We describe a protocol for modifying cell affinity of a scaffold surface in aqueous environment. The method takes advantage of titanium dioxide photocatalysis to decompose organic film in the photo-irradiated region. We show that it can be used to create microdomains of scaffolding proteins, both ex situ and in situ.
Organic contaminants adsorbed on the surface of titanium dioxide (TiO2) can be decomposed by photocatalysis under ultraviolet (UV) light. Here we describe a novel protocol employing the TiO2 photocatalysis to locally alter cell affinity of the substrate surface. For this experiment, a thin TiO2 film was sputter-coated on a glass coverslip, and the TiO2 surface was subsequently modified with an organosilane monolayer derived from octadecyltrichlorosilane (OTS), which inhibits cell adhesion. The sample was immersed in a cell culture medium, and focused UV light was irradiated to an octagonal region. When a neuronal cell line PC12 cells were plated on the sample, cells adhered only on the UV-irradiated area. We further show that this surface modification can also be performed in situ, i.e., even when cells are growing on the substrate. Proper modification of the surface required an extracellular matrix protein collagen to be present in the medium at the time of UV irradiation. The technique presented here can potentially be employed in patterning multiple cell types for constructing coculture systems or to arbitrarily manipulate cells under culture.
Semiconductor lithography processes and its derivatives — such as photolithography1,2, electron-beam lithography3-6, and microcontact printing7-10 — have now become an established tool in cell biology to grow living cells in a defined position and geometry. The patterning method relies on the use of microfabricated substrates, consisting of micro-island of cell permissive coating in a non-permissive background. Such substrate serves as a template to pattern the cells. These technologies have provided us the novel methods to engineer cells and their function at a single- and multi-cellular level, to extract the intrinsic properties of cells, and to increase the throughput of cell-based drug screening11.
The degree-of-freedom in cell patterning would greatly increase if the template pattern geometry could be altered in situ, i.e., while cells are cultured on the surface. The conventional methods for pattern formation cannot be directly applied here, since they process samples in atmosphere or in vacuum. Therefore various new surface modification techniques have been proposed, which are based, e.g., on photoreactive compounds12,13 or laser ablation5,14, just to name a few. The proposed methods have been nicely reviewed by Robertus et al.15, and more recently by Choi et al.16 and by Nakanishi17.
Here in this article, we describe a novel protocol of in-situ surface modification, which takes advantage of photocatalytic decomposition of organic molecules on a titanium dioxide (TiO2) surface18,19. In this method, a TiO2 film is inserted between the glass substrate and the organic film that interfaces the cells, and the organic film is decomposed in situ by locally irradiating ultraviolet (UV) light to a region of interest (λ < 388 nm). We show that the new protocol can be used to create micropatterns of extracellular matrix proteins and living cells both ex situ and in situ. TiO2 is biocompatible, chemically stable, and optically transparent, features of which makes it friendly to introduce in cell-culture experiments. This protocol provides a materials science-based alternative for modifying cell-culture scaffolds in cell-culture environment.
1. Preparation of TiO2-coated Glass Coverslip
2. Surface Coating with Cell-repellent Film
3. Ex-Situ Surface Patterning
4. Cell Culture
5. In-Situ Surface Patterning
Figure 1. Schematic illustration of the overall process. See text for details. Please click here to view a larger version of this figure.
Figure 2A shows a cross-sectional scanning electron microscopy (SEM) image of the sputter-deposited TiO2 film. From the observation, thickness of the film was estimated to be approximately 150 nm. Noticeable here is the flatness of the deposited TiO2 film. Further analysis by atomic force microscopy (AFM) revealed that the root-mean-square (rms) roughness of the surface was 0.2 nm (Figure 2B).
When the TiO2 surface is modified with OTS and then immersed in a serum-containing growth medium, serum albumins adsorb onto the OTS surface and inhibit subsequent protein adsorption and cell adhesion18,20. Subsequently, when focused UV light is irradiated to the surface, reactive oxygen species (ROS) are generated at the TiO2 surface21, which decompose the OTS self-assembled monolayer (SAM) on TiO2 and remove the SAM and the adsorbed albumins from the surface. This renders the irradiated region hydrophilic and permissive to protein adsorption. Direct interaction of the ROS with albumins might be occurring simultaneously, but the decomposition of OTS should be the main reaction, since OTS SAM is interfacing albumin and TiO2. In the current protocol, an extracellular matrix protein Col-IV is supplemented in the medium during irradiation, and Col-IV adsorbs preferentially to the UV-irradiated region where albumins are removed (Figure 3A). Other proteins such as fibronectin can be patterned in the same protocol (Figure 3B), indicating the versatility of this approach.
