prepolymer and curing agent were obtained from Dow Corning Corporation (Midland, MI). An aqueous stock buffer solution containing 0.2 M CaCl2, 5 mM HEPES buffer, and 0.15 mM bovine serum albumin was prepared and used in all experiments unless otherwise noted. For experiments requiring encapsulation of a fluorescent dye inside the lipid tubes, sulforhodamine B was added to the stock buffer solution to a final concentration of 1 mM. Solutions of DLPC, DMPC, or DPPC in chloroform were prepared to final concentrations of 160 mM, 150 mM, or 140 mM, respectively. In some experiments, fluorescent membrane dye, DiI(C18), was incorporated into the lipid bilayer at a final concentration of 0.11 mM. In a vacuum desiccator, 0.5 µL of stock lipid solution was dried for several hours to form a film on a glass coverslip between two 1-mm diameter circular access wells. A microfluidic channel was placed above the dried lipid film according to previously described protocols.19,20 Briefly, the trapezoidal microfluidic channel was prepared (30 µm deep, 20 µm wide at top, and 60 µm wide at base) by pouring PDMS liquid over a micromachined silicon wafer template and curing overnight at room temperature. The coverslip with the dried lipid film and access wells was gently pressed on top of the cured PDMS slab, after removal from the template, to create a sealed microfluidic channel. Figure 1a depicts the experimental setup used for forming lipid tubes in a microfluidic channel. A vacuum desiccator was placed in a water bath and its temperature monitored and equilibrated to ambient temperatures of 20-23 °C for DLPC (phase-transition temperature, Tm ) -1 °C), 40 °C for DMPC (Tm ) 24 °C), and 50 °C for DPPC (Tm ) 44 °C). A 15-µl volume of buffer was placed at each access well of the microfluidic channel to provide the aqueous medium for lipid hydration. The assembled microfluidic device was placed into the vacuum desiccator, and the desiccator was depressurized to 26 kPa (3.77 psi) for 15 min. Next, the vacuum was released, forcing the fluid to flow into the evacuated microfluidic channel and over the dried lipid film, thus hydrating the film. This filling technique was originally reported by Monahan et al.21 who developed the method as a means to introduce aqueous solutions into hydrophobic microchannels that were difficult to fill using traditional pumping methods. Phospholipid tubes were removed from the microfluidic channel and dried under vacuum onto a coverslip at ambient temperature (20-23 °C) for 20 h prior to imaging with a field emission scanning electron microscope (1.0 kV; S-4200, Hitachi, (18) Certain commercial equipment, instruments, or materials are identified in this report to specify adequately the experimental procedure. Such identification does not imply recommendation or endorsement by the National Institute of Standards & Technology nor does it imply that the materials or equipment identified are necessarily the best available for the purpose. (19) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974-4984. (20) Martynova, L.; Locascio, L. E.; Gaitan, M.; Kramer, G. W.; Christensen, R. G.; MacCrehan, W. A. Anal. Chem. 1997, 69, 47834789. (21) Monahan, J. G.; A. A.; Nuzzo, R. G. Anal. Chem. 2001, 73, 31933197.

