Three-Dimensional Printing of Photoresponsive ... - ACS Publications

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

3D Printing of Photoresponsive Biomaterials for Control of Bacterial Microenvironments Jodi L. Connell, Eric T. Ritschdorff, and Jason B. Shear Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b03440 • Publication Date (Web): 26 Oct 2016 Downloaded from http://pubs.acs.org on October 28, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

3D Printing of Photoresponsive Biomaterials for Control of Bacterial Microenvironments

Jodi L. Connell,1 Eric T. Ritschdorff,1 Jason B. Shear*

Department of Chemistry, University of Texas at Austin, 1 University Station A5300, Austin, TX 78712

1

These authors contributed equally to this work.

* Corresponding author: [email protected]

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT

Advances in microscopic three-dimensional (µ3D) printing provide a means to microfabricate an almost limitless range of arbitrary geometries, offering new opportunities to rapidly prototype complex architectures for microfluidic and cellular applications. Such 3D lithographic capabilities present the tantalizing prospect for engineering micromechanical components – for example, pumps and valves – for cellular environments comprised of smart materials whose size, shape, permeability, stiffness, and other attributes might be modified in real time to precisely manipulate ultralow-volume samples. Unfortunately, most materials produced using µ3D printing are synthetic polymers that are inert to biologically tolerated chemical and light-based triggers and provide low compatibility as materials for cell culture and encapsulation applications. We previously demonstrated feasibility for µ3D printing environmentally sensitive, microstructured protein hydrogels that undergo volume changes in response to pH, ionic strength, and thermal triggers, cues that may be incompatible with sensitive chemical and biological systems. Here, we report the systematic investigation of photoillumination as a minimally invasive and remotely applied means to trigger morphological change in protein-based µ3D-printed smart materials. Detailed knowledge of material responsiveness is exploited to develop individually addressable “smart” valves that can be used to capture, “farm,” and then dilute, motile bacteria at specified times in multichamber picoliter edifices, capabilities that offer new opportunities for studying cell-cell interactions in ultra-low-voume environments.

2


Page 2 of 18

Page 3 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION

Recent proliferation of lab-on-a-chip technologies offers opportunities to manipulate chemical and biological materials with unprecedented control. The dramatic reductions in scale provided by these platforms enable, for example, ultra-low-volume sample aliquots to be delivered to precise locales and microscopic populations of cells to be

sequestered,

patterned,

and

harvested

for

molecular

characterization.1-4

Unfortunately, most lithographic strategies in common use provide only two-dimensional control, limiting both the capabilities of resultant devices and their potential relevance to native biological environments. The growing realization that, in many situations, enhanced

capabilities

would

be

achieved

by

true,

three-dimensional

(3D)

microfabrication has motivated interest in alternate technologies. Innovations in micro-3D (µ3D) printing, in particular methods based on multiphoton photolithography (MPL),5 have provided a means to fabricate 3D objects that have essentially any conceivable geometry. By programming the movement of a tightly focused, pulsed laser beam through reagent solution, it is feasible to define elaborate objects such as cellular capture chambers, Möbius strips, interlocked blocks, and microscopic animals with submicrometer feature sizes.6-8 Although such geometrical control is visually impressive, in most instances this procedure yields inert polymeric materials whose physical properties are unresponsive to environmental conditions, including both chemical and photo-based triggers. To create 3D microstructures with greater functionality in biological and micromechanical applications, we (and others) have explored strategies for adapting µ3D printing to create microscopically defined protein-based hydrogels.9,10 These materials, formed from solutions containing concentrated protein and an efficient photosensitizer, are characterized by high porosity (and thus permeability to many solutes), tunable elasticity, biocompatibility, and versatile chemical properties, attributes that make them ideally suited for cell culture and cell encapsulation.11 We previously reported the ability to dynamically modify both size and shape of protein-based microstructures using chemical triggers, and to exploit these changes both for microactuation and manipulating bacterial microcolonies.12 Although interesting as proofof-concept studies, actuation of 3D materials required relatively large shifts in pH or ionic

