Polyion-Induced Aggregation of Lipidic-Coated Solid Polystyrene

May 17, 2008 - Universita' di Roma “La Sapienza” and Research Center SOFT-INFM-CNR. , ‡. Universita' di Perugia and Research Center SOFT-INFM-CN...
0 downloads 0 Views 104KB Size
6044

Langmuir 2008, 24, 6044-6049

Polyion-Induced Aggregation of Lipidic-Coated Solid Polystyrene Spheres: The Many Facets of Complex Formation in Low-Density Colloidal Suspensions S. Zuzzi,‡ C. Cametti,*,† and G. Onori‡ Dipartimento di Fisica, UniVersita’ di Roma “La Sapienza”, Piazzale A. Moro 5, I-00185-Rome, Italy, and Research Center SOFT-INFM-CNR, Unita’ di Roma 1, and Dipartimento di Fisica, UniVersita’ di Perugia and Research Center SOFT-INFM-CNR, Unita’ di Perugia and CEMIN (Centro Eccellenza Materiali InnoVatiVi Nanostrutturati), Via A. Pascoli, I-06123 Perugia, Italy ReceiVed February 21, 2008. ReVised Manuscript ReceiVed March 17, 2008 We have investigated the formation of a cluster phase in low-density colloidal systems formed by charged solid charged particles stuck together by an oppositely charged polyion. In analogy with what we have previously observed in the case of soft charged particles, also in this case the same basic phenomenology occurs, consisting of the presence of the two well-known characteristic phenomena of this class of colloids, that is, reentrant condensation and charge inVersion. With the aim of comparing the cluster formation in both soft and solid charged particles, we have, in previous works, employed cationic liposomes (soft particles, lipidic vesicles built up by dioleoyltrimethylammonium propane [DOTAP] lipid) and, in the present work, polystyrene particles (solid particles) covered by the same lipidic bilayer as the one of the soft particles, so that the two classes of particles share electrostatic interactions of the same nature. These charged particle clusters, where the single aggregating particles maintain their integrity without undergoing a structural rearrangement, join to a class of different aggregated structures (lamellar or inverse hexagonal phases) observed as well in the polyion-induced aggregation of oppositely charged mesoscopic particles, in particular, lipidic vesicles. Our results show that the formation of relatively large, equilibrium clusters of particles which maintain their integrity, stuck together by a polyion which acts as an electrostatic glue, is one of the many facets of the complex phenomenology underlying the interactions of charged particles with oppositely charged objects.

1. Introduction Mixtures of charged mesoscopic particles such as liposomes and oppositely charged polyions self-assemble into various structural arrangements, ranging from micelles, vesicle clusters, and lamellar or columnar and cubic mesophases, depending on different environmental chemicophysical parameters and also the different mixing protocols.1 Among these structures, the ones based on the complexation of different intact liposomes stuck together by oppositely charged polyions give rise to a cluster phase representing an important class of colloidal systems that have found relevant use as transport vehicles for drugs, enzymes, and DNA. Adsorption of polyions onto oppositely charged surfaces or onto charged colloidal particles has been studied in the last two decades by experiments and by analytical and simulation methods, thanks to the rich phenomenology2 and to the implications in biomedical applications3,4 ranging from nonviral gene therapy to, more generally, biotechnology. Even if the existence of a cluster phase in polyion-liposome colloidal suspensions has been extensively proved, also, through a series of works from our laboratory (an experimental overview * To whom correspondence should be addressed. E-mail: cesare.cametti@ roma1.infn.it. Fax: +39 06 4463158. † Universita’ di Roma “La Sapienza” and Research Center SOFT-INFMCNR. ‡ Universita’ di Perugia and Research Center SOFT-INFM-CNR. (1) Safinya, C. R.; Ewert, K.; Ahmad, A.; Evans, H.; Raviv, U.; Needleman, D.; Lin, A.; Slack, N.; George, C.; Samuel, C. Philos. Trans. R. Soc. London, Ser. A 2006, 364, 2573–2593. (2) Radler, J. O.; Koltover, I.; Salditt, T.; Safinya, C. R. Science 1997, 275, 810–814. (3) Ghirlando, R.; Wachtel, E.; Arad, J.; Minsky, A. Biochemistry 1992, 31, 7110–7119. (4) Dautzenberg, H.; Jagger, W.; Kotz, J.; Philip, B.; Seidel, C. Polyelectrolytes: Formation, characterization and applications; Hansen: Munich, 1994.

