DOI: 10.1002/cssc.200800030

Processing Conditions for the Formation of Spider SilkMicrospheresAndreas Lammel,[b] Martin Schwab,[c] Ute Slotta,[a] Gerhard Winter,[c] and Thomas Scheibel*[a]

Introduction

Numerous biopolymers have been optimized with respect todistinct applications over millions of years during evolution.For instance, spider silk exhibits extraordinary mechanicalproperties such as toughness and elasticity required to func-tion as a trap for prey, which even exceeds the mechanicalproperties of Kevlar, one of the most stable and toughest syn-thetic fibers.[1, 2] Spider silk is a material consisting of very largeproteins (>200 kDa) and has a high potential for biomedicalapplications as a result of its biocompatibility and biodegrada-bility.[3, 4] Especially in the area of new innovative and effectivedrug-delivery systems, spider silk proteins are very well suitedbecause of possible functionalization, that is, coupling ofactive agents.[5] It is feasible to develop vesicles made of spidersilk proteins that can meet the demanding requirements ofspecificity and efficiency for the delivery of active ingredientssuch as drugs and pharmaceutical proteins.[6, 7]

Results and Discussion

Generally, protein aggregation is driven by a shift of the ther-modynamic equilibrium state caused by factors such as tem-perature and ionic strength.[8] Conformational conversion ofspider silk proteins can be triggered by potassium phosphate,which is very important in terms of biocompatibility of thedrug-delivery system considering both the material and itsprocessing.[9–13] Here, the engineered spider silk protein eADF4-ACHTUNGTRENNUNG(C16) was incubated in Tris buffer (10 mm, pH 8) at0.5 mgmL�1 or 1 mgmL�1 and salted out in the presence of500 mm or 1m potassium phosphate (pH 8). It was determinedthat the required time for salting out of 50% of eADF4 ACHTUNGTRENNUNG(C16)was t50=50 s in the presence of 500 mm potassium phosphate(pH 8) and t50=30 s in the presence of 1m potassium phos-phate (pH 8; Figure 1a). Independent of the starting proteinconcentration, the equilibrium between soluble and aggregat-ed ADF4 ACHTUNGTRENNUNG(C16) is only dependent on the concentration of po-tassium phosphate. Interestingly, a constant amount of protein

per 1 mL volume remained soluble in each sample: 333 mg cal-culated in the presence of 500 mm potassium phosphate, and250 mg calculated in the presence of 1m potassium phosphate(Figure 1a, and data not shown).

Next, the morphology of the aggregates was investigated.Above a potassium phosphate concentration of 500 mm, thequantitative formation of microspheres could be detected.[13]

To analyze the influence of mixing on microsphere formation,several different methods were employed for mixing includingdialysis, simple mixing with a pipette, and micromixing withina T-mixing element. Upon mixing, the samples were analyzedwith laser diffraction spectrometry and scanning electron mi-croscopy (SEM). Accordingly, different sphere characteristicswith respect to sphere sizes, size distributions, and formationof sphere clusters could be detected depending on themethod of preparation.

At a concentration of potassium phosphate of 500 mm

(pH 8), silk spheres formed clusters with all the examined prep-aration methods (dialysis, pipette, micromixing at 2 mLmin�1

(Reynolds number, Re=85, see Experimental Section) and at50 mLmin�1 (Re=2122)). Figure 2a shows a typical SEM imageof silk spheres produced by salting out with 500 mm potassi-um phosphate (pH 8). The micrograph illustrates that spheres

Spider silk is a material consisting of very large (>200 kDa) pro-teins and has a high potential for biomedical applications as aresult of its biocompatibility and biodegradability. We report onthe influence of physicochemical factors on structure formationof the engineered spider silk protein eADF4 ACHTUNGTRENNUNG(C16), which mimicsthe known sequence of the dragline protein ADF4 from thespider Araneus diadematus. Under certain experimental condi-

tions, eADF4 ACHTUNGTRENNUNG(C16) forms stable microspheres that have been ana-lyzed with respect to sphere size, size distribution, and surface in-ertness upon different preparation methods (dialysis, pipette andmicromixing). As a result of their material strength, biocompati-bility, and the possibility of functionalization, spider silk micro-spheres have a high potential for the development of targeteddrug-delivery systems.