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

strength, both of which may have deleterious, and potentially prohibitive, impact on functioning of microdevices (e.g., microfluidic platforms) and viability of biological systems under study. Moreover, by relying on changes in solution composition, this approach to microactuation depends on how rapidly and reproducibly solutions within microenvironments of interest can be exchanged, limiting both applicability and precision. As a consequence of these limitations, alternate strategies for triggering actuation of protein-based µ3D printed materials could substantially extend the utility of protein-based 3D hydrogels as smart materials in ultra-low-volume chemical and biological applications. We report here observation and detailed examination of photo-stimulation as a means to modify the morphology of protein-based hydrogels fabricated using MPL, and show utility of this approach for controlling the function of micro-objects fashioned from these materials. Conformational responses are assessed of microstructures composed of different proteins (i.e., bovine serum albumin, avidin, lysozyme, myoglobin), photocrosslinked under various conditions, and subjected to different illumination protocols. Utility of these light-responsive protein microarchitectures as a means to capture, “farm,” and release small populations of microorganisms is demonstrated. EXPERIMENTAL SECTION Reagents. Rose Bengal (RB; 33000), methylene blue (MB; M-4159), crocin (17304), curcumin (C1386), L-ascorbic acid (A7506), poly-L-lysine solution (0.1% w/vol; molecular 150 – 300 kD; [P8920]), lysozyme from chicken egg white (“Lys”; L6876), and myoglobin from equine skeletal muscle (“myo”; M0630) were purchased from SigmaAldrich (St. Louis, MO). Bovine serum albumin (BAH64-0100) was obtained from Equitech-Bio (BSA; Kerrville, TX). HEPES sodium salt (AC21500-1000), and fluorescein (11924100) were acquired from Acros Organics (Geel, Belgium). Avidin (A-887) was purchased from Invitrogen (Carlsbad, CA). Sodium azide (SX0300-3) was obtained from EM Sciences (Gibbstown, NJ) and sodium chloride (S271-3) was purchased from Fisher Scientific (Fairlawn, NJ). Poly-L-aspartic acid sodium salt (5 – 15 kD; [151909]) was purchased from MP Biomedicals (Solon, OH). Tryptic soy broth (1010717), carbenicillin

4


Page 4 of 18

Page 5 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


disodium salt (BP2468-1), obtained from MP Biomedicals (Solon, OH) and Fisher BioReagents (Fairlawn, NJ), respectively, were used in cellular experiments. All reagents were stored according to the supplier’s specifications and used as received without further purification. Protein hydrogel fabrication. Photo-crosslinked hydrogel microstructures were fabricated on no. 1 coverglass secured in a reusable aluminum well coated with an epoxy resin using a dynamic maskbased µ3D printing technique as described previously.13,14 In brief, the output of a femtosecond mode-locked titanium-sapphire (Ti:S) laser (Tsunami, Spectra Physics) tuned to 740 nm was aligned into a dual-axis galvo scanner (Thor Labs) and rasterscanned across the face of a digital micromirror device (DMD) from a partially dismantled business projector (BenQ) functioning as an electronic reflectance mask. Complex 3D features were formed in a layer-by-layer process directed by mask images that were displayed by the DMD in a sequence synchronized to stage increments in the optical axis. The laser beam was expanded to overfill the back aperture of an oil immersion objective (Zeiss 100X Fluar, 1.3 NA) situated on an inverted microscope (Zeiss Axiovert), and a motorized focus driver (Prior Scientific) was programmed to translate the fine focus along the optical (z) axis in defined steps between scanned planes. Except for experiments specifying use of a range of fabrication conditions, microstructures were created using an average laser power (measured at the back aperture of the objective) of 33 mW, a slow axis scan time of 3.3 s, and an optical axis step size of 0.5 µm (i.e., the interval between adjacent fabrication planes). Experiments in Figure 1a-c each varied one of the these three parameters, with presented results relying on the values of the two constant parameters. Except where otherwise noted, microstructures were fabricated using a reagent solution prepared in a 20 mM HEPES (pH 7.4, 0.1 M NaCl) buffer containing either 8.5 mM Rose Bengal or 5.0 mM methylene blue, and either avidin, BSA, or lysozyme at 400 mg mL-1, or myoglobin at 40 mg mL-1. Rectangular test structures were fabricated having nominal (i.e., designed) dimensions of 15.0 µm x 20.0 µm x 8.0 µm (width x length x height). Bacterial microstructures contained chambers that were nominally 8.0-µm floor-to-ceiling with a 1.5-µm thick roof.

