You are on page 1of 15

Biomedical Microdevices 6:4, 325–339, 2004 C 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.

NanoLiterBioReactor: Long-Term Mammalian Cell Culture at Nanofabricated Scale
Ales Prokop,1, 2,∗ Zdenka Prokop,1 David Schaffer,3 Eugene Kozlov,2 John Wikswo,4, 5, 6 David Cliffel,7 and Franz Baudenbacher4
1 NanoDelivery, 2 Chemical

Inc., Nashville, TN 37211 Engineering, Vanderbilt University, Nashville, TN 37235 3 Mechanical Engineering 4 Biomedical Engineering 5 Physics & Astronomy 6 Molecular Physiology & Biophysics 7 Chemistry, Nashville, TN 37235, USA E-mail: ales.prokop@mcmail.vanderbilt.edu

Abstract. There is a need for microminiaturized cell-culture environments, i.e. NanoLiter BioReactors (NBRs), for growing and maintaining populations of up to several hundred cultured mammalian cells in volumes three orders of magnitude smaller than those contained in standard multi-well screening plates. These devices would enable the development of a new class of miniature, automated cell-based bioanalysis arrays for monitoring the immediate environment of multiple cell lines and assessing the effects of drug or toxin exposure. We fabricated NBR prototypes, each of which incorporates a culture chamber, inlet and outlet ports, and connecting microfluidic conduits. The fluidic components were molded in polydimethylsiloxane (PDMS) using soft-lithography techniques, and sealed via plasma activation against a glass slide, which served as the primary culture substrate in the NBR. The input and outlet ports were punched into the PDMS block, and enabled the supply and withdrawal of culture medium into/from the culture chamber (10–100 nL volume), as well as cell seeding. Because of the intrinsically high oxygen permeability of the PDMS material, no additional CO2 /air supply was necessary. The developmental process for the NBR typically employed several iterations of the following steps: Conceptual design, mask generation, photolithography, soft lithography, and proof-of-concept culture assay. We have arrived at several intermediate designs. One is termed “circular NBR with a central post (CP-NBR),” another, “perfusion (grid) NBR (PG-NBR),” and a third version, “multitrap (cage) NBR (MT-NBR),” the last two providing total cell retention. Three cells lines were tested in detail: a fibroblast cell line, CHO cells, and hepatocytes. Prior to the culturing trials, extensive biocompatibility tests were performed on all materials to be employed in the NBR design. To delineate the effect of cell seeding density on cell viability and survival, we conducted separate plating experiments using standard culture protocols in well-plate dishes. In both experiments, PicoGreen assays were used to evaluate the extent of cell growth achieved in 1–5 days following the seeding. Low seeding densities resulted in the absence of cell proliferation for some cell lines because of the deficiency of cell-cell and extracellular matrix (ECM)-cell contacts. High viabilities were achieved in all designs. We conclude that an instrumented microfluidics-based NanoBioReactor (NBR) will represent a dramatic departure from the standard culture environment. The employment of NBRs for

mammalian cell culture opens a new paradigm of cell biology, so far largely neglected in the literature. Key Words. nanobioreactor, long-term, mammalian culture

Background and Significance
Scaling laws, non-specific screen, and scope of work Today’s pharmaceutical industry is faced with the unprecedented challenge of managing the progress of a rapidly expanding pool of molecular targets, novel compounds, and biological assays, all of which are needed to discover and develop new drugs. High-throughput screening (HTS) in a high-density format may provide some relief. In addition, there is a clear move within the pharmaceutical industry towards increased emphasis on the cellbased assay, which requires parallel processing of multiple, smaller batches of a wide range of cell lines derived from various tissue origins. The advent of cell robots that incorporate incubators, laminar air flow, automatic seeding, feeding, trypsinizing, harvesting and counting cells, followed by the dispensing of these cells into a measurement format suitable for subsequent analysis, is becoming reality (Slater, 2001). High content screening that facilitates multiple analyses on a solitary sample is another direction that has yet to be developed in detail. New sensor concepts with increased chemical and physical sensitivity facilitates miniaturization, leading to paramount improvements in signal/background ratio, reproducibility, simplicity, and cost. While individual living cells displace picoliter volumes, the flasks, dishes, and wells in which cells are cultured in vitro possess volumes ranging from 100 µL
∗Corresponding

author.

325

such as DNA replication and messenger RNA synthesis (102 –104 s). millisecond. potential damage to shear-sensitive organisms such as mammalian cells can occur. Micro Electro Mechanical Systems (MEMS). there are a number of biological phenomena that occur quite rapidly (10−3 –10−5 seconds) at the cellular dimensions of 10 µm.326 Prokop et al. Physical hierarchy at scale-down for a stagnant spherea Relaxation time Subsystem 2000 µm 200 µm 5 ms 70 ms Oxygen quiescent (free) diffusion in/from the 0. including physical and biological phenomena. gradients at the interface will facilitate higher transfer rates and shorter relaxation times. DNA replication Response to environmental changes (temperature. no heat source is considered within the sphere) have a strong dependence upon sphere size. of which are involved in the direct response of the cell to external physical or chemical stimuli. It is for these processes that the scaling down in size of the physical environment and its concomitant variables results in an analogous biological process. It is well known that in the scaling-up of a fluidic system. In Table 2 we list results of our calculations for scaledependence of chemical and thermal diffusion time constants for two reactor configurations: a spherical reaction phase (droplet) with 2-mm diameter. The real-time dynamics of interest to this scale-down process include the faster relaxation times of enzymatic systems.. liquid circulation time. Because of this. where R is the sphere radius. E is a fractional oxygen depletion and D is oxygen diffusivity in . to 100 mL—between seven and eleven orders of magnitude greater than the cellular volume! The scaling laws that govern the microminiaturization of silicon-based electronic devices. and influence on. scaling up a bioreactor process can yield adverse results. impeller tip speed. is presented in Table 1 in terms of relaxation time or time constant. mixing and liquid homogenization times are necessarily increased to avoid such damage. such as diffusion and inertial influence. There has been substantial work on scaling bioreactors to larger volumes— commercial bioreactors have volumes from 104 to 105 liters. respectively). of a number of associated physical and biological parameters and phenomena. However. Dynamic hierarchy of physical (reactor) and biological systems (Prokop. In particular. all Table 1. or lack thereof. 1995) System (subsystem) Mixing time to homogenize liquid in a large-scale bioreactor (10–100 m3 ) Time to exchange liquid volume to 90% level (depending on growth rate) in a continuous reactor Oxygen transfer (forced. such parameters as the level of turbulent liquid motion. An example can be found in an aerated bioreactor: if the power input of its mixer remains constant. 1982. its relationship with. microfluidics. and thereby offers the greatest potential. the time constants for oxygen diffusion from these spheres (to achieve about 50% depletion or saturation) and heat transfer time constants (to heat up these spheres from ambient to 50% of the targeted 37◦ C temperature. Therefore. for the oxygen transfer time constant t = π R2 E/D. other parameters can become considerably distorted during scale-up. and nanoparticles have been studied exhaustively. 1931) Heat transfer by convection into/out of sphere 7s (Crosby. 1983). and receptor-ligand interactions. not free diffusion) Heat transfer (forced convection) Oxygen uptake rate (mammalian cells) Cell proliferation. inhibitor Receptor-ligand interaction Relaxation time (s) 104 –108 105 –106 102 –103 103 –104 104 –105 102 –104 103 –104 103 –104 101 –103 102 –101 1 10−1 –101 10−2 10−3 –10−5 10−6 –10−3 10−6 10−6 Table 2. and one with 200 µm diameter (possessing volumes of 24 µl and 24 nL. oxygen) Messenger RNA synthesis Translocation of substances into cells (active transport) Protein synthesis Allosteric control of enzyme action Glycolysis Oxidative phosphorylation in mitochondria Intracellular quiescent mass and heat transfer (dimension 10−5 m) Enzymatic reaction and turnover Bonding between enzyme and substrate. This behavior is reflected in the squared dependency on droplet size in both cases (e. in this case. many physical similarities break down because certain properties. 1961) a In practice. may become relatively distorted (Oldshue. become dominant. Those biological processes with comparatively long time constants. but are much slower at the spatial scale of a typical cell culture environment—for example. with one system having 106 liters—but not on the scaling in the other direction. and even microsecond intervals. enzyme-substrateinhibitor interactions. quiescent mass and heat transfer within the interior of a living cell. When one physical parameter is fixed. and heat transfer. The dynamic hierarchy of biological systems and subsystems. As seen from this table. but little is known about how microminiaturization of cell-culture processes to nanoliter volumes might affect the chemical and physical parameters required for maintaining cells in vitro.5 s liquid phase within a sphere (Newman.g. will not benefit from scale-down. The scale-down of a bioreactor for individual mammalian cells would benefit from the simultaneous scaling. many cellular processes occur in sub-second.