Serum albumins, which are also present in the medium at the time of UV irradiation, should also adsorb onto the irradiated region, but the results suggest that the amount is not high as that on the intact OTS SAM. To confirm this, we investigated the difference in the affinity of BSA to UV-irradiated and intact OTS SAM by fluorescence microscopy of dye-labelled BSA. For this experiment, focused UV light was irradiated to the OTS/TiO2 sample that has not been immersed in the serum-containing medium, and then the sample was immersed in a solution of fluorescein- labeled BSA. Figure 3C shows the adsorption pattern of BSA. Brighter fluorescence in the intact OTS SAM region than the UV-irradiated area indicates that the amount of adsorbed BSA is higher in the intact region. The mechanism of preferential adsorption can be understood by the difference in hydrophobicity of the two regions. The amount of BSA adsorption is known to be higher on hydrophobic surfaces22, and the intact OTS SAM region is more hydrophobic than the UV-irradiated region18.
The pattern of Col-IV thus created serves as a scaffold for cell adhesion. Figure 4A shows NGF-differentiated PC12 cells cultured on an ex-situ patterned region for 1 day. Next, UV light was irradiated to the side of a micropattern, and the cells were cultured for additional 5 days. As shown in Figure 4B-D, cells further migrated to the modified region, indicating that cell affinity of the surface was successfully modified in cell-culture environment.
Figure 2. Characterization of the TiO2 film. (A) Cross-sectional SEM image of the sample. (B) AFM image of the TiO2 surface. Please click here to view a larger version of this figure.
Figure 3. Site-selective adsorption of proteins onto the UV-irradiated region. (A) Fluorescein-labelled collagen (200 μm octagonal pattern) and (B) rhodamine-labelled fibronectin (100 μm square pattern). Preferential adsorption of these proteins relies on the absence of albumins that block protein adsorption (C). Note that in (C), OTS SAM was not immersed in serum-containing medium prior to application of fluorescein-labelled BSA. Scale bars, 100 μm (A) and 50 μm (B, C). Please click here to view a larger version of this figure.
Figure 4. Ex-situ and in-situ patterning of PC12 cells. (A) Adhesion of PC12 cells onto an ex-situ patterned region. Dotted octagon indicates the site of in-situ UV irradiation. (B–D) Migration of PC12 cells to the in-situ patterned region, at 2 days (B), 3 days (C), and 5 days (D) after irradiation. Scale bar, 200 μm. Modified with permission from Yamamoto et al.18. Please click here to view a larger version of this figure.
In our current protocol, TiO2 film was formed by RF-magnetron sputtering. We favor this method of deposition since it allows us to reproducibly prepare a photocatalytic TiO2 film with a sub-nm roughness. Although sputter deposition processes are familiar to materials scientists and electronic engineers, it may not be quite accessible to biologists. In that case, spin-coated TiO2 film would be an alternative choice23. In this method, TiO2 nanoparticles dissolved in a solvent is spread out on a substrate surface by centrifugal force. Special equipment required is a spin coater, which is much affordable than a sputtering system. We have used for ourselves a spin-coated TiO2 nanoparticle film and have confirmed that the film can also be used to carry out similar experiments (data not shown). The surface of the spin-coated TiO2 film is comparatively rough with an rms roughness of several tens of nm, and it takes some practice to uniformly coat the substrate reproducibly especially when multiple spin coatings are needed to obtain thick films. Nevertheless, the method might be the first candidate for biologists.
Thus far, we have utilized this method in patterning not only the PC12 cells but also a human pancreatic cancer cell line AsPC-1 and primary neurons obtained from embryonic rat hippocampi19. Adhesibility of AsPC-1 cells are much stronger than that of PC12 cells or hippocampal neurons, and supplementation of scaffolding molecules, such as collagen, was not required in patterning them. We yet note that non-specific adhesion of the cells to the unirradiated region was occasionally observed, which is probably due to defects in the OTS SAM.
