- Email: [email protected]
Axonal outgrowth of hippocampal neurons on micro-scale networks of polylysine-conjugated laminin
Biomaterials 22 (2001) 1049}1054
Axonal outgrowth of hippocampal neurons on micro-scale networks of polylysine-conjugated laminin L. Kam , W. Shain, J.N. Turner, R. Bizios * Department of Biomedical Engineering, Rensselaer Polytechnic Institute, 110, 8th Street, Troy, NY 12180, USA Laboratory of Nervous System Disorders, Wadsworth Center, Albany, NY 12201, USA Received 10 April 2000; accepted 7 September 2000
Abstract Microcontact printing was used to de"ne an interconnected lattice network of polylysine-conjugated laminin, a protein-polypeptide ligate that is an e!ective promoter of neuron outgrowth on material surfaces. In the presence of serum proteins, rat hippocampal neurons selectively adhered to features of polylysine-conjugated laminin as narrow as 2.6 m in width. Adhering neurons extended long axonal processes, which precisely followed and did not deviate from the prescribed patterns, demonstrating that neurons respond to this protein with high selectivity and that these techniques e!ectively provide long-range guidance of axonal outgrowth. Further examination of neuron response under serum-free cell culture conditions demonstrated that the outgrowth-promoting activity of polylysine-conjugated laminin was attributed to biologically active laminin. Together, these results demonstrate that polylysineconjugated laminin provides for high-precision guidance of neuron attachment and axon outgrowth on material surfaces in a serum-independent manner. This ability to guide hippocampal neuron response in low-density, serum-free culture with high precision is valuable for the development of advanced, neuron-based devices. 2001 Elsevier Science Ltd. All rights reserved. Keywords: Surface micropatterning; Neuron; Axonal outgrowth; Laminin; Polylysine
1. Introduction The ability to spatially control and direct the interaction of neurons with biomaterial surfaces is critical to both investigation of basic neuron cellular function and design of advanced bio-electronic devices, such as biosensors and neuronal networks. To date, both chemical and topological micropatterning of material substrates have been utilized to control neuron response [1}10]. Techniques based on micropatterning of biologically important proteins (e.g., laminin and "bronectin) are of particular interest because these biomolecules (unlike purely adhesive compounds such as polylysine) provide cues for guiding neuron response duplicate those encountered by neurons in vivo [11}14]. Recently, chemical ligation of bioactive proteins to adhesive polypeptides such as polylysine, prior to deposition on a surface by microcontact printing has been introduced as an e!ective technique for patterning proteins onto material surfaces
[15]; this initial study demonstrated enhanced outgrowth of hippocampal neurons on a simple pattern of parallel, 20-m-wide lines of polylysine-conjugated laminin. The present study examined the use of polylysine-conjugated laminin to guide outgrowth of hippocampal neurons along interconnected networks of narrow (micrometer-scale), linear features of polylysine-conjugated laminin de"ned on glass. By examining axon outgrowth of hippocampal neurons along these biomolecular patterns as a function of both pattern dimension and well as the role of serum proteins in mediating this interaction, this study identi"ed new criteria pertinent to the use of polylysine-conjugated laminin and/or other bioactive compounds to precisely guide and control neuron interaction with biomaterial surfaces. 2. Materials and methods 2.1. Polylysine-conjugated laminin
* Corresponding author. Tel.: #1-518-276-6964; fax: #1-518-2763035. E-mail address: [email protected] (R. Bizios).
Laminin was covalently ligated with polylysine using previously described techniques [15]. Brie#y, a solution
0142-9612/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 0 0 ) 0 0 3 5 2 - 5
1050
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
containing 100 g/ml laminin (EHS-mouse laminin; Boehringer Mannheim, Indianapolis, IN), 50 g/ml polylysine (33.5 kDa molecular weight; Sigma, St. Louis, MO), and 150 g/ml 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (Sigma) in phosphate-bu!ered saline (PBS, pH 7.3) was sterile "ltered then incubated at room temperature overnight. The resultant polylysineconjugated laminin in this solution was concentrated by ultra"ltration (10 kDa molecular weight cut-o! Centricon unit; Millipore, Bedford, MA) and used within hours after preparation for modi"cation of material surfaces. 2.2. Substrate preparation Micro-scale chemical patterns of either polylysine-conjugated laminin or polylysine were de"ned on borosili-
cate glass substrates using established microcontact printing techniques [15}17]; this process is outlined schematically in Fig. 1A. Brie#y, borosilicate coverslips (18mm in diameter; Fisher Scienti"c, Pittsburgh, PA) were etched in nitric acid, rinsed in deionized water, then dried overnight at 1103C. Polydimethylsiloxane elastomer (Sylgard 184; Dow Corning, Midland, MI) stamps, each containing a topological representation of a micro-scale pattern were exposed to an air plasma using a plasma sterilization unit (Harrick Scienti"c, Ossining, NY) for 15 s. The geometry of this pattern was described previously [18], and consisted of a series of linear segments arranged in an interconnected network (Fig. 1B). At every intersection, three linear segments converged with an angle of 1203 between adjacent segments, thus forming a hexagonal lattice. Each pattern contained identically sized segments measuring either 13, 22, or 43 m in length (dl , Fig. 1B) and 2.6}17 m in width (d , Fig. 1B). The plasma-treated stamps were coated with 250 g/ml of either polylysine or polylysine-conjugated laminin in aqueous bu!er (either 0.1 M borate bu!er, pH 8.5, for polylysine or PBS for polylysine-conjugated laminin) supplemented with 15% dimethylsulfoxide for 15 min, dried using a stream of argon gas, and placed in direct contact with acid-cleaned coverslips. A 60-g weight was placed on top of each stamp. After 20 min, the coverslips were carefully separated from the stamps and rinsed extensively in a HEPES-bu!ered (pH 7.3) Hank's saline solution (HBHS). The resultant, micropatterned substrates were sterilized by immersion into 10 g/ml of gentamicin in HBHS for 2 h. For experiments carried out in the presence of serum, these sterilized substrates were maintained in minimum essential medium (MEM) supplemented with 10% horse serum under standard cell culture conditions (a humidi"ed, 5% CO /95% air envi ronment, maintained at 373C) overnight before use in cell experiments. For serum-free experiments, the substrates were maintained in serum-free MEM. 2.3. Cell culture * assay of neuron attachment and axonal outgrowth
Fig. 1. Microcontact printing of biomolecules. (A) Schematic of the microcontact printing process used to create micro-scale patterns of polylysine and polylysine-conjugated laminin onto glass substrates. (B) Representative #uorescence micrograph illustrating the geometry and dimensions of the micro-scale patterns examined in the present study. Two geometric parameters that described each pattern, namely, the width (d ) and length (dl ) of the linear segments, are indicated in this image. Substrate-bound polylysine-conjugated laminin was immunofluorescently labeled with Alexa-488 and visualized as described in Section 2.3. Scale bar"50 m.