Cooper. 1990. their ability to facilitate continuous growth of selected cell lines. Table 3 compares the time constants of electrochemical sensors that we plan to use to monitor the environmental state of the cells. minimize required volumes of expensive analytical pharmaceuticals or toxins. Equipped with such sensors. 2001). but would also simplify accurate cell counting. and pH) have very short time constants. We then report on the testing of various NBR designs. beyond that of optimizing the NBR physically. (4) parallel operation of many assays realized through integration of multiple reactors on a single chip. Reaction time constants for chemical reacting systems (and cellular ones) at the nanoliter scale have been recently discussed (Bratten et al. or to apply drugs or toxins followed by the adaptive administration of a selective toxin antidote. We tested several geometric configurations (shapes). to control temperature.. (2) minimal volume requirement for analyte. Reduced NBR volumes would not only shorten the time required for diffusive mixing. the reduction of the linear scale by a factor of ten (from 2000 to 200 µm) results in a hundred-fold decrease in both relaxation times. which served as the substrate. eventually. and. and provides the ultimate rationale for microminiaturization of the cell culture system: By matching the chemical and thermal time constants of a nanoliter cell culture environment to those of the cells contained there within. 1997. hence they should adequately provide insight into the very dynamics of a variety of cellular responses. 1999). to maintain homeostasis. parallel. there was no need to selectively modify the substrate to confer a hydrophilic character to its surface. (3) reliable and reproducible operation achieved by automated sample handling. As evident from data below. The microfluidic elements were sealed to conventional glass microscope slides.. and a range of cell confinement configurations within the reactor volume.1 Instrument/probe DO probe pH probe ORP probe Size (µm) 5 (planar) 5 (planar) 5 (planar) water. redox potential—ORP.. Time constants of instruments used in this work Response time (90% change) (s) 0. (5) low-cost devices resulting from fabrication of all components onboard a single. In future versions of the NBR we plan to incorporate on-chip pumps for low flow perfusion.g. ionic concentration. The challenge. conventional cell culture chambers.1 0. response time of sensor) are reduced substantially (Manz et al. An added bonus is the possibility of studying cell populations with low cell counts whose constituents are completely detached from typical tissue environment. . although selective patterning may be necessary to confine cell growth to particular regions within the NBR. Thus it is feasible in this system to assess quickly the physiological status of a culture as soon as steady-state conditions are achieved. The advantages of miniaturized detection systems are numerous. We first describe the microfabrication techniques employed to create glass/PDMS NBRs with different geometries and area/volume ratios (A/V). or even an isothermal picocalorimeter to monitor thermodynamic response. etc.1 0. The biosensor elements of the NBR might include planar pH. for achieving thermal equilibrium. In terms of continuous-flow fluid mechanics. for the heat transfer the relationship is similar). and include: (1) reduced assay response times. Consequently. is to detect cellular response. we need only about three volume changes of the NBR content in order to achieve steady-state conditions from an engineering viewpoint (an analogy to continuous stirred tank reactor concept in reaction engineering or to a continuous culture in microbial physiology).and long-term cultivation of several mammalian cell lines in a perfused system. and for cells to grow to confluence. This approach will enable automated. and electrochemical sensing of pH. and redox potential sensors. glucose and lactate to monitor cell metabolism. For this reason. The ultimate goal is to incorporate biosensor elements into the NBR to monitor cell physiology and cell culture conditions in situ and to characterize in a nonspecific manner the metabolic activity of cells. dissolved oxygen. Experimental Design and Methods Fabrication of NBR and integration into a single functional device We used soft lithography in PDMS to fabricate NBRs with various area/volume ratios. and allow for thousands of culture chambers on a single instrumented chip. pH. and multiphasic monitoring of multiple cell lines for drug and toxicology screening. the behavior of such a scaled-down system is determined by diffusion while other times (transport time.. It is possible to totally replace the volume very rapidly when dealing with volumes as small as those contained by the NBR. it will be possible to use monitoring of external variables to determine changes in intracellular processes that heretofore have only been measurable with the much longer time constants of much larger. the NBR could be used to perform short. disposable platform. It is apparent that the electrochemical and fluorescence probes (dissolved oxygen—DO. and on cell viability measurements. or populations in controlled physical and chemical gradients.NanoLiterBioReactor 327 Table 3. facilitate closed-loop adjustments of the environment—e. provide appropriate control signals. Becker and Gartner. and to monitor their response to analytes in a massively parallel format.