Figure 1. (A) Experimental setup during phospholipid tube formation. A microfluidic device composed of PDMS microfluidic channels (grey) sealed with a glass coverslip (transparent) with holes drilled for fluidic access. The lipid film (black) on a glass coverslip is placed directly over the microfluidic channel during assembly. (B) Fluorescence micrograph of DMPC lipid tubes labeled with DiI(C18) formed using the vacuum method in the microfluidic channels showing the radial outward orientation of the tubes from a bulk lipid film. Bar ) 5 µm. Tokyo, Japan) and a confocal microscope (Axioplan 2 Imaging, Carl Zeiss, Inc., Thornwood, NY). Scanning electron microscopy (SEM) images were used to size and characterize the phospholipid tubes, and confocal microscopy was used to measure the outer diameters and confirm an internal aqueous cavity of DMPC tubes. Phospholipid tubes were stored inside the microfluidic channels at 4 °C and at ambient temperature (20-23 °C) for three months. During the first two weeks of storage, some samples were subjected to 20 min of UV exposure at 48 h intervals (254-nm hand-held UV shortwave light source, Ultraviolet Products, San Gabriel, CA) or to heat cycling at 48 h intervals. Heat treatments consisted of heating the entire microfluidic device with the phospholipid tube sample to approximately 10 °C above the phasetransition temperature of the lipid and allowing the sample to cool to room temperature for two cycles. Alternatively, in some samples, 1.5 or 0.15 mM sodium azide was added to the buffer solution after tube formation. The stability of phospholipid micro- and nanotubes was evaluated by recording the length of time that a water-filled, tubular structure was retained. The rigidity of phospholipid tubes was assessed using optical tweezers and an optical scalpel (one system was built in-house, and another was a Bioryx 200 system from Arryx, Chicago, IL). The procedure for optical manipulation of tubes was described previously by Kulin et al.22 Optical tweezers were used to pull on the lipid membrane. Several phospholipid tubes in each sample were cut with an optical scalpel once a week over a 2 month storage period.

Results Lipid micro- and nanotubes were formed inside a microfluidic channel under vacuum-generated pressuredriven flow to facilitate and direct tube growth. Figure 1 depicts the microchannel apparatus and an image of the microtubes extending from a lipid film. The lipid film was wider than the microchannel, which created a broken seal between the glass coverslip and the PDMS directly adjacent to the lipid film. Some lipid tubes formed outside of the microchannel boundaries. The tubes depicted in Figure 1b appeared to start as uniformly sized projections that grew radially from the bulk lipid. The structures contained an aqueous cavity with a lipid reservoir bulb at one end and were tethered to the lipid film at the other end. Flow in the microfluidic channel oriented the tubes and caused them to elongate axially in the channel. Figure 2 shows that phospholipid tubes are 200 nm (or less) to several micrometers in diameter, with lengths of several hundred micrometers to several centimeters as viewed by bright-field microscopy. Further examination of dried (22) Kulin, S. K. R.; Helmerson, K.; Locascio, L. Langmuir 2003, 19, 8206-8210.

10816

Langmuir, Vol. 21, No. 23, 2005

Brazhnik et al.

Table 1. Time to Rigidity for Stored PC Tubes with Different Treatments treatment

a

storage temperature

no additional treatment

20 min UV exposure every 48 h for 2 weeks after formation

0.15 or 1.5 mM NaN3 added to buffer after formation

heat cycling every 48 h for 2 weeks after formation

4 °C 20-23 °C

1 week (5)a 5 weeks (8)a

2 weeks (2)a 5 weeks (2)a

2 weeks (6)a 5 weeks (4)a

2 weeks (2)a 12 weeks (4)a

Number of samples is in parentheses.

Figure 2. (A) Fluorescence micrograph of DLPC lipid tubes labeled with membrane dye DiI(C18). Bar ) 20 µm. (B) DMPC lipid tubes labeled with membrane dye DiI(C18). Bar ) 5 µm. The tubes align with the direction of flow.