Rat hippocampal neurons were isolated and cultured under semi-serum-free conditions following established techniques [19]. Brie#y, hippocampi from embryonic (E18) Sprague}Dawley rats (Taconic Farms, Germanstown, NY) were dissected, pooled, and digested using a 0.25% trypsin solution. Neurons were released from this tissue by trituration, seeded (2.6;10 cells/cm) onto sterilized, patterned surfaces (prepared as described in Section 2.2) in MEM supplemented with 10% horse serum and allowed to adhere under standard cell culture conditions for 2 h. The coverslips and adherent neurons were then inverted and transferred to tissue culture dishes containing established cultures of astroglial cells in MEM supplemented with N2 components; para$n
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
dots, placed on the inverted coverslips prior to use in cell experiments, were used to maintain a small (1}2 mm) distance between the neurons and astrocytes. These neuron}astrocyte co-cultures were maintained under standard cell culture conditions for either 1 or 3 days for experiments carried out in the presence and absence of serum, respectively. At those times, neurons adherent to the micropatterned surfaces were "xed using warm (373C), 4% paraformaldehyde in calcium- and magnesium-free HBHS. Fixed, adherent neurons were #uorescently stained in situ, at 373C, using two di!erent techniques. For experiments carried out in the presence of serum, micropatterned laminin and adherent neurons were treated with 1% Triton X-100, blocked using 5% bovine serum albumin and then incubated with both a primary antibody to EHS-mouse laminin (rabbit anti-laminin, 1 : 500 dilution; Sigma) and a primary antibody to a neuron-speci"c chain of tubulin (clone TUJ1, 1 : 500 dilution; BAbCO, Richmond, CA). These immunochemically labeled samples were then incubated with both a biotin-conjugated antibody to mouse IgG (1 : 200 dilution; Sigma) and an Alexa-488-conjugated antibody to rabbit IgG (1 : 200 dilution; Molecular Probes) followed by streptavidin-conjugated Quantum Red2+ (1 : 50 dilution; Sigma). For serum-free experiments, adherent neurons were stained using 0.5 g/ml diIC (3) (Molecular Probes) in HEPES-bu!ered Hank's saline (HBHS) for 2 h. These diIC (3)-labeled substrates were dehydrated using graded (35, 50, 70, 95% v/v) solutions of ethanol then rehydrated in preparation for visualization by #uorescence microscopy.
1051
All #uorescently labeled samples were visualized by confocal microscopy; light of 488-nm (provided by an argon ion laser) was used to simultaneously excite the Alexa-488 and Quantum Red2+ #uorophores. Appropriate "lters were used to selectively isolate the signal emitted from these two #uorophores [16]. Light of 514-nm wavelength (provided by an argon ion laser) was used to excite the diIC (3) #uorophore; a 540-nm long-pass barrier "lter was used to selectively detect light emitted from this dye. Axons were identi"ed on the basis of morphology. Speci"cally, if the total length of the longest process elaborated by a neuron was greater than 50 m, this process was considered to be the axon [19]. From multiple images collected at random locations on select substrates, the number of neurons adherent to the patterns of polylysine-conjugated laminin and to the intervening spaces were counted separately.
3. Results After 1 day of culture (as detailed in Section 2.3), rat hippocampal neurons that were seeded in serumsupplemented media onto substrates modi"ed with polylysine-conjugated laminin selectively adhered to the biomolecular pattern (Fig. 2). As an example of this selectivity, over 95% of the neurons seeded onto surfaces modi"ed with 43-m long, 2.3-m wide segments of polylysine-conjugated laminin adhered to the biomolecular pattern. Neurons adhering to these patterns exhibited extensive axonal outgrowth, evidenced by the presence of
Fig. 2. Directed neuronal attachment and axonal outgrowth. After 1 day, rat hippocampal neurons seeded onto patterned substrates in the presence of 10% (v/v) serum selectively attached to, and extended axonal processes along, patterns of polylysine-conjugated laminin. The representative #uorescence micrographs of this "gure illustrate neuron attachment and outgrowth along patterns composed of linear segments with the following dimensions: (A) d "7.5 m, dl "13 m; (B) d "6.7 m, dl "43 m; and (C) d "2.6 m, dl "43 m. Substrate-bound laminin was immuno #uorescently labeled with Alexa-488, and appears red in this "gure. Neurons were identi"ed by immunochemical staining for a neuron-speci"c subunit of tubulin with Quantum Red, and are shown in green-yellow. These three images are shown at identical magni"cations. Scale bar"50 m.