is given in ref 5), some contradictory postulations exist in the literature, especially in the case of lipidic liposome-DNA complexes (lipoplexes). In this latter case, a commonly accepted picture considers the resulting structural arrangements as multilamellar or hexagonal phases, with a partial or complete destruction of the liposome bilayer architecture to form a variety of assemblies that host the DNA chains sandwiched between cationic bilayers.6 Liposome restructuring is known to occur during lipoplex formation, consisting of liposome fusion7,8 with a consequent release of the aqueous content.9,10 Some further experimental evidence for liposome restructuring is provided by electron microscopy which shows elongated rodlike structures11 and aggregates of globular particles.12 Besides these complex restructured arrangements, we have shown13–18 that, in an appropriate concentration regime, a different class of aggregates exists, where the individual liposome particles maintain their integrity and liposome clusters are formed (5) Bordi, F.; Cametti, C.; Sennato, S. Colloids Surf., A 2007, 306, 102–110. (6) Safinya, C. R. Curr. Opin. Struct. Biol. 2001, 11, 440–448. (7) Gershon, H.; Ghirlando, R.; Guttman, S.; Minsky, A. Biochemistry 1993, 32, 7143–7151. (8) Mok, K.; Cullis, P. Biophys. J. 1997, 73, 2534–2545. (9) Kikuchi, I.; Carmona-Ribeiro, A. J. Phys. Chem. B 2000, 104, 2829–2845. (10) Kennedy, M.; Pozharski, E.; Rakhmanova, V.; Mac Donald, R. Biophys. J. 2000, 78, 1620–1633. (11) Sternberg, B.; Sorgi, F. L.; Huang, L. FEBS Lett. 1994, 356, 361–366. (12) Eastman, S.; Siegel, C.; Tousignant, J.; Smith, A.; Cheng, S. Biochim. Biophys. Acta 1997, 1325, 41–62. (13) Sennato, S.; Bordi, F.; Cametti, C. Europhys. Lett. 2004, 68, 296–302. (14) Sennato, S.; Bordi, F.; Cametti, C. J. Chem. Phys. 2004, 121, 4936–4940. (15) Sennato, S.; Bordi, F.; Cametti, C.; Diociaiuti, M.; Malaspina, M. Biochim. Biophys. Acta 2005, 1714, 11–24. (16) Bordi, F.; Cametti, C.; Diociaiuti, M.; Sennato, S. Phys. ReV. E 2005, 71, 050401/4(RC) (17) Bordi, F.; Cametti, C.; Marianecci, C.; Sennato, S. J. Phys.: Condens. Matter 2005, 17, 3423–3432. (18) Bordi, F.; Cametti, C.; Sennato, S.; Diociaiuti, M. Biophys. J. 2006, 91, 1513–1520.

10.1021/la8005458 CCC: $40.75  2008 American Chemical Society Published on Web 05/17/2008

