Protamine Assembled in Multilayers on Colloidal Particles Can Be

Jun 18, 2004 - The capability of bound PRM to exchange with PRM in solution is attributed to its relatively small size. The demonstrated exchange may ...
12 downloads 7 Views 109KB Size
Download PDF
Biomacromolecules 2004, 5, 1580-1587

1580

Protamine Assembled in Multilayers on Colloidal Particles Can Be Exchanged and Released Sabine Hiller,* Stefano Leporatti, Andreas Schna¨ckel, Elke Typlt, and Edwin Donath Institute of Medical Physics and Biophysics, University of Leipzig, D-04103 Leipzig, Germany Received February 27, 2004; Revised Manuscript Received May 11, 2004

Biocomposite thin films assembled on colloidal particles by means of layer-by-layer adsorption have been suggested as drug carriers and diagnostic devices. Protamine (PRM)/dextransulfate (DXS) and protamine/ bovine serum albumine (BSA) multilayers were fabricated on colloidal silica and subsequently investigated by means of fluorescence activated cell sorting (FACS) and microelectrophoresis. Fluorescein labeled polyelectrolytes were embedded at different positions in the multilayers as a marker for layer growth. FACS showed that PRM and DXS formed regular growing stable multilayers, yet adsorbed PRM can be nevertheless exchanged with PRM in solution during layer formation and also after the multilayer formation has been completed. Up to 90% of the PRM pool was available for exchange. PRM together with BSA as demonstrated by SFM did not form multilayers under the applied conditions although the ζ-potential, commonly used as an indicator for stepwise adsorption, observed characteristic alternations. The capability of bound PRM to exchange with PRM in solution is attributed to its relatively small size. The demonstrated exchange may have importance in designing multilayers with smart release features. Furthermore, FACS proved to be a rather suitable means to quantify the aggregation behavior during coating and washing. Singulets, doublets, triplets, and aggregates of higher order could be clearly resolved. The aggregation of particles coated with PRM/DXS layers was higher than that of silica particles coated with PAH/PSS layers. In the first case about 50% of all recorded events are attributed to aggregats, while the PAH/PSS coating produced only about 10% aggregates. Introduction Thin organic films can be assembled layer-wise onto a variety of surfaces by means of the alternating deposition of polyanions and polycations. This novel technique, called layer-by-layer (LbL) deposition, was first introduced by Decher and co-workers in 1991.1 Later the LbL approach was transferred to colloidal particles.2 Silica, melamin formaldehyde, polystyrene latex particles, as well as biological cells have been extensively used as substrates for the film assembly.3,4 The layer-wise deposition allows the fabrication of quite complex films. A variety of functionalities may be introduced into the film. One of the most attractive features of this technique is that it provides control over the film composition in the perpendicular direction. When after coating the template core can be removed, hollow polyelectrolyte shells are obtained.5 These hollow capsules may find applications as microreactors, sensing devices, or carriers.6 For example, drugs can be encapsulated and released afterward.7 Nature as well as polymer chemistry provide an enormous wealth of materials potentially suitable for coating. Yet, only very few couples have been studied so far regarding their properties concerning LbL formation. A well established and characterized polyelectrolyte pair is polyallylamine (PAH)/ polystyrene sulfonate (PSS). It easily formes stable multi* To whom correspondence should be addressed. E-mail: reichls@ medizin.uni-leipzig.de.

layers but is not biodegradable, or biocompatible, which is a serious drawback in view of potential biomedical applications. In contrast to the fabrication of multilayers on flat substrates, the coating of colloids is more challenging. One of the key problems is to avoid particle flocculation during the multiple adsorption/washing cycles. Unfortunately, if polyelectrolyte couples other than PAH/PSS are used, one often encounters colloidal stability and layer growth problems. This is especially true when biopolymers such as proteins are employed. Although, a few experimental results demonstrating the successful assembly of proteins, enzymes, nucleic acids, or carbohydrates in polyelectrolyte multilayers have been published.8-11 However, the formation of LbL multilayers involving biopolymers is far from being properly understood. For example, the behavior of proteins during assembly is not comparable with that of synthesized polyelectrolytes. A different charge density, the amphiphilic character, and the tertiary structure of proteins may cause assembly problems and aggregation. A nonlinear layer growth12,13 and a lack of stability of the formed multilayers were described.14 Clearly, a deeper understanding of the self-assembly process is necessary to improve the LbL technology. This requires suitable methods to be established to study layer growth and properties on colloidal particles. Most characterization techniques applied in the past, such as IR, fluorescence, or Raman spectroscopy as well as microelec-

10.1021/bm049875m CCC: $27.50 © 2004 American Chemical Society Published on Web 06/18/2004

