An Abiotic Analogue of the Nuclear Pore Complex Hydrogel

COMMUNICATION pubs.acs.org/Biomac

An Abiotic Analogue of the Nuclear Pore Complex Hydrogel Sean P. Bird and Lane A. Baker* Department of Chemistry, Indiana University, Bloomington, Indiana 47405, United States

bS Supporting Information ABSTRACT: We describe an abiotic hydrogel that mimics selectivity of the nuclear pore complex. Copolymerization of peptide tetramers (phenylalanine-serine-phenylalanineglycine, FSFG) with acrylamide results in hydrophobic interactions significant enough to allow the formation of freestanding hydrogel structures. Incorporation of FSFG motifs also renders the hydrogels selective. Selective binding of importins and nuclear transport receptor
cargo complexes is qualitatively demonstrated and compared with polyacrylamide, hydrogels prepared from a control peptide, and hydrogels prepared from the nuclear pore complex protein Nsp1. These abiotic hydrogels will enable further studies of the unique transport mechanisms of the nuclear pore complex and provide an interesting paradigm for the future development of synthetic platforms for separations and selective interfaces.

’ INTRODUCTION Evolution has arrived at elegant and efficient designs for biomaterials. Mussel adhesives,1
3 spider silk,4
6 and the scales of butterfly wings7,8 demonstrate the proficiency of evolved structures. These examples have provided inspiration for the design of abiotic materials that mimic the properties of the natural materials. Here we describe an abiotic polymer hydrogel synthesized without cellular machinery that mimics the binding properties of hydrogels formed from proteins found in the nuclear pore complex (NPC). In vivo, the NPC forms the primary conduit for transport of materials across the nuclear envelope.9
12 The central channel of the NPC is ∼40 nm in diameter, yet transport of proteins is highly regulated. For proteins >30 kDa, transit through the NPC relies on assistance from nuclear transport receptors (NTRs). Proteins that are chaperoned are typically referred to as cargo proteins, and NTRs can be referred to as importins or exportins, as determined by the directionality of transport. Whereas the presence of importins, cargo proteins, and the general structure of the NPC has been known for some time, the exact mechanism of transport through the central channel remains an interesting question.13 Natively unfolded proteins rich in phenylalanine-glycine (FG) motifs extend from the interior walls of the NPC into the central channel and have been demonstrated to play a central role in observed transport properties.14 For instance, it has been shown that deletion or modification of FG motifs in S. cerevisiae can influence transport by the NPC and, in some cases, compromise viability.15 A variety of FG motifs exist and primarily consist of FG, GLFG, FSFG, and FXFG sequences, where X is any amino acid. A number of models have been proposed for the mechanism of selective transport and the role of FG-rich proteins.16,17 Remarkable reports by Frey and G€orlich have demonstrated that certain FG-rich proteins from r 2011 American Chemical Society

the NPC can form hydrogels in vitro with selectivity for importins and NTR 3 cargo complexes akin to the NPC.15,18,19 In particular, the FG-rich protein Nsp1 has been shown to form a hydrogel structure that selectively imports importins, both with and without cargo, but excludes nonspecific proteins. This discovery has lent credence to the “selective phase model”,17 in which a protein network is formed with FG motifs that provide a hydrophobic cross-link for inter- and intraprotein interactions. In this model, importins are able to dissociate these FG-cross-links, which allows transit through the protein network that otherwise restricts protein diffusion. The critical role of FG motifs in the formation of a selective hydrogel was confirmed by mutation of phenylalanine residues to serine in Nsp1, which prevented the formation of a hydrogel. The Nsp1 hydrogel thus presents an intriguing paradigm for development of synthetic materials that can selectively recognize and transport species. Significant efforts have been undertaken to realize hydrogels that incorporate biological entities (e.g., proteins, nucleic acids, etc.) for separations and sensing platforms.20 Here we seek to mimic properties exhibited by the Nsp1 hydrogel (and ultimately the NPC) through the use of peptide-modified hydrogels, as depicted in Figure 1. We have prepared FSFG tetramers that incorporate a polymerizable acryloyl functional group and subsequently copolymerized these tetramers with acrylamide monomers to realize a hydrogel structure with the potential to mimic the transport properties of the NPC.