[a] U. Slotta, Prof. Dr. T. ScheibelFAN, Lehrstuhl Biomaterialien, Universit1t BayreuthUniversit1tsstrasse 30, 95440 Bayreuth (Germany)Fax: (+49)921-55-7346E-mail : Thomas.scheibel@uni-bayreuth.de

[b] A. LammelDepartment Chemie, Lehrstuhl BiotechnologieTechnische Universit1t M;nchenLichtenbergstrasse 4, 857474 Garching (Germany)

[c] M. Schwab, Prof. Dr. G. WinterDepartment of PharmacyPharmaceutical Technology and BiopharmaceuticsLudwig-Maximilians-Universit1tButenandtstrasse 5, 81377 M;nchen (Germany)

ChemSusChem 2008, 1, 413 – 416 A 2008 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim www.chemsuschem.org 413

with a diameter of approximately 500 nm build larger clustersof aggregates. In contrast, in salting out experiments per-formed with 1m potassium phosphate (pH 8) mainly fully de-veloped microspheres could be observed (Figure 2b). Micro-mixing with higher mixing intensities led to the formation ofsmaller spheres than salting out with dialysis or simple mixingwith pipette. The SEM image in Figure 2c depicts spheres pro-duced by micromixing at 2 mLmin�1, and Figure 2d shows thespheres produced by mixing with a flow rate of 50 mLmin�1.Strikingly, dialysis, which reflects the slowest mixing conditions,resulted in the largest observed spheres with diameters greaterthan 1 mm (Figure 2b).

To quantify the size distributions, as qualitatively observedby SEM, all samples without larger aggregates were analyzedby laser diffraction spectrometry. Thus, we analyzed all samplesprepared by salting out with 1m potassium phosphate (pH 8).Figure 3a shows the size distribution of spheres produced bymixing 1m potassium phosphate (pH 8) with 1 mgmL�1

eADF4 ACHTUNGTRENNUNG(C16) solutions with a pipette as well as with micromix-ing at rates of 2 mLmin�1 and 50 mLmin�1. Figure 3b illus-trates the size distribution of spheres salted out by dialysis of0.5 mgmL�1 and 1 mgmL�1 eADF4 ACHTUNGTRENNUNG(C16) solutions against 1m

potassium phosphate (pH 8).As seen by sharp peaks, controlled mixing with a micromix-

ing device leads to a smaller size distribution as compared tomixing with pipette or salting out by dialysis. Simple mixingwith a pipette at eADF4 ACHTUNGTRENNUNG(C16) concentrations of 0.5 mgmL�1

and 1 mgmL�1 led to stable microspheres with average diame-ters of 350 nm and 510 nm, respectively (Figure 3c). By con-trolled mixing at a flow rate of 2 mLmin�1, microspheres withan average size of 290 nm (0.5 mgmL�1) and 370 nm(1 mgmL�1) were formed, whereas at a flow rate of50 mLmin�1 smaller microspheres with average diameters of250 nm and 320 nm could be detected (Figure 3c). Dialysisyielded stable spheres with average sizes of 1.6 mm and2.1 mm, respectively, with a rather broad size distribution (Fig-ure 3c).

To investigate the stability of microspheres and clustersthereof, sphere size distributions were measured with laserspectrometry over a time period of 10 min with constantmixing to prevent sedimentation of microspheres. At low po-tassium phosphate concentration (500 mm, pH 8), clusterswere formed which disassembled upon induced shear stressprovided by stirring (Figure 4). With 1m potassium phosphate(pH 8), stable microspheres could be produced independent ofthe protein concentration or the mixing condition (Figure 2c,d,Figure 3c, and Figure 4).