Protamine Exchange in Multilayers

trophoresis, are bulk techniques.10,15-18 They are not capable of recording flocculation. Often even a weak degree of aggregation between particles seriously interferes with these spectroscopic techniques caused by greatly enhanced scattering. Single particle light scattering has been quite successful in determining the layer thickness and growth, as well as the aggregation behavior, but it cannot be used routinely, since commercially available machines do not provide sufficient sensitivity and precision. Another problem is that data analysis requires a priori knowledge of the refractive index in the layer, which is a function of the water content of the layer, polymer density, etc. Neutron scattering provides, in principle, an accurate measure of the layer thickness, but is expensive and slow. In addition, rather small particles have to be used to avoid sedimentation. A promising alternative is flow cytometry. It is routinely used as a diagnostic tool in medicine and cell physiology. It is a single particle technique permitting screening of large numbers of particles within reasonable time frames.19 Standard flow cytometry devices are equipped with two scattering and three fluorescence channels. We used this technique before to follow the layer buildup on cells as well as lipid layer deposition on LbL multilayers.4,20 The main advantage of flow cytometry is that it allows us in parallel to follow layer formation and growth as well as to record the degree of flocculation. The drawback is that fluorescent labels have to be used. In view of future applications, special attention was given to coating with biopolymers. Layer growth and stability of the couples protamine/dextransulfate, introduced in ref 21, and protamine/ bovine serum albumine were investigated, and their behavior was compared with the well established polyallylamine/polystyrene sulfonate system. Protamine is a higly basic peptide which contains 75% of arginine. It is available in large quantities from fish sperm. One of its biological functions is to ensure compaction of DNA.22,23 Regarding the possible use of LbL multilayers containing protamine, it is interesting to note that protamine slows down the release of insulin when insuline/protamine complexes are used.24 Both these features of protamine are based on polyelectrolyte complex formation. Hence, protamine should be a rather suitable candidate for LbL formation with a wide range of anionic biopolymers. Materials and Methods Materials. Silica particles Ø 3.09 µm ( 0.25 µm were purchased from MicroParticles GmbH (Berlin, Germany). Fluorescein isothiocyanate (FITC) and protamine sulfate (PRM) from herring, dextran sulfate sodium salt (DXS) Mw 500 000 g/mol, bovine serum albumine (BSA), FITC-labeled bovine serum albumine (10 mol FITC/mol BSA), polyallylamine hydrochloride (PAH), polystyrene sulfonate (PSS), as well as sephadex G 25 chromatograpy gel are products from Sigma (Deisenhofen, Germany). All other chemicals were obtained from Fluka (Neu-Ulm, Germany). Coating of Silica Particles. All polyelectrolyte multilayers were assembled on silica particles at neutral pH in 0.1 M NaCl. A 150 µL silica particle suspension of a 5% stock

Biomacromolecules, Vol. 5, No. 4, 2004 1581

solution was used for coating. In every coating step, the polyelectrolyte concentration was 1 mg/mL. Protamine sulfate and dextran sulfate were adsorbed in alternating order for 10 min with gentle shaking. Particles were spun down at 500×g for 2 min and washed three times with distilled water removing the remaining in the bulk polyelectrolyte. An analogous procedure was carried out using PRM/BSA and PAH/PSS as polyelectrolyte couples. Certain coating steps involved adsorption of FITC- PRM, FITC-PAH, and FITC-BSA, respectively, employed in concentrations of 1 mg/mL. Particles for FACS and ζ-potential measurements were taken directly from the stock suspension followed by appropriate dissolution. Polyelectrolyte Multilayer Formation on Glass Substrates. Polyelectrolyte multilayers were assembled on glass micoscope slides at pH 7 in 0.1 M NaCl by alternate adsorption of dextran sulfate (DXS) and protamine (PRM) as well as protamine and bovine serum albumine (BSA). The polymer concentration was 2 mg/mL. The glasses were cleaned by using the RCA method.25 Labeling of Polyelectrolytes. Fluorescein isothiocyanate (FITC) was covalently bound to PRM as well as PAH in carbonate buffer pH 8.5. FITC, which was not utilized in the coupling reaction, was removed by gel chromatography in a sephadex G25 column. The purified FITC-polyelectrolyte was lyophylized and stored at 4 °C until further use. The concentration of the bound fluorescence dye was determined using a UV/vis spectrometer Cary 50 (Varian Instruments, U.S.A.) using the molar extinction coefficient 490 ) 80 900 M-1 cm-1. PRM has one fluorescence label per molecule, whereas the ratio of PAH was 6:1 which compares to one label per 125 monomers. Flow Cytometry. The fluorescence intensity distributions of silica particles coated with PRM /DXS, PRM/BSA, and PAH/PSS, respectively, were recorded with a flow cytometer (FACSCalibur, Becton Dickinson, U.S.A.). The fluorescence was excited with an argon laser at 488 nm. In every measurement, 10 000 particles were detected. The data were analyzed with the FACSwinmdi 2.8 software. Microelectrophoresis. The electrophoretic mobility was determined with a Zetasizer 4 (Malvern instruments, U.K,) by taking the average of 10 measurements. The mobility u was converted into a ζ-potential by using the relation ζ ) uη/, where η is the viscosity of the solution and  is the permittivity. The mobility of the coated particles was measured in 1 mmol/L KCl at neutral pH. Fluorescence Spectroscopy. For fluorescence spectroscopic investigations, 3 µm silica particles were coated with five bilayers of the couple [PRM/DXS], where four PRM layers are FITC-labeled. After finishing the sample coating and three washing steps, the particle suspension was stored at 4 °C in water containing NaN3 for inhibiting bacterial contamination. After 1 h, 4 h, 16 h, 29 h, 2 days, 4 days, and 8 days, 150 µL of the supernatant were taken for fluorescence spectroscopic investigations and replaced by new NaN3 containing water. Before every removal of supernatant, the dispersions were centrifugated for 2 min at 500×g. After 8 days, however, the supernatant was completely removed. It was replaced by a solution of 1 mg/mL