Racine. Alternatively. The diameter of the Circular NBR chamber was 825 µm and is shown in Figure 1. Both PDMS device and glass substrate were placed into a plasma sterilizer (Harrick Scientific Corp. Automated syringes pumps (WPI. and connecting microfluidics channels could be adapted according to demand. inlet and outlet ports. PA) served as the main cell substrate at the bottom of the NBR. or master. Midland. (North Andover. punched with plumbing holes. The designs were then either sent to Advance Reproductions Corp. All NBRs were fabricated with soft lithography techniques (Whitesides et al. Instead of oxygen plasma. and bonded to their substrates with plasma activation. Heat-shrinkable tubing (Advanced Polymers Inc. (2) Master production: a thin film of negative-tone photoepoxy was established on a silicon substrate. and assumed a volume of approximately 20 nL (at a depth of 45 µm). For supply and withdrawal of culture medium. They were then promptly removed and placed in mutual contact. simultaneous analyses in genomics. MI) against the SU-8 masters. was thereby established in the shape of the NBR. and may offer highly efficient.. NH) secured all plumbing connections to ensure waterproof sealing using Microtorch MT-10 (Master Appliance Corp. The mask image was reduced 100× onto high contrast slide film using a 35 mm camera. Some access ports were siliconized with SigmaCoat (Sigma Chemical Company.. and a strong. A negative relief. Fluidic elements were cast in Sylgard 184 PDMS (Dow Corning Corp. we employed simple air plasma and achieved very strong seals in the hybrid glass/PDMS device (McDonald and Whitesides. Air-plasma treatment was used to seal the PDMS blocks to their glass substrate. Once developed. the culture chamber.. The cured PDMS was peeled from the master. and metabolomics (Lee and Lee. The Perfusion Grid/Sieve NBR (PG-NBR) enclosed a similar volume. Regardless of its origin. Louis. without the seeding channel. Salem. while physicochemical patterning incorporates chemical adhesion. input and outlet ports were punched into the PDMS block with a modified 16-gauge syringe needle. Finally. FL) supplied controlled flow to the microscale devices. The Multitrap NBR (MTNBR. NY) and exposed to air plasma for 20 s. 2002). Schematic of the first generation CP-NBR with the central post. custom chrome masks were ordered per CAD design. a printout of the design was reduced in size onto highcontrast photofilm via a 35 mm camera outfitted with a 50 mm lens. The chamber incorporated a 275 µm post. Standard 1“×3” microscopic glass (Fisher Scientific. Masters were fabricated using conventional photolithography technology. The integration of such functional modules into a single device is the goal of the micro total analysis system (µ-TAS). the film was used directly as a mask in photolithography. cut into individual blocks (devices). the mask was used in contact photolithography to generate masters in SU-8 2025. as well as for cell seeding.. WI). 2004). PDMS was prepared by mixing the prepolymer and the catalyst at a ratio of 15:1. Individual devices were cut from the PDMS block. thereby activating adhesive properties of PDMS (Wang et al. or electrical/physical force imposition on cells (Jung et al. 2001). Newton. (3) Device fabrication and assembly: Liquid PDMS was cast against the master and allowed to cure. We also tested an integrated system for cell feeding and waste withdrawal. Saratosa. but such investigations may be beneficial. St. MA) where they were transferred to a chrome mask. . the access ports were plumbed with 22gauge stainless steel capillary tubes connected to flexible vinyl tubing. 1. We have not yet embarked on topographical or physicochemical substrate patterning to modulate cell phenotype and cell behavior and function. Because their manufacture was conducted entirely on-site. 2001). Ossining. and incorporated a sieve with openings ranging from 3 (Figure 2) to 8 µm.... MO) to inhibit cell attachment in their vicinity. or they were used in our on-site mask fabrication facilities. irreversible bond formed. proteomics. Figure 3) was designed larger to accommodate many miniature traps that were outfitted with sieves whose Fig. a negative-tone photoepoxy (Micro-Chem. The NBR devices were fabricated exclusively from glass and PDMS. Topographical patterning refers to the establishing of shape or texture patterns on the substrate.328 Prokop et al. 2003). The first generation of NBRs employed off-chip supply and withdrawal of nutrients and waste. In the latter case. The mixture was degassed under vacuum (20–50 mtorr) for one hour and cured for 2 h at 80◦ C. Design. Pittsburg. and fitted to a glass support. Oxygen-plasma oxidation is known to introduce silanol groups at the surfaces. Mask layouts were first drafted in AutoCAD. MA). The photoepoxy was selectively exposed through the mask and developed. The following steps outline device production: (1) Mask fabrication: masks were drawn with CAD and the designs were printed 100× larger than the final mask.

. By assembling the PG-NBR and MT-NBR. 2002). 40 ng/ml dexamethasone and FBS 10%. IA) provided the necessary UV energy. High permeability for oxygen and carbon dioxide are of benefit for our application. Shear rates were calculated for a certain NBR geometry in order to obtain a sense for ranges experienced by cells exposed to medium flow (see Culture Tests). Hem’s F12 medium with 0. PDMS is optically transparent from 240 to 1100 nm. The assembled NBRs were positioned at a distance of 5 cm from the instrument’s mercury lamp. 2.1 mg/ml G418. later we also tested 25 and 8 µm depths.. A UV-TipCleanerTM (Bioforce Nanosciences. Schematic of perfused MT-NBR with multiple trapping sieves. hydrolysate-free IMDM formulation. Molecular Probes.5 g/L sodium bicarbonate and 4. 15 mM HEPES and 0. We have previously established that these parameters are effective in bioburden removal by using B. DilI stain has been selected as a result from the testing of several possible candidates for noncytotoxic staining of cells in situ. Inc.. Schematic of PG-NBR with 3 µm sieve with a separate channel for seeding. Sterilization. . Ames. WPI). Fig. (2) CHO Epithelial cells (ATCC # CRL-1981). capable of generating nutrient gradients. thus it was not necessary to explore methods to deliver or remove these gases to/from the cell culture chambers.. 240 UV cutoff. Fluorescence microscopy.com). we succeeded in fabricating a sieve/grid system by means of soft lithography. openings were similar to those of the PG-NBR.bucksci. 10%. The box was fitted to the stage of an inverted fluorescence microscope (CK40F Olympus) equipped with a digital camera (QImaging Micropublisher MP-CLR-10).2 g/L sodium bicarbonate. PDMS and gas permeability.5 mM L-glutamine and adjusted to contain 1. Cell viability was determined via a Trypan Blue (Sigma) exclusion test. OR) stain. De Bo et al. The pictures were recorded by means of a CCD camera. the depth of the devices was 45 µm. Efimenko et al. The environment inside the box was maintained at a constant air/5% CO2 and 37◦ C (AirThermTM . In addition to device sterilization.NanoLiterBioReactor 329 Equipment assembly. DMEM supplemented with 4 mM Lglutamine and adjusted to contain 1. Acridine orange/Ethidium bromide (AO/EB. 3.5 mM sodium pyruvate. We believe that this condition aided in chamber wetting and cell seeding. subtilis spores and vegetative cells as a bioassay (see also Moisan et al. Physical characterization/shear rate. 2002. 1998). It could be applied successively in vitro. and VybrantTM DiI (Molecular Probes) and by fluorescence microscopy. DMEM/Hem’s F12 medium 1:1 supplemented with 2.. Charati and Stern. 5 µg/ml insulin. FBS 10%.5 g/L glucose. Cell lines and media. (3) Hepatocytes (ATCC # CRL-2254). Kim et al. the UV treatment was reported to render hydrophilic the originally hydrophobic PDMS surface (Wang et al. Cell culture was conducted within a custom-manufactured Plexiglas incubation box similar in design to those available from Buck Scientific (www. This configuration allowed direct observation and recording of cell status. (4) Hybridoma cells (ATCC CRL-16060). all NBR devices were sterilized under ultra-violet light just prior to use. Standard media were used to cultivate selected cell lines: (1) Mouse fibroblasts (ATCC # CRL-10225). and exposed for 10 min at the maximum power. comprising Fig. Eugene. 2003. 2000. serumfree. 2000). 2001).. 2003. 5 µg/ml transferrin. Initially.05 to 0. allowing for fluorescence microscopy (McDonald and Whitesides. 5 ng/ml selenium. PDMS is known to be highly permeable to gases (Mekel et al. FBS.. To minimize contamination.