1052
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
axonal processes measuring at least 150 m in length; moreover, these axonal processes were associated exclusively with the patterns of polylysine-conjugated laminin. Minor neurites (i.e., short, non-axonal processes) elaborated by adherent neurons were similarly selectively associated with regions of polylysine-conjugated laminin, and branched away from the cell soma along separate segments of the micro-scale patterns. The few adhering neurons on intervening regions of unmodi"ed glass remained rounded and did not elaborate cellular processes. Axonal outgrowth on these polylysine-conjugated laminin-modi"ed surfaces was qualitatively in#uenced by pattern dimension. The morphology of neurons adherent to patterns composed of short, wide segments resembled that of neurons in routine culture on unpatterned, celladhesive surfaces; this morphology is illustrated in Fig. 2A, which shows neuron outgrowth on a pattern of segments measuring 8.3 m in width (d ) and 13 m in length (dl ). On these surfaces, the axonal process elaborated by adhering neurons traversed the network of polylysine-conjugated laminin, meandering across each micropatterned segment, and avoided the small regions of unmodi"ed glass. In contrast, longer, narrower segments of polylysine-conjugated laminin provided guidance for both axonal and minor processes elaborated by adhering neurons; Fig. 2B illustrates this response on a pattern containing segments measuring 5.6 m in width (d ) and 43 m in length (dl ). As the pattern width was further reduced, the path of axonal growth was controlled with increasing precision. This response is illustrated in Fig. 2C, which illustrates a neuron adhering to a pattern containing segments measuring 2.6 m in width (d ) and 43 m in length (dl ); the axonal process remained con"ned to the prescribed pattern, including bending where multiple linear segments intersected. Directed neuron attachment and axon outgrowth on polylysine-conjugated laminin was in#uenced by, but not dependent upon, the presence of serum proteins. Under completely serum-free conditions (that is, both the substrates and the neurons were not exposed to serum either before, during, or after cell seeding), adhering neurons extended axonal processes along the patterns of polylysine-conjugated laminin (Fig. 3A). Axonal outgrowth was, however, slower than that observed in the presence of serum; in fact, neurons required several days to exhibit extensive axonal outgrowth in the absence of serum proteins (Fig. 3A). In addition, neuron attachment onto surfaces modi"ed with polylysine-conjugated laminin was less selective in the absence than in the presence of serum proteins. Occasionally, neurons adhered to regions of plain glass (which were delineated and surrounded by the patterns of polylysine-conjugated laminin) in the absence of serum (arrowheads, Fig. 3A); these cells remained rounded,
Fig. 3. Bioactivity of microcontact-printed polylysine-conjugated laminin. After 3 days of culture under serum-free conditions, hippocampal neurons selectively extended axons along patterns of polylysineconjugated laminin (A), but not on patterns of polylysine alone (B). The arrows in frame A assist in identifying the axonal process in this image. Neurons that adhered onto either plain glass or polylysine exhibited a rounded morphology, and did not elaborate cellular processes; examples of these cells are indicated by the arrowheads in frames A and B. Adherent neurons in these representative #uorescence micrographs were labeled with diIC (3). This staining procedure also resulted in labeling of the substrate-bound polylysine and polylysine-conjugated laminin, allowing concurrent visualization of the biomolecular pattern in the background. Both frames are presented at identical magni"cations. Scale bar"50 m.
and did not elaborate any processes. On surfaces patterned with polylysine alone, neurons adhered to both regions of plain glass and of polylysine (Fig. 3B); these adherent neurons remained rounded (arrowheads, Fig. 3B).
4. Discussion Microcontact printing of polylysine-conjugated laminin onto material surfaces illustrates a versatile, robust technique for controlling neuron attachment and
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
outgrowth. In particular, the present study demonstrated that patterns of polylysine-conjugated laminin can provide micrometer-scale guidance of axon outgrowth over distances of several hundred micrometers (Fig. 2C). Control over neuron responses was achieved using micrometer-scale patterns of cell-adhesive compounds (most notably polylysine [3,5,9,10] and aminosilanes [1]) on biomaterial surfaces. Attachment of rat hippocampal neurons to micropatterned polylysine alone was, however, dependent on adsorbed (but, to date, not speci"cally identi"ed) serum proteins (Fig. 3B); similar observations of serum-dependent attachment have been reported for other types of neurons [1]. In contrast, the present study demonstrated that neuron attachment and axon outgrowth on polylysine-conjugated laminin is not dependent on adsorption of serum proteins (Fig. 3A), indicating that substrate-bound laminin, in the form of polylysineconjugated laminin, is biologically active. Polylysineconjugated laminin may be useful in guiding neuron growth in low-density, serum-free cell cultures, a situation in which exogenous adhesion factors are minimal and which is ideal for studying and utilizing neuron networks. It should be noted, however, that in the absence of serum proteins in the present study, neurons attach to regions of unmodi"ed glass (Fig. 3B). This result is in sharp contrast to that obtained in the presence of serum (Fig. 2), but is in agreement with earlier reports [1] that adsorption of certain, still unidenti"ed, serum components onto material surfaces minimizes neuron attachment. Identi"cation of these serum components, which likely include serum albumin, could be used to enhance selectivity of neuron attachment on micropatterned material surfaces. An important advantage of microcontact printing techniques [20,21] is their additive nature, which allows development of spatially complex surfaces containing multiple patterns of di!erentially bioactive compounds [15,22]. Techniques for directing neuron attachment and axon outgrowth using micropatterned laminin, such as UV ablation [12,14] and selective adsorption [13], are not suited for producing multifunctionalized systems. Spatially and chemically complex systems micropatterned on substrates could provide valuable tools for the next generation of studies of neuron function. For example, investigation of how biological and/or electrical stimuli either along segments or across intersections of micro-scale patterns in#uence axon outgrowth, a presumably random process in the present study, could provide new information regarding the molecular mechanisms that direct neuron path-"nding both in vitro and in vivo. Such developments could provide important biological and materials engineering information valuable to the design of advanced, neuron-based devices.