1582

Biomacromolecules, Vol. 5, No. 4, 2004

PRM in 0.1 M NaCl and incubated for 16 h. Fluorescence spectroscopy was carried out at a Fluoromax 2 fluorescence spectrometer (Spex Ind., Inc., U.S.A.). The excitation was set at 495 nm (excitation and emission slits of 5 nm width) were used. Fluorescence spectra were detected in the range of 500-580 nm. Scanning Force Microscopy (SFM). Contact mode scanning force microscopy images of polyelectrolyte multilayers were acquired in air at room temperature using a Molecular Force Probe 3D -IO (Inverted Optical) Atomic Force Microscope (Asylum Research, Santa Barbara, CA) mounted on an Olympus IX.70 Gold coated triangular Thermomicroscope (now Veeco) Ultralevers of 0.01 N/m and 100 µm long (nominal resonance frequency of 7 kHz) are used for contact mode imaging. The images were acquired in height mode (topography) at a scan rate of 0.5-1 Hz and sampled from different macroscopically separated areas. They were processed using the MFP-3D software running under Igor Pro (Wavemetrics). Results and Discussion The LBL assembly from biopolymers onto colloidal particles can be followed by means of flow cytometry (FACS). The silica templates, which have been used, had a size quite similar to that of cells, which the FACS device is optimized for. The particular advantage of the layer-by-layer technique is the flexibility in the design of layer composition and sequence. Fluorescence labeled polyelectrolyte species were employed as markers. This provides the opportunity to directly observe the assembling process at every coating step on a single particle level. In addition, the scattering allows for a detection and quantification of particles aggregates, thus providing a means to evaluate the multilayer deposition protocol. The polyelectrolyte couples protamine sulfate (PRM)/ dextran sulfate (DXS) and protamine sulfate/bovine serum albumine (BSA) were chosen for coating and compared with the well-established polyelectrolyte system PAH/PSS. All coating steps were performed at neutral pH. Protamine sulfate was used as positive polyelectrolyte, whereas dextran sulfate and bovine serum albumine with an isoelectric point of pH ) 4.7 were selected as model negatively charged biopolymers. It is commonly assumed that recharging of the surface upon adsorption of polyelectrolytes is the prerequisite for the ongoing layer buildup. The formation of the polyanion/ polycation complex releases small counterions which is consistent with an entropy increase being essentially the driving force for layer buildup. Hence it is quite useful to follow the multilayer formation also by means of microelectrophoresis experiments. The surface of silica particles is negatively charged at neutral pH, as evident from the ζ-potential of -53 mV for the bar particle surface shown in Figure. 1. Hence, the stepwise adsorption of polyelectrolytes was started with the positively charged protamine sulfate. As expected, the ζ-potential values observe the characteristic alternating behavior upon subsequent layer deposition to-

Hiller et al.