10 mm by 10 mm samples were sectioned from the wafers using a diamond tip scribe.38 and 2. The coating densities applied were 8.5 µL/L 2-mercaptoethanol. 5 mg/L holo-transferrin. . Nafion coating. which maintained environmental temperature and gas levels. Plating experiments (for glass substrate) were conducted in standard culture setting to delineate the effect of cell seeding density on cell viability and survival. Standard 12-well microplates were used for PDMS testing. Medium perfusion was facilitated by an UltraMicroPump II nanoinjector actuated with a Micro 4 controller (both by WPI). Because we anticipate the incorporation of processed silicon in future NBR sensors. The wafers with LPCVD silicon nitride layers were provided by Motorola. and those with thermally grown silicon oxide. and viability was thereby assessed. Waste was withdrawn through channels of similar dimension. including collagen. Three cells lines were selected for detailed testing: fibroblasts. sterilized with help of 70% isopropyl alcohol. Gravity seeding (Powers et al. and all potential extracellular matrix (ECM) components. fed-batch. 4. Testing of NanoliterBioReactor design Demonstration of batch and continuous culture growth of selected cell lines within the NBR.1. 2. The sensors were intended specifically for monitoring pH and glucose activity. Naperville. in line with standard coating (Sigma). At the terminus of each experiment. at which point contact inhibition became influential. 10 mg/L insulin. and electrochemical principles (Moussy et al. and continuous feed configurations. The three cell lines were tested for viability and proliferation over extended periods of time ranging from 3 to 5 days. This population was verified with microscopy. and growth ceased. Inc.. Silicon (100) wafers with a layer of low pressure chemical vapor deposited (LPCVD) silicon nitride.44 µL/L 2aminoethanol.. glutamine-free IMDM basal medium. In all cases.. In addition to a direct microscopic count. Cells were introduced into the NBR via a static mode to achieve the desired number of cells per reactor. Untreated glass was used as the substrate for this set of tests. cured PDMS polymer.1 µg/cm2 .. 3. as well as in situ standard and fluorescence microscopy of the culture progress.. obtained via trypsinization of a stock T-flask. was filtered to 30 µm to remove cell aggregates. the NBR was enclosed within a cell culture incubator. biocompatibility tests were conducted on all compositional materials. Nalge Nunc Int. 1. and hepatocytes. 1. were the subjects of this study.67. 1994. 2000). gelatin. The PicoGreen DNA test was used to evaluate the extent of the cell growth achieved in 1–5 days following seeding. Media were vacuum de-gassed to remove dissolved gas and to prevent the formation of bubbles. Media and analytes were supplied to the cells in the NBR chamber via microchannels with cross-sectional area of approximately 250 µm2 . The pump was outfitted with a sterile 1 mL B-D syringe. Glass substrate tests were carried out in eightchamber CC2 Glass slides (Lab-Tek R Chamber SlideTM System. cell suspension. CHO cells. This procedure prevented the NBR channels and integrated filters from clogging with aggregated cells. In all cases. which served as the medium reservoir. periodic culture feeding and waste withdrawal in an overflow mode. PDMS was introduced as a layer at the bottom of plates. Steele et al. During all experiments. standard cell viability assays were performed to monitor experimental progress. IL) and standard tissue culture polystyrene dishes were employed as control substrates. The fluidic system could easily be adapted to address and control individually each chamber in a multiple-NBR arrangement. Gerritsen et al. laminin and poly-lysine. a serial dilution of a standard culture was used to seed the NBR content. Viability and proliferation were evaluated based on a DNA assay via fluorescent PicoGreen assays. additional biocompatibility tests were conducted on silicon nitride and oxide materials.. namely cell growth kinetics and total cell count. Sensing. Cells were allowed to populate the NBR content as attached cells. Prior to seeding. The end-point criteria for evaluation of culture growth and proliferation characteristics were determined by cell viability and surface coverage rates. respectively. 2002) was routinely used to introduce cells into the NBR. NBR seeding. and 10 U/mL penicillin-10 µg/mL streptomycin. fibronectin.330 Prokop et al. The subjects of this testing included the glass substrate. Medium fluid was drawn into the supply line and then delivered to the NBR chamber at the required rate.25. cells were stained with AO/EB OT Trypan blue. Integrated planar array sensors for monitoring cell culture within the NBR were fabricated using an immobilized enzyme technology. This system allowed for batch. the cells proliferated until the substrate was covered by a monolayer. while those with thermal oxide were prepared at North Carolina State Uni- versity. Biocompatibility tests of materials used for NBR fabrication Before cell culture was attempted in the NBR. The controllable external pumps provided automated. extensively (5×) washed with sterile water and coated with ECM components.0 mM glutamine. 1991. 2.

FIB—fibronectin coating. Fibroblasts. Proliferation of CHO cells on glass (and coated glass) as compared to PS (legend same as in Figure 4).) . thereby retarding cell growth to a stagnant status (Tolbert. non-treated PDMS proved to be an inferior substrate compared to polystyrene. plain glass appeared to be ample for cell culturing. For each cell line tested. Results clearly show that a minimum critical density is initially required for some cell lines to commence healthy Fig. While some growth is noted for CHO cells on plain PDMS. The cell longevity may be important for routine use in a mass-screening program. Figures 7–9 compare the results obtained from PDMS filled wells and plain PS.000/cm2 and evaluated against the glass and PS controls. Proliferation of fibroblasts on glass (and coated glass) as compared to PS( legend: PS—standard tissue cell culture quality polystyrene. Volumetric flow rates for the medium ranged from 5–50 nL/min. Fibroblasts and CHO cells were seeded at a density of 10. and 10). and silicon oxide (SiO2 ) were tested for their suitability as substrates in the NBR. Typically. silicon nitride (Si3 N4 ). For example. COL—collagen coating. 6. For all three cell lines. Next. polystyrene (PS). The results clearly show considerable improvement for the plasma-treated PDMS. (Reproduced with permission of Materials Research Society. and glass coated with ECM proteins (Figures 4–6). The actively growing cultures were perfused frequently with fresh medium. The maintenance state was induced by supplying media at lower rates. 1985). The influence of additional ECM coatings was also considered. Results and Discussion Biocompatibility testing of different materials Glass. PLL—poly(lysine) coating. growth can be arrested by removal of growth factors (serum). Seeding density experiments for three basic cell lines were conducted in standard culture environments on glass substrates (Figures 11–13). 4. We can therefore conclude that plasma treatment of our fluidic elements does not adversely affect cellular response. the growth of hepatocytes was particularly sensitive to low seed numbers (Michalopoulos et al. Very low seed densities (100–200 cells/cm2 ) prevented vigorous growth and proliferation in some cases. Finally. CHO cells 18 and 10% of the control. ECM coating improves its biocompatibility. the effect of plasma exposure on PDMS was investigated (Figures 7 Fig. GEL—gelatin coating. PDMS. Finally. This finding has significant consequence for NBR design because the main substrate material is invariably glass. Seeding density Low seeding densities retard proliferation in some cell lines because of the absence of suitable cell-cell and ECM-cell contacts.NanoLiterBioReactor 331 Protocol and system parameters for fed-batch and continuous growth configurations. and hepatocytes were first tested on bare glass substrates. 5. 1982). Fig. At the conclusion of day 3. preliminary tests were performed on silicon nitride and oxide supports (data not shown). Proliferation of hepatocytes on glass (and coated glass) as compared to PS (legend same as in Figure 4). LAM—laminin coating. fibroblasts exhibited 75 and 65% of the glass control (evaluated as DNA). the biocompatibilities of PDMS and PS were considered.. Cultures were tested under two basic conditions: actively growing populations. although some improvement was noted with ECM coated glass. and quiescent cultures whose growth was limited by contact inhibition or lack of nutrients. CHO cells.