’ RESULTS AND DISCUSSION Copolymerization of an acryloyl-terminated FSFG tetrapeptide with acrylamide was chosen as a simple route to mimic the Received: June 16, 2011 Revised: August 17, 2011 Published: August 19, 2011 3119

dx.doi.org/10.1021/bm200820x | Biomacromolecules 2011, 12, 3119–3123

Biomacromolecules

Figure 1. Illustration of hydrogels under study. A hydrogel formed from Nsp1, a nuclear pore complex protein, forms a selective gate that allows transport of importins but excludes nonspecific protein entry (left). Hydrogel formation and selectivity rely on the presence of FSFG repeats. Copolymerization of acryloyl-terminated FSFG tetramers with polyacrylamide results in a synthetic hydrogel that mimics transport of importins by Nsp1 hydrogels (right). Note that the illustration serves to highlight cross-link positions of proteins/polymer strands in the bulk hydrogels, which rely on the FSFG position.

chemical interactions of FSFG motifs in Nsp1. Acryloyl-terminated peptide tetramers were synthesized by standard FMOC chemistry and purified by HPLC. (See the Supporting Information.) When copolymerized with acrylamide monomer, the resultant structure consists of an acrylamide backbone with peptide tetramers randomly dispersed throughout the polymer.21
23 As synthesized here, the peptide is not a linear constituent of the polymer backbone (as is the case for Nsp1) but is present instead as a branch from the main polymer backbone. Initial polymerization tests are shown in Figure 2. Materials prepared from polymerization of an aqueous solution of 5% acrylamide with ammonium persulfate and tetramethylethylenediamine inside of a glass capillary are shown in Figure 2a. Under these condi tions and in the absence of bisacrylamide cross-linker (N,N0 methylenebisacrylamide), linear polyacrylamide does not form a free-standing hydrogel. Material resulting from copolymerization of acrylamide with an N-terminal acryloyl-modified FSFG peptide (aFSFG) is shown in Figure 2b. Here the formation of a freestanding hydrogel structure is observed, likely a result of hydrophobic interactions between phenylalanines in aFSFG moieties. To confirm the role of the aFSFG peptide in the formation of a freestanding gel structure, in accordance with previous reports for Nsp1,15 polymerization was also performed with an SSSG peptide with an N-terminal acryloyl modification (aSSSG). In this case, with the absence of hydrophobic interactions provided by phenylalanines, gel formation is not observed. Frey and G€orlich reported similar findings with an Nsp1 mutant, in which phenylalanines were mutated to serines. Our experiments confirm that interactions between FSFGs are critical to the formation of a freestanding polyacrylamide hydrogel absent of

COMMUNICATION

Figure 2. Optical photographs of polymerization products from (a) a solution of 5% (w/v) acrylamide monomer, (b) a solution of 5% (w/v) acrylamide and 200 mM aFSFG, and (c) a solution of 5% (w/v) acrylamide and 200 mM aSSSG. In parts a
c, no bisacrylamide crosslinker was added to solution. (d) Photograph of five 1 μL aFSFGpolyacrylamide hydrogels after soaking in a solution of GFP-tagged importin β.