gether with either DXS or BSA. Although in any case the ζ-potential was negative for the top layer being an polyanion, DXS corresponded to ζ-potentials of -40 to -50 mV and BSA to about -20, respectively,10 it is remarkable that the ζ-potential did not reach positive values when PRM was the top layer. It leveled off slightly below 0 mV. One of the possible causes might be simply an incomplete coverage of the surface with protamine. Such a simple explanation, however, would be in contradiction with regular layer growth. It is difficult to imagine that, after PRM has only adsorbed to a relatively small extent, a large amount of the polyanion could be adsorbed afterward. To understand this peculiar behavior, it is worth remembering that the ζ-potential does not necessarily reflect only the charge density of the top layer. Charges buried deeper in the multilayer may also contribute to the ζ-potential provided they are accessible to the hydrodynamic flow induced by the applied external electric field. This would be the case if there are considerable large pores or grooves in the surface. Another point is that charged groups extending out from the interface may contribute more to the measured ζ-potential than those adjacent directly to the surface. This is a consequence of their enhanced accessibility to the electroosmotic flow.3 In this context, it is important to remember that the molecular weight of the two polyelectrolyte components is quite different. Protamine sulfate has a molecular weight (Mw) of only about 5000 Da. This is 100-fold smaller than the Mw of the particular dextrane sulfate used, Mw ) 500 000 g/mol and 10-fold smaller than BSA with a molecular weight of 46 000 Da, respectively. Protamine can thus possibly easily intercalate into spaces between the macromolecules of the respective polyanion layers. Some segments of the larger polyanion layers may still protrude out into solution even after saturation adsorption of PRM. Nonetheless, the alterations of the ζ-potential evidence that PRM as well as DXS and BSA were adsorbed in a regular manner, although, on the other hand, the ζ-potential alterations alone cannot prove a regular stepwise layer growth. Equally well each incoming polyanion layer may have led to a desorption of PRM thus reproducing a negative ζ-potential without effective layer growth. This would be the case if the stability of the PRM layer was not large enough. Instead of layer-wise complex formation on the silica particles by adsorption, the next polyelectrolyte may have removed the deposited counterion forming complexes in solution instead of adsorbing onto it. To have an unambiguous proof of layer growth, other methods capable of detecting, for example, the signal of a characteristic group within the layer which increases with layer growth have to be applied. IR, Raman, or UV/vis spectroscopy have been used.15 17,26 Fluorescence spectroscopy is another widely used technique. Fluorescence labeled polyelectrolytes can be easily prepared. Subsequently, the layer buildup can be detected by the increase of the fluorescence intensity. Although fluorescence spectroscopy was used with colloidal systems,10 it is the rather intense scattering which interferes with the interpretation of results. Since the intensity of scattering scales with r4 ∼ r,6 depending on the particle size, where r is the particle radius,

Biomacromolecules, Vol. 5, No. 4, 2004 1583

Protamine Exchange in Multilayers

Figure 3. Histogram corresponding to Figure 2 of flow cytometry measurements of 3 µm in diameter silica particles coated with [PRM/ DXS]6 layers. See Figure 2. Table 1. Data Corresponding to the Histogram in Figure 3

Figure 1. ζ-potential as a function of layer number of silica particles coated with (a) PRM/DXS and (b) PRM/BSA. Measurements in 1 mM KCl, 20° C. Mean and standard deviation of 10 measurements are shown.

Figure 2. Dot plot of flow cytometry measurements of 3 µm in diameter silica particles coated with [PRM/DXS]6 layers. The 3th and 5th layer contain FITC-labeled PRM. Number of events: 10000. R1, R2, and R3 denote regions of interest (ROI). They represent singlets, doublets, and triplets of particles, respectively.

even a slight aggregation may heavily interfere with the intensity determination. In addition, the quantity of particles has to be precisely controlled. Therefore, we used flow cytometry (FACS) as described above to follow layer growth. Figure 2 shows an example of flow cytometry data of coated particles where aggregation is observed.

region

events

% of total events

all R1 R2 R3

10 000 4655 1253 719

100 46.55 12.53 7.19

mean of fluorescence intensity

variation coefficient

40.34 18.25 35.55 51.48

35.85 6.57 3.41 2.36

Particles coated with six layer pairs of the couple PRM/ DXS, where two PRM layers have been deposited from FITC-labeled PRM, are shown. The dot plot displays the FITC fluorescence intensity versus their scattering intensity in forward direction. Every displayed dot is related to the properties of one particular particle concerning its size and fluorescence intensity. The forward scattering intensity is a measure of the particle size. As it can be easily seen, the particle suspension contains particles with different scattering and fluorescence intensities, although all particles are from one and the same stock, which had been subjected to the layer fabrication protocol. Three main populations marked as R1, R2, and R3 (regions of interest ) ROI) can be identified. It is also obvious that the particle size correlates with the fluorescence intensity. The fluorescence intensity histogram corresponding to the data shown in Figure 2 is plotted in Figure 3. The histogram clearly resolves the subpopulations of particles already identified in Figure 2. The histogram data are provided in Table 1. The largest amount, almost 50% of all particles, is found in the population R1 with a mean fluorescence value of about 18 (in arbitrary units). The mean fluorescence of R2 has about twice as much fluorescence, and R3 shows almost 3-fold fluorescence compared with R1. This demonstrates that FACS measurements allow for resolving particle aggregation occurring during the layer-by-layer self-assembly. Indeed, an aggregate consisting of two particles has twice the fluorescence value than a single particle so that R1 corresponds to single particles, R2 to doublets, and R3 to triplets, respectively. All other signals correspond to aggregates of even higher order.

1584

Biomacromolecules, Vol. 5, No. 4, 2004

Hiller et al.