lipid tubes with SEM indicated that the outer diameter of the tubes was 100 nm to 10 µm. Lipid tubes did not form by simple hydration without the application of vacuumassisted flow even when all other conditions remained the same, which reveals the importance of the confined geometry and flow for creating PC tubes via this method. Tubes were also prepared with encapsulated sulforhodamine B dye dissolved in the buffer to facilitate imaging of the aqueous entrapped volume. To verify encapsulation and retention of fluorescent dye, phospholipid tubes containing the encapsulated dye were cut with the optical scalpel. When the tube was cut, the fluorescent dye leaked from the severed tube ends into the extravesicular space (data not shown). Tubes formed from pure DLPC, DMPC, or DPPC, with and without encapsulated or membrane dye, exhibited similar diameters and lengths. The time-dependent rigidity (i.e., stiffness and resistance to collapsing) of the lipid tubes was probed. Specifically, optical tweezers were used to manipulate and an optical scalpel was used to cut phospholipid tubes contained inside microfluidic channels immediately after formation and at 1 week intervals. DMPC and DPPC tubes manipulated within 48 h of preparation were easily deformed and stretched with optical tweezers. When cut with the optical scapel, a single tube was split into two tubes that retracted from one another with a concurrent rearrangement of the lipid membranes to form what appeared to be closed ends (resealed). The time-dependent rigidity in DLPC tubes was not investigated. Under some storage conditions, the membrane properties of microtubes and nanotubes differed. Figure 3a shows a cut microtube collapsing and receding in tube form away from the laser scalpel spot (visible by the bubble in the third frame) and toward the bulk lipid (reservoir). The two ends of the cut tube resealed immediately after cutting with the optical scalpel. In contrast, Figure 3b shows a lipid tube with a submicrometer diameter that did not collapse and reseal but remained open-ended when cut with the optical scalpel, and small vesicles diffused out of the cut tube. In most cases, two distinct observations were made. First, the tubes

Figure 3. (A) Sequential bright-field images of a fresh lipid tube during cutting with a laser scalpel. A bubble, visible in the third frame of the sequence, reveals the location of the laser pulse. The fourth frame shows that the cut tube splits and retracts to form two separate sealed tubes (marked by arrows). (B) Sequential images of an aged nanotube during cutting. These tubes remain open after cutting, with the second and third frames showing small round vesicles (marked by arrow) diffusing out of the tube into the surrounding solution. Bar ) 5 µm.

became rigid (i.e., resistant to stretching or deformation by the optical tweezers) before they stopped resealing when cut with the optical scapel. Second, smaller (i.e., submicrometer) tubes became rigid sooner than larger tubes (>1 µm) under the same storage conditions. Microchannel devices containing tubes were subjected to various treatments to further investigate the effect of storage conditions on tube rigidity. Table 1 shows the time (in weeks) when PC tubes became rigid, as measured by stretching with the optical tweezers, under different storage conditions. No difference in the onset of rigidity among the three lipids was observed. In general, (1) PC tubes retained their cylindrical structure throughout the 12 week study, (2) tubes stored at 4 °C became rigid faster than those stored at 23 °C, and (3) the thermal cycling during the first 2 weeks of storage caused a substantial increase in the time to rigidity for the samples stored at ambient temperature but not for those stored at lower temperatures. The surface appearance of the stored tube samples were investigated with SEM. Figure 4a shows that phospholipid tubes stored at 23 °C retained a smooth outer lamellae and uniform diameter throughout storage. In contrast, Figure 4b reveals that these tubes stored at 4 °C remained smooth, but became nonuniform in diameter. Finally, Figure 4c represents a unique case, in which chemical

Growth of PC Nanotubes in Microfluidic Channels

Langmuir, Vol. 21, No. 23, 2005 10817

Figure 4. Scanning electron micrographs of phospholipid tubes stored for 12 weeks. (A) Tubes stored at 23 °C remained smooth and uniform in diameter. (B) Tubes stored at 4 °C exhibited budding and pearling. The irregularities generally noted in these tubes are indicated with arrows. (C) Chemical treatment of phospholipid tubes with sodium azide appears to degrade the outer lipid membrane of the tubes. Bar ) 5 µm.

treatment with sodium azide led to a striated appearance of the PC tubes. Conclusions A method for preparing self-assembled phosphatidylcholine tubes of micrometer and nanometer diameters inside a microfluidic device is reported. A microfluidic channel was utilized to direct and align the growth of the lipid tubes. Using this approach, we prepared tubes from pure phosphatidylcholine lipids. Tubes prepared in this manner were up to several centimeters in length, which

is, in some cases, 40 times longer than those that have been previously reported. Interestingly, the tubes became rigid after time periods that depended on different treatments during storage. Acknowledgment. The authors gratefully acknowledge Dr. Christopher Montgomery for assistance with SEM imaging and Dr. Michael Gaitan for assistance with confocal imaging and analysis. LA047151Q