Polyion-Induced Aggregation of Lipidic Spheres

by sticking together with different liposomes, with the process being favored by the presence of oppositely charged linear polyions, independently of their chemical structure. The many facets of polyion and oppositely charged particle complex formation make the overall scenario of this phenomenology, on one hand, extremely varied and, on the other, rather intriguing for important applications in nanoscience and in various biological regulation processes. The formation of these polyion-induced (intact) vesicle aggregations is largely independent of the choice of the particular lipid or polyion to be employed, and we have strong experimental evidence about the occurrence of the above stated phenomenology both in the case of cationic dioleoyltrimethylammonium propane [DOTAP] liposomes interacting with anionic polyions (e.g., polyacrylate sodium salt)19 and in the case of anionic vesicles (hybrid niosomes) interacting with cationic polyions (e.g., R-polylysine or polyethylvinyl pyridinium bromide [PEVP]).20 In order to provide further insight into the intact liposome cluster formation, in this work, we employ spherical solid particles coated with lipids organized in bilayers and we compare their aggregation process, induced by different polyions, with the one already observed13–18 in the presence of soft particles, such as lipidic vesicles (liposomes). Whereas in this latter case a vesicle rupture (and the following structural rearrangement) is, in principle, possible, when solid particles are employed, a large class of structures are, on the contrary, hindered, in particular, those involved in the formation of lamellar LcR phases2 or of inverted hexagonal HcII liquid phases.21 It is worth noting that, with systems we are dealing with, both kinds of particles (soft vesicles and solid particles) present a surface organization built up by the same lipid (in this case a double-chain cationic lipid DOTAP), so that aggregation should share the same interface and should occur under the same electrostatic interactions. We employed hybrid assemblies composed of a negatively charged submicrometer polystyrene particles (327 nm in nominal diameter) covered by a bilayer of cationic lipids (in this case, DOTAP, to make the comparison with previous results consistent). The adsorption of cationic lipids onto anionic polystyrene particles was first suggested by CarmonaRibeiro et al.22–24 and later confirmed by Troutier et al.25 on the basis of ζ-potential measurements and fluorescent and transmission electron microscopy. A recent review of lipid membranes supported by colloidal particles can be found in ref 26. Following the nomenclature introduced by Carmona-Ribeiro, we will call these lipid bilayers supported by colloidal polystyrene particles “lipoparticles”. We have induced the cluster aggregation of the lipoparticles by means of two different linear, oppositely charged polyions, that is, polyacrylate sodium salt [PAA], 60 kD in molecular weight, and calf-thymus DNA, with a size distribution ranged from 500 to 700 nucleotide pairs. In both cases, we obtain the same basic phenomenology, consisting of the occurrence of the reentrant condensation and the charge inVersion effects. The presence of both of them is a clear print for a cluster phase (19) Bordi, F.; Cametti, C.; Diociaiuti, M.; Gaudino, D.; Gili, T.; Sennato, S. Langmuir 2004, 20, 5214–5222. (20) Sennato, S.; Bordi, F.; Cametti, C.; Marianecci, C.; Carafa, M.; Cametti, M. J. Phys. Chem. B 2008, 112, 3720–3727. (21) Koltover, I.; Salditt, T.; Ra¨dler, J. O.; Safinya, C. R. Science 1998, 281, 78–81. (22) Carmona-Ribeiro, A.; Midmore, B. Langmuir 1992, 8, 801–806. (23) Carmona-Ribeiro, A.; de MoralesLessa, M. Colloids Surf., A 1999, 153, 355–361. (24) Lincopan, N.; Espindola, N.; Carmona-Ribeiro, A. Int. J. Pharm. 2007, 340, 216–222. (25) Troutier, A.; Delair, T.; Ladaviere, C. Langmuir 2005, 21, 1305–1313. (26) Troutier, A.; Ladaviere, C. AdV. Colloid Interface Sci. 2007, 133, 1–21.

Langmuir, Vol. 24, No. 12, 2008 6045

in low-density colloidal suspensions built up by stable colloidal particle clusters. Our results support further evidence, among others, for the existence of a liposome cluster phase, where liposomal particles maintain their integrity and, in particular, the capability of preserving the aqueous inner core from the external medium. This peculiarity makes these systems eligible carriers in drug delivery techniques.