1053
Acknowledgements This work was sponsored in part by NIH NCRRR01-RR-10957, DARPA/ITO (to J.N. Turner and W. Shain) and by a special gift H71001 (to R. Bizios).
References [1] Kleinfeld D, Kahler KH, Hockberger PE. Controlled outgrowth of dissociated neurons on patterned substrates. J Neurosci 1988;8: 4098}120. [2] Clark P, Connolly P, Curtis ASG, Dow JAT, Wilkinson CDW. Topographical control of cell behavior. II: multiple grooved substrata. Development 1990;108:635}44. [3] Corey JM, Wheeler BC, Brewer GJ. Compliance of hippocampal neurons to patterned substrate networks. J Neurosci Res 1991;30: 300}7. [4] Nagata I, Kawana A, Nakatsuji N. Perpendicular contact guidance of CNS neuroblasts on arti"cial microstructures. Development 1993;117:401}8. [5] Corey JM, Brunette AL, Chen MS, Weyhenmeyer JA, Brewer GJ, Wheeler BC. Di!erentiated B104 neuroblastoma cells are a highresolution assay for micropatterned substrates. J Neurosci Meth 1997;75:91}7. [6] Branch DW, Corey JM, Weyhenmeyer JA, Brewer GJ, Wheeler BC. Microstamp patterns of biomolecules for high-resolution neuronal networks. Med Biol Eng Comp 1998;36:135}41. [7] Saneinejad S, Shoichet MS. Patterned Glass surfaces direct cell adhesion and process outgrowth of primary neurons of the central nervous system. J Biomed Mater Res 1998;42:13}9. [8] Stenger DA, Hickman JJ, Bateman KE, et al. Microlithographic determination of axonal/dendritic polarity in cultured hippocampal neurons. J Neurosci Meth 1998;82:167}73. [9] James CD, Davis RC, Meyer M, et al. Aligned microcontact printing of micrometer scale polylysine structures for controlled growth of cultured neurons on planar microelectrode arrays. IEEE Trans Biomed 2000;47:17}21. [10] Branch DW, Wheeler BC, Brewer GJ, Leckband DE. Long-term maintenance of patterns of hippocmpal pyramidal cells on substrates of polyethylene glycol and microstamped polylysine. IEEE Trans Biomed 2000;47:290}300. [11] Hammarback JA, Letourneau PC. Neurite extension across regions of low cell-substratum adhesivity: implications for the guidepost hypothesis of axonal path"nding. Dev Biol 1986;117:655}62. [12] Hammarback JA, McCarthy JB, Palm SL, Furcht LT, Letourneau PC. Growth cone guidance by substrate-bound laminin pathways is correlated with neuron-to-pathway adhesivity. Dev Biol 1988;126:29}39. [13] Clark P, Britland S, Connolly P. Growth cone guidance and neuron morphology on micropatterned laminin surfaces. J Cell Sci 1993;105:203}12. [14] Tai HC, Buettner HM. Neurite outgrowth and growth cone morphology on micropatterned surfaces. Biotechnol Prog 1998; 14:364}70. [15] Kam L, James CD, Withers G, et al. Neuron attachment and outgrowth on microcontact-printed polylysine-conjugated laminin. J Neurosci Meth 2001, in press. [16] Kam L. Modulation on neuron and astroglial cell function by micropatterning and immobilization of select biomolecules on biomaterial surfaces. Doctoral dissertation, Rensselaer Polytechnic Institute, Troy, NY, 1999. [17] James CD, Davis RC, Kam L, et al. Patterned protein layers on solid substrates by thin stamp microcontact printing. Langmuir 1998;14:741}4.
1054
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
[18] Kam L, Shain W, Turner JN, Bizios R. Correlation of astroglial cell function on micro-patterned surfaces with speci"c geometric parameters. Biomaterials 1999;20:2343}50. [19] Banker G, Goslin K, editors. Culturing nerve cells. Cambridge: The MIT Press, 1991. p. 251. [20] Singhvi R, Kumar A, Lopez GP, et al. Engineering cell shape and function. Science 1994;264:696}8.
[21] Bernard A, Delamarche E, Schmid H, Michel B, Bosshard HR, Biebuyck H. Printing patterns of proteins. Langmuir 1998;14: 2225}9. [22] Wheeler BC, Corey JM, Brewer GJ, Branch DW. Microcontact printing for precise control of nerve cell growth in culture. J Biomech Engng 1999;121:73}8.