Figure 4. Fluorescence intensity histograms of 3 µm in diameter silica particles coated with [PRM/DXS]6. Each histogram represents 10000 particles. The sequence of coating with FITC-PRM is seen in Table 2. Table 2. Sequence of Coating with FITC-PRM layer no. 0 1 3 5 7 9 11

1

2

sample no. 3

4

FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM

5 FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM FTIC-PRM

In a second set of experiments, we applied the FACS technique to evidence the layer growth. Figure 4 displays various histograms of particle fluorescence. The particles were coated with the layer couple PRM/DXS. Six different samples having 12 layers in total were measured. Sample 0 did not contain any fluorescent labeled PRM. Nevertheless, the observed fluorescence at very low intensities (note, that the scale is logarithmic) may be due to self-fluorescence or represent a crosstalk with scattered light visible at the employed large amplification of the device. In sample 1, only the third layer is FITC labeled in sample 2, the third and fifth layer, the following samples contain 3, 4, and 5 FITCPRM layers, respectively. All fluorescence distributions have a similar shape, the peaks for singlets, doublets, triplets and higher aggregates are resolved. The fluorescence intensity increases with the number of fluorescent layers. However, the fluorescence intensity corresponding to five layers of labeled polyelectrolyte is more than 100-fold as much as for one layer, where it should be only 5-fold as large. To understand this behavior, further experiments were conducted involving also PAH/PSS and PRM/BSA multilayers on silica particles. The idea was to find out why the fluorescence continuosly decreased with subsequent adsorption of nonfluorescent layers. In Figure 5a-c, the observed fluorescence intensity values are shown as a function of the number of embedded FITClabeled layers. Black squares correspond to samples coated with six layer pairs PRM/DXS, PAH/PSS, and PRM/BSA, that is after completing the procedure of layer fabrication. The open circles represent the fluorescence intensities of the samples which were directly measured after the last fluorescent layer has been adsorbed, that is after the “italic” step

Figure 5. Fluorescence intensity as a function of layer number. (a) Silica particles coated with PRM/DXS, (b) PAH/PSS, (c) PRM/BSA. Scheme of coating of all samples is given in Table 3. In all three panels, the full squares correspond to particles with 12 layers in total (bold fields in Table 3). Open circles represent samples where coating was stopped after adsorption of one, two, three, four, or five FITClabeled polyelectrolyte layers (italic fields in Table 3).

in Table 3. Naturally, the total number of layers in these samples varies between 3 and 11. Each of the investigated three polyelectrolyte couples shows a senario concerning layer growth, which is different from the two others. The FACS data of the PRM/DXS are shown in Figure 5a. The fluorescence intensity increases with

Biomacromolecules, Vol. 5, No. 4, 2004 1585

Protamine Exchange in Multilayers Table 3. Scheme of Layer Deposition Protocol for FACS Measurementsa layer no. 1 2 3 4 5 6 7 8 9 10 11 12

1

2

sample no. 3

4

(a) Couple Protamine Sulfat/Dextran Sulfate PRM PRM PRM PRM DXS DXS DXS DXS FITC-PRM FITC-PRM FITC-PRM FITC-PRM DXS DXS DXS DXS PRM FITC-PRM FITC-PRM FITC-PRM DXS DXS DXS DXS PRM PRM FITC-PRM FITC-PRM DXS DXS DXS DXS PRM PRM PRM FITC-PRM DXS DXS DXS DXS PRM PRM PRM PRM DXS DXS DXS DXS

5 PRM DXS FITC-PRM DXS FITC-PRM DXS FITC-PRM DXS FITC-PRM DXS FITC-PRM DXS

1 2 3 4 5 6 7 8 9 10 11 12

(b) Couple Polyallylamine Hydrochloride/Polystyrene Sulfonate PAH PAH PAH PAH PAH PSS PSS PSS PSS PSS FITC-PAH FITC-PAH FITC-PAH FITC-PAH FITC-PAH PSS PSS PSS PSS PSS PAH FITC-PAH FITC-PAH FITC-PAH FITC-PAH PSS PSS PSS PSS PSS PAH PAH FITC-PAH FITC-PAH FITC-PAH PSS PSS PSS PSS PSS PAH PAH PAH FITC-PAH FITC-PAH PSS PSS PSS PSS PSS PAH PAH PAH PAH FITC-PAH PSS PSS PSS PSS PSS

1 2 3 4 5 6 7 8 9 10 11 12

(c) Couple Protamine Sulfat/Bovine Serum Albumine PRM PRM PRM PRM PRM FITC-BSA FITC-BSA FITC-BSA FITC-BSA FITC-BSA PRM PRM PRM PRM PRM BSA FITC-BSA FITC-BSA FITC-BSA FITC-BSA PRM PRM PRM PRM PRM BSA BSA FITC-BSA FITC-BSA FITC-BSA PRM PRM PRM PRM PRM BSA BSA BSA FITC-BSA FITC-BSA PRM PRM PRM PRM PRM BSA BSA BSA BSA FITC-BSA PRM PRM PRM PRM PRM BSA BSA BSA BSA BSA

a The samples in the highlighted fields were used for FACS measurements (compare Figure 5a-c).