.400 cells/cm2 of the NBR area. NBR design: CP-NBR. The fluidic elements were sealed against a glass substrate on which the cells were growing. One is called the “circular NBR with central post” (CP-NBR). 8. Fig. The CP-NBR chamber contains a net volume of 20 nL. 1987. 7. and conditioned media. the “perfusion grid NBR” (PG-NBR).) Fig. so far largely neglected in the literature. another. and tested. manufactured. The addition of the post improved fluid distribution by eliminating dead zones populated with nonperfused cells. These observations are of great consequence for initiating the cell growth in the NBR environment. (Reproduced with permission of Materials Research Society. Fig. 280 and 1. 1982). the employment of the NBRs for mammalian cell culturing opens a new paradigm of cell biology. growth.. The above results serve as models for cell behavior within the NBR.. Lowering down the chamber height to 8 µm. CHO) are capable of developing a colony from a single seeded cell (Konrad et al. Pomp et al. 5. The CP-NBR involved an 825 µm diameter culture chamber of 40 µm height. For CP-NBR (see below). Michalopoulos et al. Proliferation of fibroblasts on PDMS (legend same as in Figure 4). PG-NBR. fitted with a central post of diameter 275 µm.g. 9.332 Prokop et al. On the other hand. growth factors. MT-NBR Several NBR designs were conceived. 1996. we employed between 10–20 cells per seed. One way of improving cell plating efficiency is to employ ECM. Summarizing. and third version. somewhat on the low side. The chamber was connected to inlet and outlet ports by 100 µm channels (also of 40 µm height). Proliferation of hepatocytes on PDMS (legend same as in Figure 4). Park et al. the “multitrap NBR” (MT-NBR). . some cell lines (e. Proliferation of CHO cells on PDMS (legend same as in Figure 4). 1977.. some success was noted. 20 and 50 cells seed per NBR corresponded to about 140. Typically. The perfusion PG-NBR was difficult to design without the cell leakage between the sieve and the substrate..

NanoLiterBioReactor 333 Fig. Fig. 13. 10. Reproducibility of using such design was not assured. Fig. (See contrast with a static T -flask culture environment with huge media layer above the attached Fig. Culture tests Batch growth within the NBR is not a viable possibility because the medium volume is extremely small. . The MT-NBR would allow a dynamic seeding avoiding the cell attachment to coated areas of the substrates. with more material attached to the glass bottom. and cells could therefore suffer from a nutrient limitation even in early culture stages. however. Effect of plating density on proliferation of CHO cells. Proliferation of fibroblasts on coated PDMS and plasma-treated PDMS (legend same as in Figure 4). Only when the design of sieve posts was changed to a more robust one. we were able to fabricate such NBRs without any leaking. 12. 11. Effect of plating density on proliferation of fibroblasts (note the lower detection limit is as low as 10 cells). Effect of seeding density on proliferation of hepatocytes.

Aggregate use is reserved for the MT-NBR design only. cell viabilities were much lower as accessed by Trypan blue stain. demonstrated that the glass/PDMS chamber is a satisfactorily biocompatible environment. For this reason. 1998). Flow rates conducive to cell quiescence were observed to be on the order of 5 nL/min for all cells lines we considered.. We did not make any attempt to follow their function and did not establish the maintenance of differentiated phenotype. including a continuous removal of mitotic cells by shear. although the filtering device appeared perfect. 2 day culture. We have reserved these studies for the MT-NBR.) However. We observed that some rounded-off freshly-trypsinized cells (resulting from T -flask culture). while the other cell lines required close to 16 h. Typically. We also intentionally avoided the use of hepatocyte aggregates (Parsons-Wingerter and Salzman. with focal adhesion points. in which nutrient gradients can easily be established. We have not yet addressed the option of cell patterning. 5 and 8 µm allowed some penetration due to cell deformation. 1992). 14. could penetrate the sieve perforations and completely squeeze through in some instances. The PG-NBR was particularly adequate in retaining these cells and maintaining reasonable viabilities in a perfused state lasting up to 5 days. cells were perfused at a rate corresponding to one-fourth to one-half of the chamber volume per minute. we employed static seeding just prior to the start of the feeding process to allow cell attachment. For fibroblasts. Typically. Beyond that time. It may not be so critical at this stage of the project. van Kooten et al. Typical examples of fibroblast and CHO cultures (in CP-NBR) are illustrated in Figures 14 and 15.334 Prokop et al. such experiments. These values are too low to affect cell physiology (Prokop and Bajpai.g. The mean diameters of cells right after the trypsinization were thus assessed and found to be in the range of 11–18 µm for fibroblasts. in line with our results. Nevertheless. only fed-batch and continuous modes were used in the PG-NBR. viability rates averaged 95% under a continuous feed regime. It should be mentioned that the shear stresses cells experienced in the NBRs ranged from 5·10−5 to 1·10−2 dynes/cm2 for flow rates of 5–50 nL/min. Fibroblasts in CP-NBR. which were used as seed. A few experiments were performed on hybridomas. but feeding rates lower than this led to poor viability. Hepatocytes were cultured 10 days (Figure 17). low seed (20 cells). exemplary employment of MT-NBR is in Figure 16. This way a quiescence state of the remaining cells (for CHO cells) was obtained. Higher perfusion rates were also tested. For all three cell lines. fibroblasts required a 2–3 hr attachment period.. 1993) as a seed material in order to minimize clogging of NBR channels. 11–25 µm for CHO and 16–30 µm for hepatocytes. . higher flow rates yielded elongated cell morphology and lining up with the fluid flow stream. although Fig. Special consideration of cell deformability is relevant to the sieve design in both PG-NBRs and MT-NBRs. We thereby concluded that sieve openings of 3. Very few viability measurements of mammalian cells on glass or PDMS substrates have been reported (e. cells. The attachment period was not necessary in the case of the PG-NBR because its design offers total cell retention.