5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

After fabrication

was

complete,

reagent

Page 6 of 18

solutions

were

removed

and

microstructures were rinsed 5 times (2 mL per wash), incubated for 5 min, and washed an additional 5 times, all with the same HEPES buffer used to prepare the reagent solution. For experiments examining perturbations to the environment using charged poly-amino-acids, proteins, or dyes, such solutions were washed into the sample well after the initial set of rinses and 5-minute incubation in buffer. Macroscopic structural integrity of microstructures was maintained for periods of at least many hours. Detailed, quantitative studies of solute release rates from protein matrixes have not been performed; however, some matrixes (e.g., RB/BSA) were visually observed to retain similar optical densities over the course of many hours. Moreover, RB/BSA test structures that were washed, left in buffer for 24 h, and washed again before illumination, expanded to the same mean A/A0 value as structures that were washed and irradiated immediately after fabrication. Visible-light illumination conditions. Samples were incubated in charged-species solutions for 15 min and then either illuminated, or washed, incubated, and washed in buffer again prior to illumination. Protein structures were irradiated in the same sample well used for fabrication in 2 mL of buffer using the microscope’s broadband transmitted-light illumination system. To irradiate a sample, all filters and the DIC polarizer were removed from the illumination pathway, the field and condenser irises were fully opened, and the 100-watt tungstenhalogen lamp was set to the highest voltage (12 V). Except where otherwise stated, samples were illuminated using these conditions for a period of 15 min. Filters used for limiting the spectral bandwidth of illumination (Figures 2 and 4) were placed in the illumination pathway between the tungsten-halogen lamp housing and the sample. Imaging and data analysis. Samples were imaged post fabrication, at the end of each wash step, and immediately before and after each illumination using a 40X air objective (Olympus, Plan Apo, 0.95 NA) and an ORCA-Flash2.8 CMOS camera (Hamamatsu). As a quantitative measure of test structure size, cross-sectional area (A) was determined from brightfield images acquired by focusing at the top of each structure, a process that was aided by

6


Page 7 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fabricating a low-profile (triangular) marker on structure roofs. The swelling ratio, A/A0, was calculated by dividing the area of a structure measured after irradiation (A) by the area measured before irradiation (A0), and provides a facile and reproducible means to assess expansion/contraction extent using two-dimensional imaging. Determination of volumetric swelling ratios would additionally need to account for expansion or contraction along the optical axis, which often is similar (but generally not identical) to changes in each of the radial dimensions. Each reported swelling ratio represents the mean of at least four separate structures and the error is given as one standard deviation. Bacterial cell-culture conditions. Pseudomonas aeruginosa (P. aeruginosa) is a common environmental pathogen characterized by a well-studied set of population-dependent adaptations, including onset of quorum sensing and reversible resistance to certain antibiotic classes.13,15 In the current studies, the PA01 strain carrying the pMRP9-1 plasmid to constitutively express GFP was grown aerobically overnight to saturation at 37°C in tryptic soy broth (TSB) containing carbenicillin (300 µg mL-1) for plasmid selection. PA01 cells were then diluted in TSB and grown to mid-exponential phase; a 2-mL aliquot of this culture was then added to a sample well containing BSA microchambers, each designed with an entry pore for bacterial access. For all experiments, bacteria were given 5 min to swim into the microchambers before the sample was illuminated, a procedure that expanded the walls of BSA microchambers and, as a consequence, pinched closed their entry pores. After this initial illumination, samples were washed with fresh TSB to remove cells not captured within structures. In the experiment shown in Figure 4a, the sample was incubated at 37°C after illumination and only removed to image and record the growth of the confined bacterial populations over an 8 h. period. For the experiment using multi-component protein structures designed to “catch-and-release” bacteria shown in Figure 4c, the sample was illuminated for 15 min with a TRITC excitation filter to close the entry pore to the 0.3 pL chamber. In this demonstration, P. aeruginosa cells trapped in microchambers were incubated at ambient temperature for only 1 h. prior to releasing motile cells into the 4.3 pL chamber by illuminating the sample for 15 min through the Texas Red emission filter.