Conclusions

We conclude that for the production of spider silk micro-spheres by salting out, a high concentration of potassiumphosphate is required to obtain stable spheres. Sphericalgrowth stops when the protein concentration is below theequilibrium of solubility, leading to the fact that the sphere di-ameter does not increase anymore.[13] Generally, the spheresize is increased with increasing protein concentration

Figure 1. Aggregation assay of eADF4 ACHTUNGTRENNUNG(C16) protein. a) Aggregation behaviorof eADF4 ACHTUNGTRENNUNG(C16) as a function of time at 25 8C and at protein and potassiumphosphate concentrations as indicated. b) Aggregation of eADF4 ACHTUNGTRENNUNG(C16) as afunction of potassium phosphate concentration and temperature after 1 hincubation. At lower salt concentrations, the aggregation is accelerated withincreasing temperature. 50% of the protein is aggregated at (381�4.2) mm

and 10 8C, at (205.5�4.3) mm and 30 8C, and at (85.2�1.7) mm and 60 8C.The volume of each experiment was kept constant at 1 mL.

Figure 2. SEM images of microspheres produced by different methods.a) Micromixing of 500 mm potassium phosphate (pH 8) with eADF4ACHTUNGTRENNUNG(C16) so-lution (0.5 mgmL�1) at a flow rate of 50 mLmin�1. b) Dialysis of eADF4ACHTUNGTRENNUNG(C16)solution with a protein concentration of 1 mgmL�1 against 1m potassiumphosphate (pH 8). c) Micromixing of 1m potassium phosphate (pH 8) witheADF4ACHTUNGTRENNUNG(C16) solution (1 mgmL�1) at a flow rate of 2 mLmin�1. d) Micromixingof 1m potassium phosphate (pH 8) with eADF4 ACHTUNGTRENNUNG(C16) solution (1 mgmL�1) ata flow rate of 50 mLmin�1.

414 www.chemsuschem.org A 2008 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem 2008, 1, 413 – 416

T. Scheibel et al.

(Figure 5). The sphere size can be further adjusted by con-trolled micromixing of eADF4 ACHTUNGTRENNUNG(C16) with 1m potassium phos-phate (pH 8) inside a T-mixing element (Figure 5).

The obtained microspheres represent a new class of bio-mimetic materials that are an alternative to microcapsules forthe incorporation of active ingredients such as therapeutic pro-teins. In contrast to spider silk microcapsules, the productionof microspheres is much simpler as no formation of vesicles isrequired.[6,7] Incorporation of active ingredients into spider silkmicrospheres can be realized by adding a solution with the de-sired molecules before microsphere formation upon addition

of potassium phosphate. Studies are underway to reveal infor-mation about the likely influence of guest molecules on spidersilk spheres with respect to structural and phase-separation be-havior. As targeted drug delivery demands the presence andexposure of signal molecules, in a next step ADF4 ACHTUNGTRENNUNG(C16) mi-crobeads can be functionalized on their surface with respectivemolecules similar to a process described previously for themodification of surfaces of films made of ADF4 ACHTUNGTRENNUNG(C16).[5] Presum-ably, mechanical stability of the newly detected microspheresis much higher than that of the previously described microcap-sules.[6, 7] Owing to material strength and biocompatibility,spider silk microspheres have a high potential for the develop-ment of targeted drug-delivery systems.

Experimental Section

Engineering of eADF4ACHTUNGTRENNUNG(C16): The amino acid sequence of eADF4(C16) was adapted from the natural sequence of ADF4 from Ara-neus diadematus. The repetitive part of ADF4 is generally com-posed of a single conserved repeat unit displaying only slight var-

Figure 3. Size analysis of microspheres salted out with 1m potassium phos-phate (pH 8). a) Size distribution of eADF4 ACHTUNGTRENNUNG(C16) microspheres prepared bydifferent methods: simple mixing with a pipette and controlled mixing witha micromixing device at flow rates of 2 mLmin�1 and 50 mLmin�1. b) Sizedistribution of eADF4ACHTUNGTRENNUNG(C16) microspheres prepared by dialysis. c) Calculatedmean sphere size of eADF4ACHTUNGTRENNUNG(C16) microspheres prepared by different meth-ods as indicated.