bisacrylamide. In experiments here, 200 mM peptide was chosen to mimic the overall FSFG content of the protein Nsp1. A key difference between the FSFG acrylamide polymer prepared here and the protein Nsp1 is solubility. Highly basic environments or the formation of trifluoroacetic acid salts are required to solubilize Nsp1 at concentrations necessary for hydrogel formation.18,19 The FSFG acrylamide polymer is soluble in aqueous solution, and thus bisacrylamide was employed in gel polymerizations for in vitro experiments described further below. Fluorescence microscopy of tagged proteins has proven to be a powerful tool for elucidation of transport mechanisms of the Nsp1 hydrogel and confirmation of selectivity.14,18,19 Figure 2d shows a Petri dish that contains a glass microscope slide on which five spots of the aFSFG hydrogel (bisacrylamide cross-linked) have been placed. When exposed to a solution of a GFPtagged nuclear import protein, GFP-importin β (GFP-Impβ), the importin is concentrated in the hydrogel. This occurs to the extent that the color of the aFSFG hydrogel turns from clear (Figure 2b) to green (Figure 2d). To examine the selectivity of the abiotic hydrogels further, we prepared cross-linked acrylamide and aFSFG- and aSSSG-acrylamide hydrogels. Additionally, histidine-tagged Nsp1 protein was expressed and purified, as previously reported.19 Hydrogels of each of these materials were prepared (Supporting Information) and exposed to protein solutions to test selectivity. Comparison of GFP-Impβ is the simplest case. When found in vivo, Impβ serves to transport molecules across the nuclear envelope through the NPC.9 For in vitro experiments here, Impβ serves to verify that a hydrogel structure with selective binding has been achieved. A second protein, HSET was also tagged with GFP. For in vitro experiments here, HSET serves two functions. First, GFP-tagged HSET (GFP-HSET) in the absence of any import proteins serves as a control for selectivity in binding of the Nsp1 hydrogel. Lack of fluorescence, as observed via microscopy, from the interior of an Nsp1 hydrogel indicates that a barrier to free diffusion of appreciable quantities of GFP-HSET into the hydrogel on the 3120

dx.doi.org/10.1021/bm200820x |Biomacromolecules 2011, 12, 3119–3123

Biomacromolecules

COMMUNICATION

Figure 3. Fluorescence micrographs of hydrogels exposed to protein solutions. For all images, the microscope is focused to a plane just above the surface of the glass slide on which the hydrogel rests. The left of each image is the solution; the right of each image is the hydrogel. Accumulation of GFP-tagged protein at the hydrogel/solution interface is indicated by green fluorescence. In the absence of accumulation, the hydrogel/solution interface has been indicated by a dashed white line (interface determined by brightfield microscopy). All acrylamide hydrogels were prepared with bisacrylamide crosslinker. No peptide indicates a pure acrylamide hydrogel, aFSFG indicates aFSFG-acrylamide hydrogel and aSSSG indicates aSSSG-acrylamide hydrogel. Details related to hydrogel preparation and image acquisition are available as Supporting Information.

time scale of these experiments has been achieved. When found in vivo, HSET serves a key role in spindle formation during mitosis.24 Importin β binds to HSET through a second importin, importin α (Impα), to dissociate HSET from spindles, and as such, HSET possesses a nuclear localization sequence (NLS).25 Therefore, in a second function, GFP-HSET serves as confirmation of Nsp1 hydrogel selectivity. When Impα and Impβ (untagged) are added to a solution, enrichment of GFP-HSET into the hydrogel further demonstrates that selective binding has been obtained. Selective binding is especially evident because the complex Impβ 3 Impα 3 GFP-HSET is much larger than GFPHSET alone. Employed in this manner, entry of GFP-HSET into the hydrogel in the presence of importins indicates that a structure with selective binding can be obtained. Results of these experiments are shown as a micrograph table in Figure 3. Each column represents a different hydrogel: the Nsp1 protein, acrylamide (labeled no peptide), aFSFG-acrylamide (labeled aFSFG), and aSSSG-acrylamide (labeled aSSSG). Each row represents respective hydrogels after exposure to a different protein solution for a period of 1 h, with microscope gain settings and solution concentrations identical for each image. The first row is after exposure to a solution of GFP-Impβ, the second row is after exposure to GFP-HSET, and the third row is after exposure to Impβ 3 Impα 3 GFP-HSET. In these images, the microscope is focused just above the plane of the microscope slide on which the hydrogel rests such that the edge of the circular hydrogel that is in the field of view appears as an arc. For all images, the hydrogel portion is on the right side of the image, and the solution that bathes the hydrogel is present on the left side of the image. From these images, the selectivity in binding of the Nsp1 hydrogel is observed for both the nuclear import protein Impβ and cargo (GFP-HSET) chaperoned by the Impβ 3 Impα complex because a bright green fluorescence accumulates at the solution/hydrogel interface. In the absence of import proteins,