7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Safety considerations. Various chemical used in these studies can cause skin irritation and/or are internally toxic to humans, and standard laboratory protocols were followed to prevent exposure. All toxic chemicals were disposed of using approved protocols at the University of Texas. Laser light used in these experiments can caused damage to eyes and skin, and approved university protocols for avoiding exposure were observed. RESULTS AND DISCUSSION Similar to synthetic polymeric materials that are sensitive to light and various other stimuli,16,17 the degree of photo-induced expansion and contraction of multiphotonfabricated protein-based hydrogels can be tuned by modifying experimental parameters. In the current study, the photoresponsive materials

properties was

found

of to

these strongly

depend on fabrication and illumination protocols.

To

systematically

characterize these properties for a representative protein-based hydrogel type, rectangular test structures were fabricated using a solution of Rose Bengal (RB) and BSA for various protein

and

concentrations (average)

laser

photosensitizer (data

not

powers,

shown), slow-axis

scan frequencies, and optical axis step sizes (Figure 1). Under all conditions, structures

printed

from

RB/BSA

Figure 1. Photo-expansion of RB/BSA hydro-gels. The photoresponse of test structures printed from RB/BSA response is tunable as a function (a) laser power, (b) slow axis scan frequency, (c) optical axis step size, and (d) illumination time. Error bars represent one standard deviation for a minimum of four structures.

solutions were found to undergo swelling when stimulated with broad-band illumination. Test structure swelling ratio (A/Ao) was greatest for conditions resulting in low protein crosslinking — i.e., for relatively low average laser powers (Figure 1a), high scan frequencies (Figure 1b), and large steps between adjacent fabrication layers along the

8


Page 8 of 18

Page 9 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


optical

(z)

axis

(Figure

1c).

Additionally, the swelling ratio was observed to increase throughout a 30-min illumination time, although the expansion

rate

decreased

considerably after ~10 – 15 min (Figure 1d). To investigate the spectral dependence of photoresponse, we examined effects of placing three different

spectral

filters

in

the

illumination pathway (Figure 2, top panel): a 400-nm bandpass filter, a (narrow-band) tetramethylrhodamine isothiocyanate

(TRITC)

excitation

filter, and a 625-nm longpass filter. Essentially no expansion occurred when RB/BSA microstructures were exposed using the bandpass filter or the longpass filter, whereas use of the TRITC filter retained nearly half of the swelling response observed when RB/BSA microstructures were

Figure 2. Influence of spectral isolation on photoresponse of RB/BSA microstructures. Top panel: Swelling ratios of structures illuminated using different spectral filters. Substantial photoexpansion is observed in the absence of a filter (white) and when using a TRITC excitation filter (dark gray), but not when using a 400-nm bandpass filter (light gray) or a 625-nm longpass filter (hatched). Error bars represent the standard deviation for a minimum of four structures. Middle panel: Transmittance spectra of the three filters used in this study. Bottom panel: RB absorption spectrum, which overlaps extensively with the TRITC excitation filter transmittance spectrum.

irradiated with the full spectrum tungsten-halogen output. These results indicate that RB/BSA-based-microstructure photoresponse relies on absorption of residual RB (i.e., dye that is retained within the hydrogel matrix after washing). As shown in Figure 2, middle and lower panels, only the TRITC excitation filter efficiently transmits spectral components within the RB visible absorption range. Protein hydrogels can be µ3D printed using various high-solubility proteins and photosensitizers. We therefore investigated responsiveness of protein-based test structures fabricated for a range of protein-photosensitizer combinations using broad-