Figure 4. Conformational stability of microspheres produced by micromixingof eADF4 ACHTUNGTRENNUNG(C16) solution (0.5 mgmL�1) with potassium phosphate (pH 8) at aflow rate of 2 mLmin�1. Stable spheres are formed upon salting out of theprotein with 1m potassium phosphate (pH 8). After 10 min stirring, no de-crease or increase in particle size could be detected. In case of micromixing500 mm potassium phosphate (pH 8) with eADF4ACHTUNGTRENNUNG(C16) solution(0.5 mgmL�1), clusters were formed which slowly disassembled upon stir-ring.

Figure 5. Influence of protein concentration and mixing intensity on micro-sphere formation produced by salting out with 1m potassium phosphate(pH 8). Larger spheres can be produced with increasing protein concentra-tions, whereas the sphere size can be decreased with increasing mixing in-tensity. Scale bars : 1 mm.

ChemSusChem 2008, 1, 413 – 416 A 2008 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim www.chemsuschem.org 415

Spider Silk Microspheres

iations. Repetitive elements in the sequence of ADF4 display apoly ACHTUNGTRENNUNGalanine-rich motif with high glycine content, which wasnamed C. Our previously established engineering approach al-lowed the combination and multimerization of the single motifs,resulting in the eADF4ACHTUNGTRENNUNG(C16) protein which comprises 16 repeats ofthe sequence GSSAAAAAAAASGPGGYGPENQGPSGPGGYGPGGP re-sulting in a molecular mass of 48 kDa.[9,10] The protein was purifiedas described previously to yield a purity of over 98%.[10]

Dialysis : Lyophilized protein eADF4ACHTUNGTRENNUNG(C16) was dissolved in 6m gua-nidinium thiocyanate. Dialysis was performed against 10 mmtris(hydroxymethyl)aminomethane ACHTUNGTRENNUNG(Tris)/HCl, pH 8, at 4 8C usingmembranes with a molecular weight cut-off at 6000–8000 Da(Spectrum Laboratories, Rancho Dominguez, USA).

Determination of protein concentration: The concentration ofeADF4ACHTUNGTRENNUNG(C16) solutions was determined by UV measurements at20 8C using a Cary100 spectrophotometer (Varian Medical Systems,Palo Alto, USA). The molar extinction coefficient of eADF4ACHTUNGTRENNUNG(C16) at276 nm and 20 8C was employed (e=40974m

�1 cm�1).

Preparation of eADF4ACHTUNGTRENNUNG(C16) microspheres: Four different prepara-tion methods were employed for inducing salting out: 1) simplemixing (pipette) of 2m potassium phosphate (pH 8) with eADF4-ACHTUNGTRENNUNG(C16) solution; 2) mixing of 2m potassium phosphate (pH 8) witheADF4ACHTUNGTRENNUNG(C16) solution in a micromixing device under laminar flowconditions (2 mLmin�1, Re=85); 3) mixing of 2m potassium phos-phate (pH 8) with eADF4 ACHTUNGTRENNUNG(C16) in a micromixing device under tur-bulent flow conditions (50 mLmin�1, Re=2122); and 4) dialysis ofeADF4ACHTUNGTRENNUNG(C16) solution against 1m potassium phosphate (pH 8).

Micromixing device: Constant volume flows of educt solutions (po-tassium phosphate (pH 8) and eADF4ACHTUNGTRENNUNG(C16) solution) were generat-ed by two identical syringe pumps (Model 100DX, Teldyne Isco,Inc. , USA). Pumps were operated by a digital controler (ISCO SeriesD, Teldyne Isco, Inc. , USA). Mixing and precipitation was performedin a T-shaped mixing element with a circular mixing zone with a di-ameter of 500 mm (Peek Tee, P-728 Threads 10–32, Upchurch Scien-tific). The feed tubes were 500 mm in diameter and located oppo-site to each other. To characterize the flow through the mixing ele-ment, a mixing Reynolds number is defined according to Equa-tion (1), where u is the average fluid velocity, d is the diameter ofthe mixing zone, v is the dynamic viscosity of water, and Q is thevolumetric flow rate.