no accumulation is observed for GFP-HSET (Figure 3, column 1, middle), where the hydrogel/solution interface has been indicated by a dashed white line, as determined by bright field microscopy. In fact, for all hydrogels, the GFP-HSET complex does not accumulate at the interface of the hydrogel (middle row) in the absence of import proteins. For the case of acrylamide with no peptide, no fluorescence accumulation is observed for any of the GFP-tagged proteins. Likewise, for the aSSSG hydrogel, no accumulation is observed. For the aFSFG hydrogel, selectivity that mimics the case of Nsp1 is observed in that GFP-Impβ and Impβ 3 Impα 3 GFP-HSET accumulate into the hydrogel, but GFP-HSET is excluded in the absence of import proteins. Additional experiments related to the aFSFG hydrogel and Nsp1 are shown in Figure 4. A fluorescence micrograph of an aFSFG hydrogel that was incubated with GFP-Impβ for 1 h is shown in Figure 4a. To determine the depth of entry, we measured cross sections (indicated by a dotted line) of the hydrogel/solution interface. Results of measurements for Nsp1 or aFSFG-hydrogels incubated with GFP-Impβ or Impβ 3 Impα 3 GFP-HSET for 1 h are shown in Figure 4b. Entry depths for complexes are shown in each case. For Nsp1 gels, GFP-Impβ enters 58 μm (full-width half-maximum) and Impβ 3 Impα 3 GFP-HSET enters 69 μm. For aFSFG-polyacrylamide, GFPImpβ enters 31 μm and Impβ 3 Impα 3 GFP-HSET enters 62 μm. These data suggest that diffusion into the Nsp1 hydrogel and the aFSFG-polyacrylamide hydrogel occur on a similar spatial and temporal scale, although it should be noted that measurement of intragel diffusion is required for quantitative comparisons. Alterations in diffusion within the aFSFG-polyacrylamide hydrogel could be achieved through the concentration of aFSFG present, and these results are presently being investigated further. We speculate that further control over cross-link density will allow control over intragel rates of diffusion. 3121

dx.doi.org/10.1021/bm200820x |Biomacromolecules 2011, 12, 3119–3123

Biomacromolecules

COMMUNICATION

Fluorescence recovery after photobleaching (FRAP) was also employed to compare acrylamide hydrogels prepared with aFSFG to hydrogels prepared from the protein Nsp1. In previous reports, FRAP has been utilized to probe the mobility of proteins in Nsp1 hydrogels.15 Shown in Figure 4c are the results from a FRAP experiment performed on an aFSFG hydrogel incubated with Impβ 3 Impα 3 GFP-HSET for 1 h. The relatively high mobile fraction indicates that the NTR 3 cargo complex is free to diffuse through the aFSFG hydrogel, a phenomenon also observed in Nsp1 hydrogels.

’ CONCLUSIONS We have demonstrated that when acrylamide is copolymerized with an acryloyl-modified FSFG peptide in the presence of bisacrylamide, a hydrogel is realized that mimics the selective binding of hydrogels prepared from proteins found in the NPC. Even in the absence of the overall sequence of NPC proteins, incorporation of this short peptide fragment imparts biomimetic transport properties to the polyacrylamide hydrogel. Comparison with control hydrogels and the protein Nsp1 clearly demonstrate that selective binding of importins and NTR 3 cargo complexes can be attained. The aFSFG hydrogel presents an interesting venue for further exploration of the role of FG repeats in the NPC. Questions related to mechanisms of binding and diffusion within the aFSFG hydrogel are being examined presently in further detail through variation of the bisacrylamide cross-linker and aFSFG concentrations in prepared hydrogels. Copolymerization with varied chemical constituents (e.g., charge, hydrophobicity) is expected to expand our knowledge of the nuclear import process and understanding of the selectivity of the NPC. Further efforts to adapt the unique transport paradigm of the NPC to platforms with unique separation or recognition properties are ongoing. ’ ASSOCIATED CONTENT

bS

Supporting Information. Experimental details. This material is available free of charge via the Internet at http://pubs.acs. org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Tel: 812-856-1873.