9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

band illumination (Figure 3). As with RB/BSA, some reagent combinations yielded structures that undergo photo-expansion. Others, however, responded to illumination with substantial decreases in structure volume. Notably, for all combinations examined here, we found that the charge of the photosensitizer and net charge of the protein (in pH 7.4 HEPES) together were predictive of whether photo-expansion or photocontraction occur. Combinations of a photosensitizer and protein that both carry a net positive or net negative charge respond to light by contracting or expanding, respectively (Figure 3a). If the charge of the photosensitizer is opposite to the net charge of the protein (e.g., MB/BSA and RB/avidin), a condition that may result in more extensive ion pairing and a hydrogel matrix having relatively few charge centers, no significant volume changes are observed in response to illumination. Although RB is negatively charged and myoglobin is nearly neutral at pH 7.4,18 no expansion is observed for this combination in response to illumination, suggesting that the overall charge density of resultant test structures may be relatively low. However, RB/myoglobin structures expand in response to light at pH 9.0, as predicted by the net negative charge of both species at this pH (Supporting Information, Table S1). Also as expected from the results reported in Figure 1, the magnitude of both expansion and contraction is lower for test structures fabricated at higher versus lower laser powers (48 mW vs 33 mW). Light-induced responses can be dramatically modified by tuning the local charge environment of hydrogels by incubating structures in solutions of various charged species. In a series of studies, we observed that the photoresponse of protein hydrogels can be enhanced, ameliorated, reversed, or inhibited by incubating hydrogels in solutions containing charged polyamino acids, proteins, and dyes. For example, hydrogels fabricated from unresponsive photosensitizer-protein combinations (MB/BSA and RB/avidin) can be altered to contract or expand by illuminating microstructures in the presence of polyamino acids carrying large positive (poly-L-lysine, PLL) or negative (poly-L-aspartic acid, PLA) charges, respectively (Figure 3b). Moreover, charge-induced photoresponses can be sequentially modified by multiple washes/illuminations using solutions containing oppositely charged poly-aminoacids. This concept is demonstrated in Figure 3c using MB/BSA-based hydrogels with sequential illuminations in (1) PLL followed by PLA to contract then expand structures

10


Page 10 of 18

Page 11 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(filled squares, bottom axis), and (2) PLA then PLL to expand then contract structures (open circles, top axis). Figure 3d shows general trends for perturbing hydrogels with different types of charged solutes; specific data for multiple proteins and photosensitizer reagents, different charged-species washes and pH values, is provided in Table S1. Collectively, these data suggest that, in many instances, charge state of a hydrogel may provide predictive information regarding the effect of irradiation on hydrogel volume. Additional observations are consistent with a model in which volume changes are mediated primarily by photochemical, rather than photothermal, mechanisms: swelling responses are not diminished when hydrogels are immersed in solutions at temperatures below 5°C, but are inhibited in the presence of a number of reactive oxygen species quenchers (crocin, curcumin, sodium azide, ascorbic acid; Figure S1, Table S2). Overall, our findings are consistent with a hypothesis in which excited-state photosensitizers (and/or their downstream products) cause hydrogel contraction via formation of additional crosslinks within positively charged protein matrix regions, and enable hydrogel expansion through breakage of pre-existing crosslinks within negatively charged protein matrix regions. Both photocleavage and photoformation of bonds have been reported within synthetic polymer networks as a result of complex sets of reactions between photoinitiators and photo-generated acids and bases.19-22 In the current studies, protein photocrosslinking and photocrosslink cleavage may occur at different molecular sites within a given protein hydrogel, with the extent of such opposing reactions varying according to pH and the pKa values of relevant chemical groups, as well as charged solutes that may bind to the hydrogel matrix. Depending on which pathways dominate, different macroscopic behaviors (expansion or contraction) would be observed.