the number of FITC-PRM layers adsorbed. This is generally consistent with a stepwise layer growth of PRM and DXS. However, the rate of increase of fluorescence is dependent on how the measuring protocol was conducted. Whereas the fluorescence increases almost linearly (with the exception of the very first PRM layer) when the fluorescence of the particles was recorded directly after coating of the last fluorescent PRM layer, a different result was obtained when the fluorescence was measured after all 12 layers were finally adsorbed. In the latter situation, a nonlinear growth of fluorescence was observed. The fluorescence increased slightly for the first three samples, and only the samples having four or more FITC-PRM layers showed a steep increase of fluorescence with layer number. How this can be explained? Fluorescence quenching caused by overlapped polyelectrolyte layers can be ruled out, also a desorption of polyelectrolyte layers is quite unlikely. Both possibilities would not explain the differences between the two measuring schemes, because finally for the case of five fluorescent layers almost the same fluorescence was observed for both protocols. The only reasonable explanation is that the fluorescence labeled PRM can be exchanged for nonlabeled PRM offered in excess during the subsequent coating cycles. For example, consider a situation of three FITC-PRM layers

Figure 6. Scanning force microscopy (SFM) image of (PRM/BSA)3 polyelectrolyte multilayers on glass.

paired with DXS. The next two PRM layers were then fabricated by means of adsorbing nonlabeled PRM. Obviously, this was sufficient to almost completely, at least in logarithmic scale, exchange the FITC-PRM for nonlabeled PRM. It has to be stressed that this took place not only for the top layer but also for PRM layers within the multilayer. Such a behavior was not observed with FITC-labeled PAH (Figure 5b). Here the two different measuring protocols revealed the same fluorescence increase as a result of a regular layer growth without any indication of exchange of PAH. The small difference in fluorescence intensity can certainly be attributed either to a slightly different sensitivity of the FACS machine or to minute differences in the coating efficiency. The results obtained with the PRM/BSA system (Figure 5c) are quite different with both the PRM/DXS and the PAH/ PSS system. There was no fluorescence growth with layer number; notwithstanding, the ζ-potential showed the typical oscillations characteristic for polyelectrolyte adsorption. Probably BSA will be removed by means of interaction with bulk PRM molecules during the coating. Although certainly not fully compatible with the coating of silica spheres, nevertheless, SFM imaging of PRM/BSA multilayers, adsorbed on glass, might help us to understand the observed lack of fluorescence growth. Figure 6 provides the SFM image of PRM/BSA multilayers of a glass surface which has undergone three absorption cycles with protamine and BSA. It is obvious that, instead of a multilayer, polyelectrolyte complexes of different size and shape have been produced. This together with the absence of fluorescence growth proves that a regular growth of a multilayer did not occur with the BSA/PRM couple. Lack of layer growth composed only of globular proteins has been reported by Lvov et al.27 Dissolution of polyelectrolyte multilayers in the presence of bulk polyelectrolyte molecules is well-known.28 To verify the exchange between bound and free PRM a special experiment was designed. Silica particles were coated with PRM/DXS[FITC-PRM/DXS]4. After compleating the polyelectrolyte absorption, the multilayer covered particles were stored in water and shaken from time to time. 150 µL of the supernatant was removed after 1 h, 4 h, 16 h, 29 h,

1586

Biomacromolecules, Vol. 5, No. 4, 2004

Figure 7. Fluorescence intensities measured at 518 nm emission of supernatants of 3 µm in diameter silica particles coated with PRM/ DXS [FITC-PRM/DXS]4 after various incubation protocols. Supernatants were collected after 1 h, 4 h, 16 h, 29 h, 2 days, 4 days, and 8 days incubation in water. The last column displays the maximum fluorescence intensity of the supernatant following 16 h incubation in the presence of PRM.

48 h, and 4 days and replaced with water. After 8 days, the supernatant was replaced with 300 µL of PRM (1 mg/mL) solution and incubated for another 16 h. The various supernatants were investigated by means of fluorescence spectroscopy. The result is presented in Figure 7. The fluorescence intensity of all supernatants was negligibly small regardless of the storage time of the sample. At most a fluorescence intensity of the order of 105 arbitrary units was observed. This proves that itself did not desorb over a larger period of time. When, however, water was replaced by PRM, significant fluorescence was found in the supernatant (Figure 7). This is a clear proof that PRM in solution was exchanged with the labeled one in the interior of the multilayer. The binding energy of PRM and DXS may be weaker than between PAH and PSS allowing for exchange but not for decomposition of the layer. Possibly the small size of PRM plays an important role. It facilitates diffusion into the layer interior and it also makes transient desorption more likely to occur. It is further remarkable that during DXS adsorption a loss of fluorescence related to a possible desorption of PRM was not observed. This proves the stability of the layer itself and rules out the possibility that the increased ionic strength alone leads to layer decomposition. Concerning details of the exchange process and the time necessary for exchange, more experiments clearly have to be done. From the coating experiments, it follows that the exchange is fast enough to take place at least partly within the 10 min allowed for PRM to adsorb. An incubation of 16 h was, on the other hand, not long enough to completely exchange FITC-PRM. FACS measurements still revealed the presence of some bound FITC-PRM (data not shown). It is also likely that the deeper the PRM is located within the multilayer the more slow or even impossible the exchange should be. Another issue is that, strictly speaking, FITCPRM and PRM may have a slightly different binding behavior due to the presence of one FITC per PRM. Comparing the aggregation behavior of silica particles coated with PAH/PSS with these covered with PRM/DXS and PRM/BSA as investigated by flow cytometry, we found remarkable differences. Particles coated with PRM/DXS exhibit a stronger aggregation than those coated with PAH/

Hiller et al.