’ ACKNOWLEDGMENT Financial support of the National Science Foundation, DMR Biomaterials Award 0906843, and the Research Corporation (Cottrell Scholar’s Award to L.A.B) are acknowledged. We thank Ms. Alicia Friedman for comments and Prof. Wookyoung Lee for assistance in peptide synthesis. ’ REFERENCES Figure 4. (a) Fluorescence micrograph of an aFSFG hydrogel after exposure to GFP-Impβ. The dotted line represents where fluorescence intensity was integrated to determine entry depth of GFP-Impβ. (b) Fluorescence intensity versus distance plots for each protein under study in aFSFG-polyacrylamide and Nsp1 hydrogels. (c) FRAP plot demonstrating the mobility of Impβ 3 Impα 3 GFP-HSET in an aFSFG hydrogel.

(1) Deming, T. J. Curr. Opin. Chem. Biol. 1999, 3, 100–105. (2) Lee, H.; Dellatore, S. M.; Miller, W. M.; Messersmith, P. B. Science 2007, 318, 426–430. (3) Yu, M. E.; Deming, T. J. Macromolecules 1998, 31, 4739–4745. (4) Karageorgiou, V.; Kaplan, D. Biomaterials 2005, 26, 5474–5491. (5) O’Brien, J. P.; Fahnestock, S. R.; Termonia, Y.; Gardner, K. C. H. Adv. Mater. 1998, 10, 1185–1195. (6) Vollrath, F.; Knight, D. P. Nature 2001, 410, 541–548. 3122

dx.doi.org/10.1021/bm200820x |Biomacromolecules 2011, 12, 3119–3123

Biomacromolecules

COMMUNICATION

(7) Huang, J. Y.; Wang, X. D.; Wang, Z. L. Nano Lett. 2006, 6, 2325–2331. (8) Parker, A. R.; Townley, H. E. Nat. Nanotechnol. 2007, 2, 347–353. (9) G€orlich, D.; Kutay, U. Annu. Rev. Cell Dev. Biol. 1999, 15, 607–660. (10) Paine, P. L.; Moore, L. C.; Horowitz, S. B. Nature 1975, 254, 109–114. (11) Rexach, M.; Blobel, G. Cell 1995, 83, 683–692. (12) Wente, S. R.; Rout, M. P. Cold Spring Harbor Perspect. Biol. 2010, 2. (13) Weis, K. Cell 2007, 130, 405–407. (14) Patel, S. S.; Belmont, B. J.; Sante, J. M.; Rexach, M. F. Cell 2007, 129, 83–96. (15) Frey, S.; Richter, R. P.; G€orlich, D. Science 2006, 314, 815–817. (16) Lim, R. Y. H.; Huang, N.-P.; K€oser, J.; Deng, J.; Lau, K. H. A.; Schwarz-Herion, K.; Fahrenkrog, B.; Aebi, U. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 9512–9517. (17) Ribbeck, K.; G€orlich, D. EMBO J. 2001, 20, 1320–1330. (18) Frey, S.; G€orlich, D. EMBO J. 2009, 28, 2554–2567. (19) Frey, S.; G€orlich, D. Cell 2007, 130, 512–523. (20) Bird, S. P.; Baker, L. A. Analyst 2011, 136, 3410–3418. (21) Ehrick, J. D.; Deo, S. K.; Browning, T. W.; Bachas, L. G.; Madou, M. J.; Daunert, S. Nat. Mater. 2005, 4, 298–302. (22) Miyata, T.; Asami, N.; Uragami, T. Nature 1999, 399, 766–769. (23) Kamarun, D.; Zheng, X. W.; Milanesi, L.; Hunter, C. A.; Krause, S. Electrochim. Acta 2009, 54, 4985–4990. (24) Cai, S.; Weaver, L. N.; Ems-McClung, S. C.; Walczak, C. E. Mol. Biol. Cell 2009, 20, 1348–1359. (25) Kalderon, D.; Roberts, B. L.; Richardson, W. D.; Smith, A. E. Cell 1984, 39, 499–509.

3123

dx.doi.org/10.1021/bm200820x |Biomacromolecules 2011, 12, 3119–3123