11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. Test structures fabricated using multiple protein/photosensitizer combinations exhibit different responses to broad-band (15 min) illumination. (a) Micro-3D-printed structures fabricated using a negatively charged photosensitizer and protein reagent combination (RB/BSA) expand when illuminated in a pH 7.4 solution, while structures printed using positively charged photosensitizer/protein reagent combinations (MB/Av, MB/Lys) undergo contraction. Structures created using reagent combinations in which aqueous photosensitizer and protein have opposing net charges (MB/BSA, RB/Av, RB/Myo) do not undergo significant volume changes in response to illumination. Light-induced responses were examined for test structures fabricated using two Ti:S fabrication powers, 33 mW and 48 mW. (b) Perturbation of hydrogel charge through addition of polyamino acids can induce photoresponses for relatively neutral structures. MB/BSA (dark gray) and RB/Av (light gray) structures do not undergo light-mediated volume changes at pH 7.4 (left bars). However, both sets of structures contract in the presence of positively charged polymer (PLL, middle bars) and expand in the presence of negatively charged polymer (PLA, right bars) in response to illumination. (c) Initially low-charge MB/BSA structures can contract and then expand (black squares, bottom axis) or expand and then contract (open circles, top axis) as a result of sequential washes with highly charged polyamino acids (0.075% w/vol PLL and 0.5 mg -1 mL PLA) and sequential exposures to light. Error bars represent the standard deviation for a minimum of four structures. (d) Perturbation of hydrogel photoresponse using charged aqueous solutes. This summary table shows trends on the light-responsive behavior of protein-based hydrogels when perturbing microstructure charge via incubation in photosensitizers or polyamino acids, or by changes to the buffer pH. For detailed information, see Table S1.

12


Page 12 of 18

Page 13 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


To exploit hydrogel photoexposure as a means to rationally manipulate cellular microenvironments, we first examined cell tolerance of typical illumination conditions when confined within protein-based microchambers. Structures were printed using RB/BSA, with chamber walls immediately surrounding the entry pore fabricated using a larger plane-to-plane step size than the chamber walls and roof (0.50 µm vs. 0.30 µm). This larger step size resulted in lower hydrogel crosslink density in proximity to the pore and, as predicted in Figure 1c, greater photoresponsiveness (i.e., photo-mediated expansion). The resultant structures were incubated with P. aeruginosa bacteria, a motile environmental pathogen associated with chronic infections, for example, in wounds, cystic fibrosis lungs, and air passageways. After a small number of cells had passed into a picoliter chamber (Figure 4a, upper left panel), structures were exposed to broad-band light to promote expansion of hydrogel material primarily in the region adjacent to the pore. Using this approach, bacteria could be isolated from the bulk population (Figure 4a, upper right panel) and allowed to divide within highly confined spaces. Trapped cells displayed generation times similar to P. aeruginosa within culture flasks not subject to illumination, reaching extremely high densities within µ3D printed structures within hours (Figure 4a, lower panels). At such cell densities, enhanced chemical and physical interactions between microbial cells (including P. aeruginosa and various other pathogenic bacteria) can promote phenotypic adaptations associated with enhanced microbial virulence, such as quorum sensing and population-dependent antibiotic resistance. Although these responses are known to play important roles in pathogenicity, most information regarding onset and termination of group behaviors has been acquired on macroscopic populations, a scale that may not be relevant to transmission and early development of many infections. Our capacity to fabricate µ3D printed materials that respond differentially to a sequence of illumination triggers enables more sophisticated, dynamic control of bacterial confinement within picoliter spaces. Such a capability set could be used, for example, to investigate how changes in cell-cell interaction volumes at specific points during phenotypic transformation may influence phenotypic outcome. To dynamically modify the volume in which small, rapidly growing populations of P. aeruginosa interact, we designed an arrangement comprised of two adjacent

13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

confinement regions, a sub-picoliter antechamber for capturing motile bacteria and “farming” them to high densities, and a much larger (4.3 pL) container into which cells are released following population growth. To provide temporal control over cellular confinement, the antechamber was designed with two photo-switchable pores: (1) An initially open entry pore (for capturing bacteria from bulk solution) composed of material, M1, that expands to close the pore when irradiated with spectral component, λ1; and (2) An initially closed release pore (leading to the larger container) composed of material, M2, that contracts to open the pore when exposed to spectral component, λ2. From materials characterization studies, we initially chose to fabricate M1-based pores using RB/BSA and M2-based pores using MB/Lys, where M1 and M2 ideally could be triggered independently using different spectral components from the microscope illuminator. Figure 4b demonstrates the ability to excite RB and MB with low channel cross-talk using TRITC excitation and Texas Red emission filters, respectively. This potentially straightforward strategy, however, is complicated by the fact that the first material type that is fabricated (either M1 or M2) is necessarily bathed in the reagent used to fabricate the second material type. As suggested from the studies described in Figure 3, such reagent exposure caused prohibitive changes to the photoresponsive behavior of the material that was printed first. Although this effect could not be fully reversed through extensive rinsing with buffer alone, we found that by printing M1 (using RB/BSA) then M2 (using MB/Lys), then rinsing the completed set of structures with a low concentration of RB (1.0 mM for 4 min), we could sufficiently restore M1’s capacity to undergo photoexpansion without eliminating M2’s ability to undergo photocontraction in response to MB excitation. Potential expansion of the M2-based release pore during illumination through the TRITC filter as a result of postfabrication incubation in RB (see, for example, Table S1), was prevented by fabricating M2 using a somewhat higher laser power (43 mW) than was used to fabricate M1 (38 mW).