Patterning primary neurons were much more challenging since their adhesibility is weak. In fact, the critical step of the current process lies in the proper adsorption of scaffolding molecules. A simple supplementation of collagen or laminin was insufficient to increase affinity of the modified region for the neurons. Conventional culture of hippocampal neurons relies on positively-charged peptide poly-lysine to coat the surface of the coverslip24. In this context, inducing selective attachment of poly-lysine to the UV-irradiated area is desired, but was impossible since poly-lysine adsorbed on the intact OTS SAM region as well. We tried other surface coatings, such as polyethylene glycol-silane SAMs and perfluoro-silane SAMs, to find that none of these were anti-fouling to poly-lysine. To solve this issue, we very recently showed that a cross-linking molecule could be used to induce active adsorption of the scaffolding molecules onto the UV-irradiated region19. Such selection of scaffolding molecules and optimization of its concentration should be carried out for each experiment.
UV irradiation dose should also be optimized since it strongly affects the amount of protein adsorption. In our case, fluorescence intensity measurement was used to derive an optimal UV dose of 200 J cm-2 18. The value is likely to vary depending on experimental parameters such as the thickness and the crystallinity of the TiO2 film.
Among various techniques of in-situ surface modification developed so far, the major advantage of the present method is that it does not require organic synthesis of specialized molecules for the surface processing. The method presented here is, in principle, compatible with organic films prepared from the vast library of commercially-available polymers, peptides, and SAM precursors. It is also favorable in a sense that the surface-modification process can be carried out using conventional lab equipment, such as a wide-field fluorescence microscope equipped with a mercury arc lamp, as was shown here. Using an UV laser as the light source, subcellular resolution down to ~3 μm can be achieved (data not shown).
Some limitations of the current method include the irreversibility of the surface modification, need for supplementing scaffolding molecules in the medium, and generation of ROS. The process described here is applicable to non-permissive to permissive conversion of the surface, but the conversion is one-way and cannot be reversed. For experiments that require reversal conversion of cell affinity, other methods15 need to be considered. In addition, our process requires scaffolding molecules such as collagen to be bath-applied to the cell culture medium. Since the population of pre-patterned cells would inevitably be exposed to the molecules, this may limit some application of this process. Photocatalyzed generation of ROS, such as hydroxyl radicals, at the UV-irradiated region may be toxic to nearby cells, but, at least, this effect was not observed in our experiments. Even if this becomes prominent for other applications, this problem should be solved by supplementing oxygen scavengers in growth medium. Since the oxygen scavenging molecules would not be located between the TiO2 and the non-permissive molecules on its surface, they are expected to eliminate excessive ROS without affecting the patterning process kinetics.
The authors have nothing to disclose.
Authors thank Mr. Kotaro Okubo for the kind assistance with SEM imaging. This work was supported by the Japan Society for the Promotion of Science Grant-in-Aid for Basic Research (B) (20310069), Grant-in-Aid for Research Activity Start-up (25880021), and by research grants from the Kurata Memorial Hitachi Science and Technology Foundation and the Nippon Sheet Glass Foundation for Materials Science and Engineering.
Glass coverslip | Warner Instruments | CS-15R15 | 15 mm, #1.5 thickness |
Diamond scriber | Ogura Jewel Industry | D-Point Pen | |
RF sputtering system | ANELVA | SPC350 | |
TiO2 sputtering target | Kojundo Chemical Lab | Titanium (IV) oxide, target | Purity, 99.9% |
Plasma reactor | Yamato | PR301 | |
n-octadecyltrichlorosilane (OTS) |
Aldrich | 104817 | |
Toluene | Wako | 204-01866 | |
Tissue-culture dish (35 mm) | Greiner | 627160 | |
Tissue-culture dish (60 mm) | BD Falcon | 353002 | |
Type-IV collagen | Nitta Gelatin | Cellmatrix Type IV | |
D-PBS | Gibco | 14190-144 | |
Dulbecco's modified Eagle's medium (DMEM) | Gibco | 11885-084 | |
Fetal bovine serum | Gibco | 12483-020 | Heat-inactivate and pass through a 0.22 mm filter before use |
Horse serum | Gibco | 26050-088 | Pass through a 0.22 mm filter before use |
Penicillin-streptomycin (100x) | Nacalai tesque | 26253-84 | |
7S nerve growth factor (NGF) | Alomone Labs | N-130 | |
Bovine serum albumin (BSA) | Sigma | A2153 | |
EDTA | Dojindo | N001 | Stock solution in 0.5 M |
TiO2 nanoparticle | Tayca | TKD-701 |