Axonal outgrowth of hippocampal neurons on micro-scale networks of polylysine-conjugated laminin L. Kam , W. Shain, J.N. Turner, R. Bizios * Department of Biomedical Engineering, Rensselaer Polytechnic Institute, 110, 8th Street, Troy, NY 12180, USA Laboratory of Nervous System Disorders, Wadsworth Center, Albany, NY 12201, USA Received 10 April 2000; accepted 7 September 2000
Abstract Microcontact printing was used to de"ne an interconnected lattice network of polylysine-conjugated laminin, a protein-polypeptide ligate that is an e!ective promoter of neuron outgrowth on material surfaces. In the presence of serum proteins, rat hippocampal neurons selectively adhered to features of polylysine-conjugated laminin as narrow as 2.6 m in width. Adhering neurons extended long axonal processes, which precisely followed and did not deviate from the prescribed patterns, demonstrating that neurons respond to this protein with high selectivity and that these techniques e!ectively provide long-range guidance of axonal outgrowth. Further examination of neuron response under serum-free cell culture conditions demonstrated that the outgrowth-promoting activity of polylysine-conjugated laminin was attributed to biologically active laminin. Together, these results demonstrate that polylysineconjugated laminin provides for high-precision guidance of neuron attachment and axon outgrowth on material surfaces in a serum-independent manner. This ability to guide hippocampal neuron response in low-density, serum-free culture with high precision is valuable for the development of advanced, neuron-based devices. 2001 Elsevier Science Ltd. All rights reserved. Keywords: Surface micropatterning; Neuron; Axonal outgrowth; Laminin; Polylysine
1. Introduction The ability to spatially control and direct the interaction of neurons with biomaterial surfaces is critical to both investigation of basic neuron cellular function and design of advanced bio-electronic devices, such as biosensors and neuronal networks. To date, both chemical and topological micropatterning of material substrates have been utilized to control neuron response [1}10]. Techniques based on micropatterning of biologically important proteins (e.g., laminin and "bronectin) are of particular interest because these biomolecules (unlike purely adhesive compounds such as polylysine) provide cues for guiding neuron response duplicate those encountered by neurons in vivo [11}14]. Recently, chemical ligation of bioactive proteins to adhesive polypeptides such as polylysine, prior to deposition on a surface by microcontact printing has been introduced as an e!ective technique for patterning proteins onto material surfaces
[15]; this initial study demonstrated enhanced outgrowth of hippocampal neurons on a simple pattern of parallel, 20-m-wide lines of polylysine-conjugated laminin. The present study examined the use of polylysine-conjugated laminin to guide outgrowth of hippocampal neurons along interconnected networks of narrow (micrometer-scale), linear features of polylysine-conjugated laminin de"ned on glass. By examining axon outgrowth of hippocampal neurons along these biomolecular patterns as a function of both pattern dimension and well as the role of serum proteins in mediating this interaction, this study identi"ed new criteria pertinent to the use of polylysine-conjugated laminin and/or other bioactive compounds to precisely guide and control neuron interaction with biomaterial surfaces. 2. Materials and methods 2.1. Polylysine-conjugated laminin
* Corresponding author. Tel.: #1-518-276-6964; fax: #1-518-2763035. E-mail address: [email protected] (R. Bizios).
Laminin was covalently ligated with polylysine using previously described techniques [15]. Brie#y, a solution
0142-9612/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 0 0 ) 0 0 3 5 2 - 5
1050
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
containing 100 g/ml laminin (EHS-mouse laminin; Boehringer Mannheim, Indianapolis, IN), 50 g/ml polylysine (33.5 kDa molecular weight; Sigma, St. Louis, MO), and 150 g/ml 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (Sigma) in phosphate-bu!ered saline (PBS, pH 7.3) was sterile "ltered then incubated at room temperature overnight. The resultant polylysineconjugated laminin in this solution was concentrated by ultra"ltration (10 kDa molecular weight cut-o! Centricon unit; Millipore, Bedford, MA) and used within hours after preparation for modi"cation of material surfaces. 2.2. Substrate preparation Micro-scale chemical patterns of either polylysine-conjugated laminin or polylysine were de"ned on borosili-
cate glass substrates using established microcontact printing techniques [15}17]; this process is outlined schematically in Fig. 1A. Brie#y, borosilicate coverslips (18mm in diameter; Fisher Scienti"c, Pittsburgh, PA) were etched in nitric acid, rinsed in deionized water, then dried overnight at 1103C. Polydimethylsiloxane elastomer (Sylgard 184; Dow Corning, Midland, MI) stamps, each containing a topological representation of a micro-scale pattern were exposed to an air plasma using a plasma sterilization unit (Harrick Scienti"c, Ossining, NY) for 15 s. The geometry of this pattern was described previously [18], and consisted of a series of linear segments arranged in an interconnected network (Fig. 1B). At every intersection, three linear segments converged with an angle of 1203 between adjacent segments, thus forming a hexagonal lattice. Each pattern contained identically sized segments measuring either 13, 22, or 43 m in length (dl , Fig. 1B) and 2.6}17 m in width (d , Fig. 1B). The plasma-treated stamps were coated with 250 g/ml of either polylysine or polylysine-conjugated laminin in aqueous bu!er (either 0.1 M borate bu!er, pH 8.5, for polylysine or PBS for polylysine-conjugated laminin) supplemented with 15% dimethylsulfoxide for 15 min, dried using a stream of argon gas, and placed in direct contact with acid-cleaned coverslips. A 60-g weight was placed on top of each stamp. After 20 min, the coverslips were carefully separated from the stamps and rinsed extensively in a HEPES-bu!ered (pH 7.3) Hank's saline solution (HBHS). The resultant, micropatterned substrates were sterilized by immersion into 10 g/ml of gentamicin in HBHS for 2 h. For experiments carried out in the presence of serum, these sterilized substrates were maintained in minimum essential medium (MEM) supplemented with 10% horse serum under standard cell culture conditions (a humidi"ed, 5% CO /95% air envi ronment, maintained at 373C) overnight before use in cell experiments. For serum-free experiments, the substrates were maintained in serum-free MEM. 2.3. Cell culture * assay of neuron attachment and axonal outgrowth
Fig. 1. Microcontact printing of biomolecules. (A) Schematic of the microcontact printing process used to create micro-scale patterns of polylysine and polylysine-conjugated laminin onto glass substrates. (B) Representative #uorescence micrograph illustrating the geometry and dimensions of the micro-scale patterns examined in the present study. Two geometric parameters that described each pattern, namely, the width (d ) and length (dl ) of the linear segments, are indicated in this image. Substrate-bound polylysine-conjugated laminin was immunofluorescently labeled with Alexa-488 and visualized as described in Section 2.3. Scale bar"50 m.