14


Page 14 of 18

Page 15 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Functional cell enclosures. The light responsive properties of protein hydrogels can be used to corral and release bacterial populations within picoliter volumes. (a) P. aeruginosa cells swim through an open aperture in a RB/BSA structure (top left), and are then trapped by swelling the aperature closed with a 15 min exposure to light (top right). Bacteria divide at a normal rate to fill the cavity over ~8 h (bottom panels). The pink region in the top left panel shows the region of lower-density hydrogel. (b) RB and MB absorption spectra, and transmittance spectra of TRITC excitation and Texas Red emission filters, demonstrate the ability to excite these photosensitizes using distinct spectral components with minimal cross-talk. (c) A hybrid bacterial edifice comprised of a RB/BSA pore (structure bottom) and a MB/Lys pore (paired ovals) demonstrates the ability to sequentially close and open chambers to capture, grow, then dilute P. aeruginosa. Bacteria swim into a small antechamber through Pore 1 (1o) in its open state, and are allowed to accumulate for a short period (left panel). Illumination of the edifice through a TRITC excitation filter causes swelling and transition of Pore 1 to a closed state (1c) while the MB/lys aperture, Pore 2, remains in a closed state (2c), thereby restricting cells to a 0.3 pL chamber (middle panel). Following division of a period of 1 h., cells are released into a 4.3 pL chamber by illuminating the sample through a Texas Red emission filter to open the MB/Lys pore (2o; right panel). Bacteria in all images are false-colored green for visualization. Scale bars, 10 µm.

Figure 4c demonstrates use of these materials to sequentially manipulate the interaction volume of P. aeruginosa cells on demand within a multi-chambered edifice. Here, the outer walls of both chambers, including the antechamber entry pore (Pore 1), as well as the roof and internal support pillars in the larger chamber, were µ-3D printed as a monolithic structure composed of M1, while the release pore (Pore 2) was fabricated in a second step as two M2 oval columns attached to the internal walls of the M1 monolith. Following fabrication, initial rinsing (which results in a small increase in

15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 18

hydrogel dimensions), and post-fabrication treatment with RB, Pore 1 was in an open state (“1o”) while the release pore was closed to bacterial transit (“2c”) – a configuration that allowed P. aeruginosa cells to enter and collect in the 0.3-pL antechamber at the lower portion of the image (Figure 4c, left panel). Following accumulation of a small number of cells, the edifice was illuminated for 15 min using the TRITC excitation filter, causing the RB/BSA-based material surrounding the entry pore to expand and pinch closed (“1c”) without modifying the release pore (still “2c”), thereby isolating the cells in the antechamber (Figure 4c, middle panel). Cell division was allowed to proceed for ~1 h. in this chamber to create a dense bacterial microcolony; the structure was then exposed for an additional 15 min using the Texas Red emission filter, causing Pore 2 to open (“2o”) via selective contraction of M2, thereby releasing motile cells into the 4.3 pL chamber (Figure 4c, right panel). CONCLUSIONS Various properties of µ3D printed protein hydrogels provide advantages in micromechanical and cellular confinement studies, including biocompatibility, tunable stiffness, and, of particular relevance to the current work, permeability to aqueous solutes

and

environmental

responsiveness.