2. Experimental Section 2.1. Materials. The cationic lipid dioleoyltrimethylammonium propane [DOTAP] was purchased from Avanti Polar Lipids Inc. (Alabaster, AL) and used without further purification. Anionic polystyrene particles were purchased from Sigma Chem. Co. and used as supplied. Particles were functionalized with sulfate groups, and they are surfactant free. The surface charge density, as given by the manufacturer, was of the order of 0.3 µC/m2. The mean particle diameter determined by the supplier from transmission electron microscopy was 2R ) (327 ( 10) nm. The mean hydrodynamic diameter determined by dynamic light scattering measurements was 2R ) (316 ( 10) nm, with a polydispersity defined by Poly ) µ2/〈Γ〉2, where µ2 is the second cumulant of the scattered light correlation function fitted by the cumulant analysis and 〈Γ〉 is the average decay rate (of about 0.036). Poly(acrylate) sodium salt was purchased from Sigma Chem. Co. as an aqueous solution 25% w/w. The molecular weight is MW ) 60 kD. Double-stranded calf-thymus DNA, purchased from Sigma Chem. Co. and used without further purification, was dissolved in water. The purity of the DNA employed was analyzed by means of ultraviolet spectrophotometry (A260/A280 )1.9, with A260 and A280 being the absorbance at 260 and 280 nm, respectively). 2.2. Vesicle Preparation. Unilamellar liposomes were prepared by using a standard lipid film hydration method. The DOTAP lipid was dissolved in chloroform-methanol (1:1 v/v) at a concentration of 10 mg/mL. After solvent evaporation, dried lipid films were hydrated with deionized water (electrical conductivity of less than 10-6 mho/cm, at room temperature). In order to form unilamellar vesicles, the lipid solution was sonicated at a temperature of 25 °C for 1 h at the pulsed-power mode until the solution appeared optically transparent in white light. The solution was then filtered through a Millipore polycarbonate filter of 0.45 µm in size. Liposome size and size distribution obtained from dynamic light scattering measurements give an average diameter of about 80 nm with a moderate polydispersity of about 0.2, as expected for a rather homogeneous particle suspension. 2.3. Dynamic Light Scattering Measurements. Dynamic light scattering [DLS] measurements were carried out using a Brookhaven instrument (Brookhaven, NY) equipped with a 10 mW HeNe laser at a 632.8 nm wavelength at a temperature of (25.0 ( 0.2) °C. Correlation data were collected at 90° relative to the incident beam, and delay times from 0.8 µs to 10 s were explored. Correlation data were fitted using the non-negative least-squares [NNLS]27 or CONTIN28,29 algorithms, supplied with the instrument software. The average hydrodynamic radius of the diffusing objects was calculated from the diffusion coefficient D and the Stokes-Einstein relationship, R ) (KBT)/(6πηD), where KBT is the thermal energy and η is the solvent viscosity. 2.4. ζ-Potential Measurements. The electrophoretic mobility of simple liposomes, polystyrene particles, and polyion-induced coated polystyrene particle aggregates (lipoparticles) was measured by means of laser microelectrophoresis in a thermostatted cell (T ) 25 °C) using a Malvern Zetasizer Mod. instrument (Malvern, U.K.). Since the vesicle and polystyrene particle diameter is much larger (27) Lawson, C. L.; Morrison, I. D. SolVing least squares problems. A FORTRAN program and subroutines called NNLS; Prentice-Hall: Englewood Cliffs, NJ, 1974. (28) Provencher, S. Comput. Phys. Commun. 1982, 27, 213–227. (29) Provencher, S. Comput. Phys. Commun. 1982, 27, 229–242.

6046 Langmuir, Vol. 24, No. 12, 2008

Zuzzi et al.