Unfortunately. Fig. some visible outside of traps). 16. 15. A partial solution was achieved with the silicone coating as previously mentioned. we could access a measurement of several metabolic uptake rates across the NBR). Jung et al. The importance of micropatterning is discussed in more detail elsewhere (Folch and Toner.. spindle shape cells and few rounded mitotic cells. 2001). no patterning effort will ever produce a complete absence of cells without difficulty. 2000. CHO cells in CP-NBR.g. .. provided that channels in and out of the NBR are free of cells.NanoLiterBioReactor 335 Fig. (Reproduced with permission of Materials Research Society.) some advantages could be visualized (e. Freshly seeded CHO cells in MT-NBR (cells captured in small traps. 5 day culture.

In fact. We also intend to perfect our sensing capabilities (particularly for gradient studies) via NBRs boasting integrated sensors. we now realize that this configuration provides an interesting opportunity for the study of physiology in confined spaces. NBR reuse Special attention was given to cell removal and lysis following the employment of the NBR. Growth in confined space An observation on the contact inhibition of cells in confined spaces resulted from NBR sieve design work. all three cells lines featured very limited proliferation and quiescence over period of 5 days. as will mitotic cell removal in the circular NBR with open inlets and outlets. While our original consensus was that the utility of the NBR would be limited when the chamber is packed full with cells. few cells penetrated through the sieve channel on the right). no attempts were exerted to characterize the quiescent state of cells by a molecular means. Future research efforts will concentrate on expanding the culture experiments to include longer-term assays. proliferation. Hepatocytes in PG-NBR (4d culture growing in aggregates. We hope to enable extension of this concept into a rational tool for studying fundamental cell biology phenomena.and nano-fabrication is considered an enabling technology that is expected to generate significant new . General discussion Micro. We also intend to explore the effects of shear stress and gradients on cell physiology within the multitrap and grid NBRs. cell debris and ECM components are typically left behind and contaminate the NBR space. At this time. These efforts are aimed at obtaining an improved fundamental understanding of cell physiology at the small-scale number density that the NBR affords. and accordingly grew in size asymmetrically. Sensing in NBR and outlook Evaluation of NanoBioReactor temporal physiologic responses and determination of the sensor sensitivity to interventions that alter the homeostatic state of three cell lines in continuous culture is an interesting proposition. Under these conditions. Further studies will also include considerations such as cell differentiation (expression of specific function) vs. In an effort to achieve a complete lysis of cells and of the ECM components from the NBR interior. Fig. some attempts were made to limit the sieve leakage by lowering the chamber height to 8 µm. although the low cost of fabrication renders it an easily disposable device. a confined space may help to mimic the three-dimensional cell-cell contact behavior. featured a “flat” morphology while they were squeezed between the NBR’s ceiling and floor.336 Prokop et al. The effect of complete cell retention in the PG-NBR will be further investigated. we tested several gentle lysis reagents designed to release nucleic acids and proteins. A successful solution would enable a reuse of the NBR. As mentioned. However. Trypsin itself was partially effective. 17. especially larger ones such as hepatocytes. Cells.

28 d PC12 cells. sensors. 2003 Hafner.. Viravaidya and Shuller. 24 h culture Mouse fibroblasts. The enabling aspects of micro/nano technologies in the field of biology are best defined as study of biological phenomena at the micro. 2003 Jager et al. The ultimate objective is to design and control experiments at the micrometer scale. membrane chamber. glass/polycarbonate. surface detachment. The latter three aspects are briefly reviewed in this section. no growth Xenopus.1 mL 750 µL 0. separating. 2002 Johannessen et al. 2004. adipocytes. silicon. 16 h Primary rat hepatocytes. cardiomyocytes.. Morphological changes. polystyrene. L2 rat lung cells. DO. coli. deformability. Substrate materials (from both fabrication and cell culture points of view) often in- clude biocompatible materials such as silicon dioxide and nitride.. PDMS. 14 d Powers et al. silicon. 14 d.. 1995 Zanzotto et al. manipulating and characterizing cells include cellular adhesion. These objects can be stationary or moving structures. Hep2G. PDMS and other polymers.12 mL 2.NanoLiterBioReactor 337 product/market opportunities through integration of science and technology. pumps. glass/PDMS. metabolism and secretion (Zieziulewicz et al. polycarbonate/PDMS. and particle gathering are additional physiologic activities relevant Table 4. CHO. glass/PDMS.4 nL 6 nL 20 nL 0. pH and optidal density Light & fluorescence microscopy N/A 27 nL 5–50 µL 21 nL Microfluidic single-cell analysis Single cell clinic Perfused 3-D liver culture Bacterial culture Cell-based silicon sensor Capillary cultivation system Picocalorimetry at single cell level Microvascular mimic Single-cell monitoring NanoBioReactor 0. Some novelties introduced via nanotechnology will be discussed later. medical devices and biosensors. 2004 Hediger et al. necrotic growth. signal transduction. 8 d Mammalian cells..2 nL 16 nL 5. biochemistry. The scope of this field is to enabling novel capabilities in various applications by creating objects with dimensions in the range of nanometers and micrometers to millimeters (Voldman et al. PDMS. PDMS. Status of micro. 2004 1. 10 h Microbial. 2002 Huang et al. 2003 Grodrian et al. etc. motility. short-term Adipocytes. glucose.3 mL In situ sensing DO Comments Rat hepatocyte. 2003). 2002 Borenstein et al. PDMS/silicon. 6 d E. short-term measurement Fibroblasts.. 2002 Leclerc et al.and nano-scales. 2001 Weibezahn et al.. channels.... 2000. primary rat hepatocytes. apoptosis. Eklund et al. 22 h Colorectal carcinoma. 2000. several days Recombinant cells.. pC Primary rat perfused hepatocytes. Sin et al. few hours Epithelial cells. not individually addressable Jurkat. short-term Endothelial cells. 2002 Wheeler et al. Balcarcel and Clark..7 nL N/A 0.. lactate production.5 nL 0. 2004 This paper ...and nanobafricated cell culture devices Device function Perfusion mimic of pharmacokinetic animal behavior Well-plate adapted for CO2 measurement Well-plate adapted microbioreactor Bioartificial liver in gradient reactor Cytosensor Microphysiometer Biomedical diagnostic system for epithelial cells First microreactor liver mimic High-throughput membrane-aerated microbioreactor Perfused 3-D liver culture Volume 4. U937 cells... plastic. silicon. Possible applications include areas of molecular biology. 6 h HepG2. 2004 Yang and Balcarcel. cell biology. oxygen and glucose consumption Impedance (growth) None DO.. 10 d Mouse L cells.8 µL CO2 production Biomass. polyimide.7–22. silicon. 2002 and 2004 Chang et al. Manipulations of the physical world at this scale with the intent of detecting. 1999). glass. 2004. 2003 Brischwein et al. 2003 Maharbiz et al. pH None Acidification rate.05 nL 10–100 nL Ca flux Impedance (growth) None Impedance (growth) Acidification None Picocalorimetry (pC) None Amperometric pH. 14 d Bacterial fermentations Reference Ghanem and Shuller. 2004 Allen and Bhatia. namely growth chambers.