Although

environmentally

sensitive

3D-defined materials previously have been developed that undergo controllable size and/or shape changes, the environmental cues used to trigger such responses often are damaging or otherwise disruptive to cellular systems. In the current studies, we demonstrate capabilities for tailoring the properties and exposure protocols of µ3D printed materials to achieve tunable photoresponsive behaviors for remotely manipulating, in a minimally invasive manner, living cells within defined picoscopic cell culture volumes. Rational design of complex, multi-component microfabricated environments comprised of such protein-based smart materials offers exciting opportunites for investigating a range of cell-cell interactions. ACKNOWLEDGMENTS. We gratefully acknowledge support for these studies by the Welch Foundation (Grant F-1331) and the Army Research Office (Grant W911NF-13-1-0199).

REFERENCES

16


Page 17 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

Barbulovic-Nad, I.; Yang, H.; Park, P.S.; Wheeler, A.R. Lab Chip 2008, 8, 519– 526. Gach, P.C., Attayek, P.J., Whittlesey, R.L., Yeh, J.J., Allbritton, N.L. Biosensors & Bioelectronics 2014, 54, 476–483. Pei, J.; Nie, J.; Kennedy, R.T. Anal. Chem. 2010, 82, 9261–9267. Croushore, C.A.; Sweedler, J.V. Lab Chip 2013,13, 1666-1676. Denk, W.; Strickler, J.H.; Webb, W.W. Science 1990, 248, 73–76. Spivey, E.C.; Ritschdorff, E.T.; Connell, J.L.; McLennon, C.A.; Schmidt, C.E.; Shear, J.B. Adv. Funct. Mat, 2013, 23, 333–339. LaFratta, C.N.; Li, L.; Fourkas, J.T. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 8589–8594. Kawata, S., Sun, H.-B., Tanaka, T. & Takada, K. Nature 2001, 412, 697–698. Pitts, J.D.; Campagnola, P.J.; Epling, G.A.; Goodman, S.L. Macromolec. 2000, 33, 1514–1523. Kaehr, B.; Shear, J.B. J. Am. Chem. Soc. 2007 129,1904–1905. Connell, J.L.; Ritschdorff, E.T.; Whiteley, M.; Shear, J.B. Proc. Natl. Acad. Sci. USA 2013, 110, 18380–18385. Kaehr, B.; Shear, J.B. Proc. Natl. Acad. Sci. USA 2008, 105, 8850–8854. Connell, J.L.; Wessel, A.K.; Parsek, M.R.; Ellington, A.D.; Whiteley, M.; Shear, J.B. mBio 2010, 1, e00202-e00212. Nielson, R; Kaehr, B.; Shear, J.B. Small 2009, 5, 120–125. Parsek, M.R.; Greenberg, E.P. Trends Microbiol. 2005, 13, 27–33. Watanabe, T.; Akiyama, M.; Totani, K.; Kuebler, S.M.; Stellacci, F.; Wenseleers, W.; Braun, K.; Marder, S.R.; Perry, J.W. Adv. Funct. Mat. 2002 1, 611–614. Lendlein, A.; Shastri, V.P. Adv. Mater. 2010, 22, 3344–3347. Malamud, D.; Drysdale, J.W. Anal. Biochem. 1978, 86, 620–647. Fouassier, J.P. et al. Materials 2010, 3, 5130–5142. Moon, S.Y.; Kim, J.M. J. Photochem. Photobiol. C: Photochem. Rev. 2007, 8, 157–173. Mallick, A.; Purkayastha, P.; Chattopadhyay, N. J. Photochem. Photobiol. C: Photochem. Rev. 2007, 8, 109–127. Suyama, K.; Shirai, M. Prog. Polymer Sci. 2009, 34, 194–209.

Notes. The authors declare no competing financial interest. Supporting Information. Table S1: Detailed effects of charged solutes on protein hydrogel photoresponsive properties. Figure S1: Structures of reactive oxygen species (ROS) quenchers. Table S2: Effects of ROS quenchers on RB/BSA structure expansion.

For TOC only

17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

18


Page 18 of 18

Three-Dimensional Printing of Photoresponsive ... - ACS Publications

Recommend Documents