Rat hippocampal neurons were isolated and cultured under semi-serum-free conditions following established techniques [19]. Brie#y, hippocampi from embryonic (E18) Sprague}Dawley rats (Taconic Farms, Germanstown, NY) were dissected, pooled, and digested using a 0.25% trypsin solution. Neurons were released from this tissue by trituration, seeded (2.6;10 cells/cm) onto sterilized, patterned surfaces (prepared as described in Section 2.2) in MEM supplemented with 10% horse serum and allowed to adhere under standard cell culture conditions for 2 h. The coverslips and adherent neurons were then inverted and transferred to tissue culture dishes containing established cultures of astroglial cells in MEM supplemented with N2 components; para$n
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
dots, placed on the inverted coverslips prior to use in cell experiments, were used to maintain a small (1}2 mm) distance between the neurons and astrocytes. These neuron}astrocyte co-cultures were maintained under standard cell culture conditions for either 1 or 3 days for experiments carried out in the presence and absence of serum, respectively. At those times, neurons adherent to the micropatterned surfaces were "xed using warm (373C), 4% paraformaldehyde in calcium- and magnesium-free HBHS. Fixed, adherent neurons were #uorescently stained in situ, at 373C, using two di!erent techniques. For experiments carried out in the presence of serum, micropatterned laminin and adherent neurons were treated with 1% Triton X-100, blocked using 5% bovine serum albumin and then incubated with both a primary antibody to EHS-mouse laminin (rabbit anti-laminin, 1 : 500 dilution; Sigma) and a primary antibody to a neuron-speci"c chain of tubulin (clone TUJ1, 1 : 500 dilution; BAbCO, Richmond, CA). These immunochemically labeled samples were then incubated with both a biotin-conjugated antibody to mouse IgG (1 : 200 dilution; Sigma) and an Alexa-488-conjugated antibody to rabbit IgG (1 : 200 dilution; Molecular Probes) followed by streptavidin-conjugated Quantum Red2+ (1 : 50 dilution; Sigma). For serum-free experiments, adherent neurons were stained using 0.5 g/ml diIC (3) (Molecular Probes) in HEPES-bu!ered Hank's saline (HBHS) for 2 h. These diIC (3)-labeled substrates were dehydrated using graded (35, 50, 70, 95% v/v) solutions of ethanol then rehydrated in preparation for visualization by #uorescence microscopy.
1051
All #uorescently labeled samples were visualized by confocal microscopy; light of 488-nm (provided by an argon ion laser) was used to simultaneously excite the Alexa-488 and Quantum Red2+ #uorophores. Appropriate "lters were used to selectively isolate the signal emitted from these two #uorophores [16]. Light of 514-nm wavelength (provided by an argon ion laser) was used to excite the diIC (3) #uorophore; a 540-nm long-pass barrier "lter was used to selectively detect light emitted from this dye. Axons were identi"ed on the basis of morphology. Speci"cally, if the total length of the longest process elaborated by a neuron was greater than 50 m, this process was considered to be the axon [19]. From multiple images collected at random locations on select substrates, the number of neurons adherent to the patterns of polylysine-conjugated laminin and to the intervening spaces were counted separately.
3. Results After 1 day of culture (as detailed in Section 2.3), rat hippocampal neurons that were seeded in serumsupplemented media onto substrates modi"ed with polylysine-conjugated laminin selectively adhered to the biomolecular pattern (Fig. 2). As an example of this selectivity, over 95% of the neurons seeded onto surfaces modi"ed with 43-m long, 2.3-m wide segments of polylysine-conjugated laminin adhered to the biomolecular pattern. Neurons adhering to these patterns exhibited extensive axonal outgrowth, evidenced by the presence of
Fig. 2. Directed neuronal attachment and axonal outgrowth. After 1 day, rat hippocampal neurons seeded onto patterned substrates in the presence of 10% (v/v) serum selectively attached to, and extended axonal processes along, patterns of polylysine-conjugated laminin. The representative #uorescence micrographs of this "gure illustrate neuron attachment and outgrowth along patterns composed of linear segments with the following dimensions: (A) d "7.5 m, dl "13 m; (B) d "6.7 m, dl "43 m; and (C) d "2.6 m, dl "43 m. Substrate-bound laminin was immuno #uorescently labeled with Alexa-488, and appears red in this "gure. Neurons were identi"ed by immunochemical staining for a neuron-speci"c subunit of tubulin with Quantum Red, and are shown in green-yellow. These three images are shown at identical magni"cations. Scale bar"50 m.