Re ¼ u d=v ¼ 4Q=ðp d vÞ ð1Þ

In the discussed experiments, mixing was conducted at flow ratesof 2 mLmin�1 and 50 mLmin�1, which correspond to Reynoldsnumbers of Re=85 and Re=2122, respectively. The flow inside themixing zone can be considered turbulent owing to the impingingof the two solutions. The generated suspensions were collected in1.5-mL Eppendorf tubes.

Determination of sphere size: Particles sizes and their distributionswere determined with laser diffraction spectrometry (Horiba, Parti-ca LA-950, Japan). Refractive indices of 1.33 for water and 1.60 forprotein were taken for computation of particle sizes. In addition, adry specimen of each preparation was analyzed by scanning elec-tron microscopy (SEM) to confirm sphere formation and spheresizes. In SEM analyses, eADF4 ACHTUNGTRENNUNG(C16) microspheres were immobilizedon Thermanox plastic cover slips (Nagle Nunc, USA), which hadbeen coated with gold by sputtering under vacuum, and analyzedwith a JSM 5900 LV scanning electron microscope (JEOL Ltd. ,Japan, at 20 kV).

Acknowledgements

This work was supported by the Deutsche Forschungsgemein-schaft (SCHE 603/4-3 to T.S.) and the International GraduateSchool of Science and Engineering (IGSSE to A.L.) within the EliteNetwork of Bavaria.

Keywords: biopolymers · drug delivery · protein engineering ·proteins · self-assembly

[1] F. Vollrath, Rev. Mol. Biotechnol. 2000, 74, 67–83.[2] G. H. Altman, F. Diaz, C. Jakuba, T. Calabro, R. L. Horan, J. Chen, H. Lu, J.

Richmond, D. L. Kaplan, Biomaterials 2003, 24, 401–416.[3] B. Panilaitis, G. H. Altman, J. Chen, H.-J. Jin, V. Karageorgiou, D. L.

Kaplan, Biomaterials 2003, 24, 3079–3085.[4] R. L. Horan, K. Antle, A. L. Collette, Y. Wang, J. Huang, J. E. Moreau, V.

Volloch, D. L. Kaplan, G. H. Altman, Biomaterials 2005, 26, 3385–3393.[5] D. Huemmerich, U. Slotta, T. Scheibel, Appl. Phys. A 2006, 82, 219–222.[6] K. D. Hermanson, D. Huemmerich, T. Scheibel, A. R. Bausch, Adv. Mater.

2007, 19, 1810–1815.[7] K. D. Hermanson, M. B. Harasim, T. Scheibel, A. R. Bausch, Phys Chem.

Chem. Phys. 2007, 9, 6442–6446.[8] V. N. Uversky, Eur. J. Biochem. 2002, 269, 2–12.[9] T. Scheibel, Microb. Cell Fact. 2004, 3, 14.

[10] D. Huemmerich, C. W. Helsen, S. Quedzuweit, J. Oschmann, R. Rudolph,T. Scheibel, Biochemistry 2004, 43, 13604–13612.

[11] U. Slotta, S. Hess, K. Spieß, T. Stromer, L. Serpell, T. Scheibel, Macromol.Biosci. 2007, 7, 183–188.

[12] J. H. Exler, D. Huemmerich, T. Scheibel, Angew. Chem. 2007, 119, 3629–3632; Angew. Chem. Int. Ed. 2007, 46, 3559–3562.

[13] U. Slotta, S. Rammensee, S. Gorb, T. Scheibel, Angew. Chem. 2008, DOI:10.1002/ange.200800683; Angew. Chem. Int. Ed. 2008, DOI: 10.1002/anie.200800683.

Received: December 20, 2007Revised: February 11, 2008Published online on April 18, 2008

416 www.chemsuschem.org A 2008 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemSusChem 2008, 1, 413 – 416

T. Scheibel et al.