A. 89 (2001).and long-term cultures in confined spaces. Chaudhury. Chang.H.J.R. Glaser. R. M. 306 (2002). 253 (2003). Silicon wafers were kindly provided by Bridget Rogers. Crosby. Huang.M. Microengineering of cellular interactions. Interface Sci. K. Table 4 lists various micro-.W. 19. Svoboda. M. Anal. E. Lee. Wolf. and J. Sedlak. 234 (2003). 111 (2001). Sprakel. Gartner. Kohler. to this scale. Kim. I. Acknowledgments This work was supported by in part by a National Institutes of Health grant 5 R43 RR016124-02 to NanoDelivery.J.E. Towards electronic Petri dishes and picolitre-scale singlecell techniques. Rev. W. and by the Vanderbilt Institute for Integrative Biosystems Research and Education. Wang.A. and M. 226 (1999). Progr. 69.P. and D.T. Kros. Bourova. We anticipate that the NBR will demonstrate itself as a technology suitable for incorporation into massively parallel. Z.J. B. Chip.-H.A. Biocompatibility evaluation of sol-gel coatings for subcutaneously implantable glucose sensors. Diffusion of gases in silicone polymers: Molecular dynamics simulations.P. D.) with spatial and temporal control. Lundstrom. Clonal variation in colony morphology and growth of CHO cells cultured on agar.338 Prokop et al. Eklund. Vacanti. King.J. H. C. and D. Only very few devices listed in this table allow for in situ physiologic measurements or for perfusion mode of operation. Efimenko. Micromachined nanocalorimetric sensor for ultra-low-volume cell-based assays. D. and quantal release monitoring of single cells on a microfluidic device. 215. A. Topographical and physicochemical modification of material surface to enable patterning of living cells. Jung. (6) Determine the maximum duration of stable cell culture in the NBR. Prog. Magnusson. etc. and E. Sayah. 303 (2003). H. A. Weaver. Otto. M. Bioelectr. Cytosensor microphysiometer: Technology and recent applications. Characterization of a perfusion reactor utilizing mammalian cells on microcarrier beads.H. M. Zhang.K. Horisberger. Future iterations of the NBR should incorporate ways to (1) Automate cell loading into the NBR. Sci. and R. V. I.T. Membrane.H. J. Chem. Chem. Cui.A. W. Jager. A microphysiometer for simultaneous measurement of changes in extracellular glucose. Taylor. Peterson. J.. (5) Extend the number of cell lines that can be cultured stably.M. Giuliano.M. H. and tissue-based devices. 16. K. Kozlov. Roth.G. Akin. Microdev. 174 (2002). Macromol. and L. J. P. 71 (2000). Crit. and J. 689 (2001). 21.E. J. 31. Chem. Owen. 483 (2004). W.and Nanoengineering. Investigation of the permeability and selectivity of gases and volatile organic compounds for polydimethylsiloxane membranes. Konrad. S. De Bo. 2190 (2002). Microdev.. Nolte. E. Z. P. E. Wallace. Folch and M. and J. D. Inc. E. Segmented flow generation by chip reactors for highly parallelized cell cultivation.V. 2003). References J. 226. E.H. Th. A. Cell 10. including on-chip pumps and valves. A. lactate production. Pang. 519 (2004). Analyt. constructed with the aim to monitor short. R. Poly(dimethylsiloxane) (PDMS) and silicon hybrid biochip for bacterial culture. Cabala.-W. Physiologic functions that can be readily monitored via multicomponent analyte systems include pH fluctuations. 82. Special designs accommodate the generation of complex gradients (e. Clark. Metze. Biomed.M. milli-. Modular microsystem for epithelial cell culture and electrical characterization. Becker and C. 21. M. Sci. Biotechnol. R. Lab. 15. Gerritsen. E. D. Chemical Engineering Department. 17. P. De Neve. Allen and S. (9) Conduct comprehensive metabolic flux analyses using the optimized. Grothe. and (10) Determine the requirements for construction of NBR arrays. Biosens.A.-F. Borenstein. L. Bratten.M.L. and J. oxygen consumption. Transport. Jansen. F. D.A. Formation of steady-state oxygen gradients in vitro. Micromachining sensors for electrochemical measurement in subnanoliter volumes. Experiments in Transport Phenomena (Wiley. Lutterman. Cobbold.-J. Hydrophobic recovery of polydimethylsiloxane elastomer exposed to partial electrical discharge. (7) Enhance automation capabilities to operate multiple NBRs. and acidification rate. Biomed.D. Cooper. Surface modification of Sylgard-184 Poly(dimethyl siloxane) networks by ultraviolet and ultraviolet/ozone treatment. I. Toner.A. Cooper. Bashir. high-content.-K. Terai. cell-. Mrksich. Nicolau and A. Analyt. and B. Thompson. 305 (1977). (3) Optimize control parameters for each cell line. K. The cell clinic: Closable microvials for single cell studies.C. L. 177 (2002). Storrie.R. 76. 149 (2000). 471 (2000). Van Langenhove. In Biomedical Applications of Micro. Ladisch.-L. and M. Biosensors can be embedded within such devices to allow for molecule-. glucose utilization. Microfabrication technology for vascularized tissue engineering.M. 281 (2003).R. Hafner. H. P. location. 82. Shuler. Functional cellular assays with multiparametric silicon sensor.H. S. S. and O. Metabolic screening of mammalian cell cultures using well-plates.-E. J. Cahrati and S. Brischwein. 253 (1997). Biotechnol. Inganas. Biomat.E.. Bioeng. Chem. however.g. 231 (2000). 227 (2000). Cheng. Hediger. Balcarcel and L. Biomed Microdev 4.R. SPIE 4937. Prokop. and D. multiphasic. of substrate. J. Hijs. Cliffel. Revs. Anal. 98 (2003). K. and J. 6. (8) Incorporate highly-versatile microfluidics into the NBR. Proc.W. 3. J.R. Vakelecom. Adams.F. Polymer based micro-reactors.M. K. Henkel. Immerstrand. and redox potential (Zieziulewicz et al. Taylor.M. 16. Bhatia. Chang. Ghanem and M. A. Coll. Interface. has been unprecedented in last few years. J. Biotechnol. H. chemokines. Genzer.L. Biosensors Bioelectr. E. 76. A. Pieters. T. Stern. Craighead. Cooper. M. Weinberg. J. Johannessen. lactate. Kapur. W. and nanoliter scale culture devices. The emergence of such devices. Biotechnol. Trends Biotechnol. fully instrumented NBR. Pruuost. M. M. Grodrian.A. Coll. Biotechnol. 167 (2002). and J. M.J. oxygen. C. toxicology screening systems. KaazempurMofrad. . 254. 5529 (1998).W. 4. (4) Add additional sensors and control loops to NBR. Molec. Ann Rev Biomed Eng 2.M. 74. Dick. Gobbold.N. New York. J. Motrescu. (2) Develop single and multi-parameter control loops for NBR system variables.R. J. 1961). J. edited by D.