1052
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
axonal processes measuring at least 150 m in length; moreover, these axonal processes were associated exclusively with the patterns of polylysine-conjugated laminin. Minor neurites (i.e., short, non-axonal processes) elaborated by adherent neurons were similarly selectively associated with regions of polylysine-conjugated laminin, and branched away from the cell soma along separate segments of the micro-scale patterns. The few adhering neurons on intervening regions of unmodi"ed glass remained rounded and did not elaborate cellular processes. Axonal outgrowth on these polylysine-conjugated laminin-modi"ed surfaces was qualitatively in#uenced by pattern dimension. The morphology of neurons adherent to patterns composed of short, wide segments resembled that of neurons in routine culture on unpatterned, celladhesive surfaces; this morphology is illustrated in Fig. 2A, which shows neuron outgrowth on a pattern of segments measuring 8.3 m in width (d ) and 13 m in length (dl ). On these surfaces, the axonal process elaborated by adhering neurons traversed the network of polylysine-conjugated laminin, meandering across each micropatterned segment, and avoided the small regions of unmodi"ed glass. In contrast, longer, narrower segments of polylysine-conjugated laminin provided guidance for both axonal and minor processes elaborated by adhering neurons; Fig. 2B illustrates this response on a pattern containing segments measuring 5.6 m in width (d ) and 43 m in length (dl ). As the pattern width was further reduced, the path of axonal growth was controlled with increasing precision. This response is illustrated in Fig. 2C, which illustrates a neuron adhering to a pattern containing segments measuring 2.6 m in width (d ) and 43 m in length (dl ); the axonal process remained con"ned to the prescribed pattern, including bending where multiple linear segments intersected. Directed neuron attachment and axon outgrowth on polylysine-conjugated laminin was in#uenced by, but not dependent upon, the presence of serum proteins. Under completely serum-free conditions (that is, both the substrates and the neurons were not exposed to serum either before, during, or after cell seeding), adhering neurons extended axonal processes along the patterns of polylysine-conjugated laminin (Fig. 3A). Axonal outgrowth was, however, slower than that observed in the presence of serum; in fact, neurons required several days to exhibit extensive axonal outgrowth in the absence of serum proteins (Fig. 3A). In addition, neuron attachment onto surfaces modi"ed with polylysine-conjugated laminin was less selective in the absence than in the presence of serum proteins. Occasionally, neurons adhered to regions of plain glass (which were delineated and surrounded by the patterns of polylysine-conjugated laminin) in the absence of serum (arrowheads, Fig. 3A); these cells remained rounded,
Fig. 3. Bioactivity of microcontact-printed polylysine-conjugated laminin. After 3 days of culture under serum-free conditions, hippocampal neurons selectively extended axons along patterns of polylysineconjugated laminin (A), but not on patterns of polylysine alone (B). The arrows in frame A assist in identifying the axonal process in this image. Neurons that adhered onto either plain glass or polylysine exhibited a rounded morphology, and did not elaborate cellular processes; examples of these cells are indicated by the arrowheads in frames A and B. Adherent neurons in these representative #uorescence micrographs were labeled with diIC (3). This staining procedure also resulted in labeling of the substrate-bound polylysine and polylysine-conjugated laminin, allowing concurrent visualization of the biomolecular pattern in the background. Both frames are presented at identical magni"cations. Scale bar"50 m.
and did not elaborate any processes. On surfaces patterned with polylysine alone, neurons adhered to both regions of plain glass and of polylysine (Fig. 3B); these adherent neurons remained rounded (arrowheads, Fig. 3B).
4. Discussion Microcontact printing of polylysine-conjugated laminin onto material surfaces illustrates a versatile, robust technique for controlling neuron attachment and
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
outgrowth. In particular, the present study demonstrated that patterns of polylysine-conjugated laminin can provide micrometer-scale guidance of axon outgrowth over distances of several hundred micrometers (Fig. 2C). Control over neuron responses was achieved using micrometer-scale patterns of cell-adhesive compounds (most notably polylysine [3,5,9,10] and aminosilanes [1]) on biomaterial surfaces. Attachment of rat hippocampal neurons to micropatterned polylysine alone was, however, dependent on adsorbed (but, to date, not speci"cally identi"ed) serum proteins (Fig. 3B); similar observations of serum-dependent attachment have been reported for other types of neurons [1]. In contrast, the present study demonstrated that neuron attachment and axon outgrowth on polylysine-conjugated laminin is not dependent on adsorption of serum proteins (Fig. 3A), indicating that substrate-bound laminin, in the form of polylysineconjugated laminin, is biologically active. Polylysineconjugated laminin may be useful in guiding neuron growth in low-density, serum-free cell cultures, a situation in which exogenous adhesion factors are minimal and which is ideal for studying and utilizing neuron networks. It should be noted, however, that in the absence of serum proteins in the present study, neurons attach to regions of unmodi"ed glass (Fig. 3B). This result is in sharp contrast to that obtained in the presence of serum (Fig. 2), but is in agreement with earlier reports [1] that adsorption of certain, still unidenti"ed, serum components onto material surfaces minimizes neuron attachment. Identi"cation of these serum components, which likely include serum albumin, could be used to enhance selectivity of neuron attachment on micropatterned material surfaces. An important advantage of microcontact printing techniques [20,21] is their additive nature, which allows development of spatially complex surfaces containing multiple patterns of di!erentially bioactive compounds [15,22]. Techniques for directing neuron attachment and axon outgrowth using micropatterned laminin, such as UV ablation [12,14] and selective adsorption [13], are not suited for producing multifunctionalized systems. Spatially and chemically complex systems micropatterned on substrates could provide valuable tools for the next generation of studies of neuron function. For example, investigation of how biological and/or electrical stimuli either along segments or across intersections of micro-scale patterns in#uence axon outgrowth, a presumably random process in the present study, could provide new information regarding the molecular mechanisms that direct neuron path-"nding both in vitro and in vivo. Such developments could provide important biological and materials engineering information valuable to the design of advanced, neuron-based devices.