PSS. A washing processes with 0.1 M NaCl solutions during preparation enhanced the aggregation in the PRM/DXS system. Whereas we found singlets, doublets, triplets, and higher aggregates in the PRM/DXS suspensions, the PAH/ PSS particle population contained mostly singlets and only 10% of doublets (data not shown). The particles treated with PRM and BSA showed nearly no aggregation. Currently, work is under way to detail the aggregation behavior during coating. The observed exchange of labeled PRM is quite important concerning the application of multilayer coated colloids and capsules as release systems. If the exchange of PRM with FITC-PRM is possible, it is quite conceivable that exchange between different polyions might be possible as well. Clearly, more systematic studies are necessary to understand the details of this process. As above-mentioned, systems exist which even decompose in the presence of bulk polyelectrolytes. Maybe in the case of PRM with BSA, the rate of dissolution is faster than layer formation. Then growth is not possible. This may also have practical importance regarding required biodegradability. Conclusion It has been demonstrated with PRM, a small basic protein, that constituents of polyelectrolyte multilayers can be exchanged with bulk compounds. This is an important finding regarding possible applications of multilayer coated particles as release systems. It should be possible by properly designing the multilayer to tune the release over longer time frames. Flow cytometry has been shown to be a valuable technique to investigate details of the multilayer growth. By means of FACS measurements many problems associated with scattering in bulk suspensions can be overcome. In addition, the aggregation behavior, which is rather important for practical applications, can be followed. Acknowledgment. This work was supported by a grant from the German Ministry of Education and Science, BMBF 0312011C. Dr. M. Wo¨tzel, Institute for Clinical Imunology and Transfusion Medicine of the University of Leipzig, is acknowledged for providing the FACS facilities. References and Notes (1) Decher, G. Fuzzy Nanoassemblies: Toward Layered Polymeric Multicomposites. Science 1997, 277, 1232. (2) Sukhorukov, G. B.; Donath, E.; Lichtenfeld, H.; Knippel, E.; Knippel, M.; Budde, A.; Mohwald, H. Layer-by-Layer Self-Assembly of Polyelectrolytes on Colloidal Particles. Colloids Surf. A-Physicochem. Eng. Aspects 1998, 137, 253. (3) Donath, E.; Walther, D.; Shilov, V. S.; Knippel, E.; Budde, A.; Lowack, K.; Helm, C. A.; Mohwald, H. Nonlinear Hairy Layer Theory of Electrophoretic Fingerprinting Applied to Consecutive Layer-by-Layer Polyelectrolyte Adsorption Onto Charged Polystyrene Latex Particles. Langmuir 1997, 13, 5294. (4) Neu, B.; Voigt, A.; Mitlohner, R.; Leporatti, S.; Gao, C. Y.; Donath, E.; Kiesewetter, H.; Mohwald, H.; Meiselman, H. J.; Baumler, H. Biological Cells As Templates for Hollow Microcapsules. J. Microencapsulation 2001, 18, 385. (5) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davies, S. A.; Mohwald, H. Novel Hollow Polymer Shells by Colloid-Templated Assembly of Polyelectrolytes. Angew. Chem., Int. Ed. Engl. 1998, 37, 2202.