vol. K. 20.D. 335 (2001). and J.F. Saltzman. D.K. Bier.J. 129 (1987). Y. and T. Techniques. 97– 123. 9. A. and L. P. 95 (1995). 1 (1998). Y.A. 64. Biotechnol. Miniaturized total chemical analysis systems: A novel concept for chemical screening. Biotechnol. A. Determination of carbon dioxide production rates for mammalian cells in 24-well plates. Zare. Smith. Ingber. Schrier. Jamil. Bio. A. Griffith. 289 (2004). diffusion. Wike. and D. Y. Tabrizian. Holtz. K. Microbiol. J. 491 (2002). Bioeng. A.M. W. J. Steele. Res.A. M. J. New York. J. P. Kaazempur-Mofrad. A. Biomed. Thames. 7. M. Oncol. Feder and W. 1442 (2003). 27. Microfabrication in biology and medicine.H. Harrison. Membrane-aerated microbioreactor for high-throughput bioprocessing. 203 (1931). Weibezahn. Underwood. 257 (2002).D. 750 (2004). 20.L. 531 (1991). and R. and K. 87. New York. Sakai.J. Anticancer. J. 286 (2004). pp. 1985). G.D. Organs. 14 (Suppl. 5. Biotechnol. 415 (2000). Eletrophoresis 24. Z. J.F.I. M.R. A. Artif. Account Chem. Bajpai.J. Rev. Johnson. Widmer. Dodwell. 3249 (2003). 121 (1996). Parsons-Wingerter and W. Barbeau.D. M.H. Clonogenic assay combined with flow cytometric cell sorting for cell-cycle analysis of human leukemic colony-forming cells. Schiller. J.M. Kurzawski. von Recum.F. Prokop and R. and M.W. Michalopoulos. R. 1 (2001). W. Shuller. H. Szita. and T. K.R. 1992). 4673 (1982).J. J. and K. Park. Part I: Product. Whitesides. 74. Zieziulewicz. Liao. van Kooten. Yang and R. 1. S. Int. A miniaturized Nafion-based glucose sensor: In vitro and in vivo evaluation in dogs. Polymer. E. Th. Hoogstraten. and R. Rajotte. 85. Inst. Dvelaar. Biomed. K. Shuller. W. N. Whitesides.L. A. Gray.F. Sci. Low-temperature sterilization using gas plasmas: A review of the experiments and an analysis of the inactivation mechanisms. 78. Lessard. and M. Biotechnol.R. edited by A.D. The design and fabrication of three-chamber microscale cell culture devices with integrated dissolved oxygen sensors. Phys. and W.E.I. A microfabricated array bioreactor for perfused 3D liver culture. Kostov. I.D.M. Appl. Kamm. Liver regeneration studies with rat hepatocytes in primary culture. and I. J. Lawrence. Biomat. G. 235 (2003). Jiang. Cell culture in 3-dimensional microfluidic structure of PDMS (polydimethylsiloxane). J. Poly(dimethylsiloxane) as a material for fabricating microfluidic devices. Schubert. and T. K.A. 70 (2001). 401 (1999). Jensen. Exp. Radiother. E. Dertinger. Throndset. Graber. Unfricht. Pomp. Balcarcel. T. Mater. Gas sorption. Perfusion culture systems for production of mammalian cell biomolecules. Tolbert. A. Kalezi. Prokop. M. F. Oleschuk. Sakai. G. Sin. S.H. In Large Scale Mammalian Culture. 165– 232.A. and H. Appl.R.A. J. Influence of silicone (PDMS) surface texture on human skin fibroblast proliferation as determined by cell cycle analysis. J.N. Laskin (Academic Press.NanoLiterBioReactor 339 E. Microfluidic device for single-cell analysis. Rao. 35.) 43.T. Maharbiz. Fujii. Domansky. N. J. Biotechnol. Amer. Leach. Prokop. W. Knedlitschek. Microdev. Takayama. Chin.C. H. Wheeler. Wack. Microbiol. Manger. K.W. Whitesides. Biotechnol. 8. Advan. Biomed. Biotechnol. B. Sinskey. Kimler. E. Bondar. 1). Bioeng.A. Kligerman. K.F. The sensitivity of biocatalysts to hydrodynamic shear stress. P. A. Fujii. T. Annu. Sensors Act B1. Cancer. Trans..C. Moreau. Daridon. Oldshue. Res. C. Bioeng. A. Engr. A. McDonald and G. Pelletier.B. A. Progr. Tolbert (Academic Press. P.G. R.M. H.L. Keasling. Progr. Viravaidya and M. A. . J. Gen. 376 (2004). Surface characterization using chemical force microscopy and the flow performance of modified polydimethylsiloxane for microfluidic device applications. Int. Yahia. Challenges in commercial biotechnology.R. 600 (1993). Schmidt. D. pp. process and market discovery.Y. Polymer.V. Abdulali-Kanji. and L. Stolz. Annu. S41 (1995).J. Brock.M. R. Phamaceut. Moisan. K. A.G. A. Cianciulli. Micro total analysis system (µ-TAS) in biotechnology. T. Res. 109 (2002). Clin. Whelan. Hadjout. Shrinking the biologic world—Nanobiotechnologies for toxicology. Res.J. Cancer Res 42. 36. Boccazzi. Freeman. Cell density dependent plating efficiency affects outcome and interpretation of colony forming assays. Incorporation of 3T3-L1 cells to mimic bioaccumulation in a microscale cell culture analog device for toxicity studies. M. and A. Int. Jirtle. Lynes. S. 243 (2004). Nagai. Cytotoxicity tests for high-throughput drug discovery. Adhesion and growth of cultured human endothelial cells on perfluorosulphonate: Role of vitronectin and fibronectin in cell attachment. C. Novotny. R. V.E.H. Leer. D. Powers. Zanzotto. Mekel. Y. Y. Horton.B.L. Fluid Mixing Technology (McGraw-Hill.R. Rev. Biomater. J. Progr.D. and P. Chem. 40. In Advances in Applied Microbiology. Anal Chem 75.C. J.-H. Howe.L. Eng. Pinnau. 1 (1982). Growth versus function in the three-dimensional culture of single and aggregated hepatocytes within collagen gels. Ouwerkerk. Microfluidic PDMS (Polydimethylsiloxane) bioreactor for large-scale culture of hepatocytes.Y. 244 (1990). 12. 590 (2004). B. M.C. Upadhyaya.K. J. I. Ostuni. G. New York. 226.M. (Appl. Lee. and D. Biomed.G. Leclerc.M.H. Microbioreactor arrays with parametric control for high-throughput experimentation. M. Syst. Farrell. Curr.R. Manz. Biotechnol. X. edited by J. and permeation in poly(dimethylsiloxane). J. and R. 37. Opin. Sci. Lee and S. Biotechnol. Progr. Newman. Eng. Strom. 12. Slater.D. 17. and A.J. N. 20.J. 88 (1994). B. 338 (2004). Moussy. Toxicol. Soft lithography in biology and biochemistry. 1983). Capitano. W. 40. 3. Reconstruction of tissue layers in mechanically processed microstructures. P. B 38. Leclerc. S. Wang. Norris. Voldman. Systems analysis and synthesis in biology and biotechnology. G.