1053
Acknowledgements This work was sponsored in part by NIH NCRRR01-RR-10957, DARPA/ITO (to J.N. Turner and W. Shain) and by a special gift H71001 (to R. Bizios).
References [1] Kleinfeld D, Kahler KH, Hockberger PE. Controlled outgrowth of dissociated neurons on patterned substrates. J Neurosci 1988;8: 4098}120. [2] Clark P, Connolly P, Curtis ASG, Dow JAT, Wilkinson CDW. Topographical control of cell behavior. II: multiple grooved substrata. Development 1990;108:635}44. [3] Corey JM, Wheeler BC, Brewer GJ. Compliance of hippocampal neurons to patterned substrate networks. J Neurosci Res 1991;30: 300}7. [4] Nagata I, Kawana A, Nakatsuji N. Perpendicular contact guidance of CNS neuroblasts on arti"cial microstructures. Development 1993;117:401}8. [5] Corey JM, Brunette AL, Chen MS, Weyhenmeyer JA, Brewer GJ, Wheeler BC. Di!erentiated B104 neuroblastoma cells are a highresolution assay for micropatterned substrates. J Neurosci Meth 1997;75:91}7. [6] Branch DW, Corey JM, Weyhenmeyer JA, Brewer GJ, Wheeler BC. Microstamp patterns of biomolecules for high-resolution neuronal networks. Med Biol Eng Comp 1998;36:135}41. [7] Saneinejad S, Shoichet MS. Patterned Glass surfaces direct cell adhesion and process outgrowth of primary neurons of the central nervous system. J Biomed Mater Res 1998;42:13}9. [8] Stenger DA, Hickman JJ, Bateman KE, et al. Microlithographic determination of axonal/dendritic polarity in cultured hippocampal neurons. J Neurosci Meth 1998;82:167}73. [9] James CD, Davis RC, Meyer M, et al. Aligned microcontact printing of micrometer scale polylysine structures for controlled growth of cultured neurons on planar microelectrode arrays. IEEE Trans Biomed 2000;47:17}21. [10] Branch DW, Wheeler BC, Brewer GJ, Leckband DE. Long-term maintenance of patterns of hippocmpal pyramidal cells on substrates of polyethylene glycol and microstamped polylysine. IEEE Trans Biomed 2000;47:290}300. [11] Hammarback JA, Letourneau PC. Neurite extension across regions of low cell-substratum adhesivity: implications for the guidepost hypothesis of axonal path"nding. Dev Biol 1986;117:655}62. [12] Hammarback JA, McCarthy JB, Palm SL, Furcht LT, Letourneau PC. Growth cone guidance by substrate-bound laminin pathways is correlated with neuron-to-pathway adhesivity. Dev Biol 1988;126:29}39. [13] Clark P, Britland S, Connolly P. Growth cone guidance and neuron morphology on micropatterned laminin surfaces. J Cell Sci 1993;105:203}12. [14] Tai HC, Buettner HM. Neurite outgrowth and growth cone morphology on micropatterned surfaces. Biotechnol Prog 1998; 14:364}70. [15] Kam L, James CD, Withers G, et al. Neuron attachment and outgrowth on microcontact-printed polylysine-conjugated laminin. J Neurosci Meth 2001, in press. [16] Kam L. Modulation on neuron and astroglial cell function by micropatterning and immobilization of select biomolecules on biomaterial surfaces. Doctoral dissertation, Rensselaer Polytechnic Institute, Troy, NY, 1999. [17] James CD, Davis RC, Kam L, et al. Patterned protein layers on solid substrates by thin stamp microcontact printing. Langmuir 1998;14:741}4.
1054
L. Kam et al. / Biomaterials 22 (2001) 1049}1054
[18] Kam L, Shain W, Turner JN, Bizios R. Correlation of astroglial cell function on micro-patterned surfaces with speci"c geometric parameters. Biomaterials 1999;20:2343}50. [19] Banker G, Goslin K, editors. Culturing nerve cells. Cambridge: The MIT Press, 1991. p. 251. [20] Singhvi R, Kumar A, Lopez GP, et al. Engineering cell shape and function. Science 1994;264:696}8.
[21] Bernard A, Delamarche E, Schmid H, Michel B, Bosshard HR, Biebuyck H. Printing patterns of proteins. Langmuir 1998;14: 2225}9. [22] Wheeler BC, Corey JM, Brewer GJ, Branch DW. Microcontact printing for precise control of nerve cell growth in culture. J Biomech Engng 1999;121:73}8.