Protamine Exchange in Multilayers (6) Dahne, L.; Leporatti, S.; Donath, E.; Mohwald, H. Fabrication of Micro Reaction Cages With Tailored Properties. J. Am. Chem. Soc. 2001, 123, 5431. (7) Qiu, X. P.; Donath, E.; Mohwald, H. Permeability of Ibuprofen in Various Polyelectrolyte Multilayers. Macromol. Mater. Eng. 2001, 286, 591. (8) Szyk, L.; Schwinte, P.; Voegel, J. C.; Schaaf, P.; Tinland, B. Dynamical Behavior of Human Serum Albumin Adsorbed on or Embedded in Polyelectrolyte Multilayers. J. Phys. Chem. B 2002, 106, 6049. (9) Caruso, F.; Schuler, C. Enzyme Multilayers on Colloid Particles: Assembly, Stability, and Enzymatic Activity. Langmuir 2000, 16, 9595. (10) Caruso, F.; Mohwald, H. Protein Multilayer Formation on Colloids Through a Stepwise Self-Assembly Technique. J. Am. Chem. Soc. 1999, 121, 6039. (11) Lu, C.; Luo, C.; Cao, W. Fabrication of Ultrathin Films Based on Chitosan and Bovine Serum Albumin and Their Stability Studied With the Radio-Labeled Methodol. Colloids Surf. B: Biointerfaces 2002, 25, 19. (12) Picart, C.; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.; Schaaf, P.; Voegel, J. C.; Lavalle, P. Molecular Basis for the Explanation of the Exponential Growth of Polyelectrolyte Multilayers. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12531. (13) Boulmedais, F.; Ball, V.; Schwinte, P.; Frisch, B.; Schaaf, P.; Voegel, J. C. Buildup of Exponentially Growing Multilayer Polypeptide Films With Internal Secondary Structure. Langmuir 2003, 19, 440. (14) Ladam, G.; Schaaf, P.; Decher, G.; Voegel, J. C.; Cuisinier, F. J. G. Protein Adsorption Onto Auto-Assembled Polyelectrolyte Films. Biomol. Eng. 2002, 19, 273. (15) Cui, X. Q.; Pei, R. J.; Wang, X. Z.; Yang, F.; Ma, Y.; Dong, S. J.; Yang, X. R. Layer-by-Layer Assembly of Multilayer Films Composed of Avidin and Biotin-Labeled Antibody for Immunosensing. Biosens. Bioelectron. 2003, 18, 59. (16) Fang, M.; Grant, P. S.; McShane, M. J.; Sukhorukov, G. B.; Golub, V. O.; Lvov, Y. M. Magnetic Bio/Nanoreactor With Multilayer Shells of Glucose Oxidase and Inorganic Nanoparticles. Langmuir 2002, 18, 6338. (17) Brynda, E.; Houska, M. Multiple Alternating Molecular Layers of Albumin and Heparin on Solid Surfaces. J. Colloid Interface Sci. 1996, 183, 18.

Biomacromolecules, Vol. 5, No. 4, 2004 1587 (18) Schwinte, P.; Voegel, J. C.; Picart, C.; Haikel, Y.; Schaaf, P.; Szalontai, B. Stabilizing Effects of Various Polyelectrolyte Multilayer Films on the Structure of Adsorbed/Embedded Fibrinogen Molecules: An ATR-FTIR Study. J. Phys. Chem. B 2001, 105, 11906. (19) Boeck, G. Current Status of Flow Cytometry in Cell and Molecular Biology. Int. ReV. Cytol. 2001, 204, 239. (20) Moya, S.; Donath, E.; Sukhorukov, G. B.; Auch, M.; Baumler, H.; Lichtenfeld, H.; Mohwald, H. Lipid Coating on Polyelectrolyte Surface Modified Colloidal Particles and Polyelectrolyte Capsules. Macromolecules 2000, 33, 4538. (21) Balabushevich, N. G.; Tiourina, O. P.; Volodkin, D. V.; Larionova, N. I.; Sukhorukov, G. B. Loading the Multilayer Dextran Sulfate/ Protamine Microsized Capsules With Peroxidase. Biomacromolecules 2003, 4, 1191. (22) Brewer, L.; Corzett, M.; Lau, E. Y.; Balhorn, R. Dynamics of Protamine 1 Binding to Single DNA Molecules. J. Biol. Chem. 2003, 278, 42403. (23) Dunne, M.; Bibby, D. C.; Jones, J. C.; Cudmore, S. Encapsulation of Protamine Sulphate Compacted DNA in Polylactide and Polylactide-Co-Glycolide Microparticles. J. Controlled Release 2003, 92, 209. (24) Yip, C. M.; Brader, M. L.; Frank, B. H.; DeFelippis, M. R.; Ward, M. D. Structural Studies of a Crystalline Insulin Analogue Complex With Protamine by Atomic Force Microscopy. Biophys. J. 2000, 78, 466. (25) Graf, K.; Riegler, H. Molecular Adhesion Interactions Between Langmuir Monolayers and Solid Substrates. Colloids Surf. A-Physicochem. Eng. Aspects 1998, 131, 215. (26) Li, W.; Xian, M.; Wang, Z.; Sun, C.; Zhao, M. Alternate Deposition of Horseradish Peroxidase and Bipolar Pyridine Salt on the Solid Surface to Prepare Electrocatalytically Active Enzyme Thin Film. Thin Solid Films 2001, 386, 121. (27) Lvov, Y. M. Electrostatic Layer-by-Layer Assembly of Proteins and Polyions. In Protein Architecture: Interfacing Molecular Assemblies and Immobilization Biotechnology; Lvov, Y. M., Mohwald, H., Eds.; Dekker: New York, 2000; Chapter 6. (28) Kovacevic, D.; van der Burgh, S.; de Keizer, A.; Stuart, M. A. C. Kinetics of Formation and Dissolution of Weak Polyelectrolyte Multilayers: Role of Salt and Free Polyions. Langmuir 2002, 18, 5607.

BM049875M