Figure 1. ζ-Potential of DOTAP-coated polystyrene particles as a function of the ratio Av/Ap at the temperature of 25.0 °C. The fractional volume of the polystyrene particles is maintained constant to the value Φ ) 0.001, and the DOTAP concentration is varied from 0.005 to 0.80 mg/ mL. Two independent sets of measurements are reported: (O) Av/Ap between 0.3 and 10 and (0) Av/Ap between 2 and 45. The values of the ζ-potential of the DOTAP liposomes and the bare polystyrene particles are also shown.

than the Debye screening length, the ζ-potential was calculated from the measured electrophoretic mobility u by means of the Smolukovsky equation

ζ)

4πηu 

(1)

where  and η are the permittivity and the viscosity of the aqueous phase, respectively. 2.5. Lipoparticle Preparation and Characterization. Lipoparticles were prepared by adding a polystyrene particle suspension (316 nm in diameter, from dynamic light scattering) to a preformed DOTAP vesicle (80 nm in diameter) suspension. The relative proportion of the polystyrene particles and liposomal vesicles in the whole sample is characterized by a ratio Av/Ap (the notation was first introduced by Carmona-Ribeiro et al.22,23) where Ap is the total surface area per unit volume provided by the polystyrene particles (the area is calculated from the fractional volume Φ and the mean radius Rp by TEM, according to Ap ) 3Φ/Rp) and Av is the total surface area per unit volume of vesicles (calculated from the surface area a0 per DOTAP molecule at the air-water interface according to Av ) NCDOTAPa0/MWDOTAP, where CDOTAP is the DOTAP concentration, MWDOTAP is its molecular weight, and N is Avogadro’s number). The progressive adsorption of DOTAP bilayer fragments onto polystyrene particles changes the sign of the particle surface charge, shifting its value from negative (the initial value of the bare polystyrene particles) to positive, and favors an interparticle aggregation close to the overall charge neutralization. Further addition of DOTAP liposomes stabilizes the polystyrene particles at a size and ζ-potential consistent with polystyrene particles covered by oppositely charged DOTAP bilayer fragments. In order to follow the whole adsorption process during the addition of DOTAP liposomes to a polystyrene particle suspension, we have varied the ratio Av/Ap from 0.3 to 45. We employed a polystyrene suspension with a fractional volume Φ ) 0.001 and a DOTAP liposome suspension at a lipid concentration varying from CDOTAP ) 5 × 10 -3 mg/mL to CDOTAP ) 0.8 mg/mL. These conditions correspond to polystyrene particle excess at the beginning and, conversely, DOTAP liposome excess at the end of the process. The results concerning the charge of the lipoparticles and their hydrodynamic size are shown in Figures 1 and 2. As can be seen, the ζ-potential progressively increases from the negative value of the bare polystyrene particles (ζ ) -55.8 mV) to positive values as the liposomal vesicle concentration is increased. From values of Av/Ap larger than Av/Ap ) 18, the ζ-potential remains approximately constant to a positive value of about ζ ) 23 mV (Figure 1). The same trend was observed by Troutier et al.25 in the case of the cationic lipid dipalmitoyl trimethylammonium propane

Figure 2. Size distribution of the resulting DOTAP-coated polystyrene particles (lipoparticles) for different values of the ratio Av/Ap, in the range Av/Ap g 18. The upper panel shows the initial size distribution of both the DOTAP liposomes and polystyrene particles before mixing. At Av/Ap ) 18, the DOTAP liposome component has practically disappeared and only a larger size component (lipoparticles) is present.

[DPTAP], with a change of the ζ-potential from ζ ) -58 mV (bare polystyrene particle) to ζ ) 22 mV (lipoparticles). The effective surface charge of the DOTAP-coated polystyrene particles can be derived from the relationship between the surface charge and the ζ-potential (approximated to the surface potential) for spherical particles in a salt-free medium, recently proposed by Ohshima.30 This relationship reads

ζ)

(

( )(

KBT 1 ze ln ze 6Φ ln(Φ) KBT

2

Q 4π0R

)) 2

(2)

where KBT is the thermal energy, ze is the ionic charge, 0 is the dielectric constant of free space, and Q is the electric charge smeared over the particle of radius R dispersed in an aqueous medium of permittivity . For particles 316 nm in diameter, at a concentration of Φ ) 0.8 × 10 -3 in an aqueous salt-free phase, at a potential of ζ ) 22 mV, it corresponds from eq 2 an effective surface charge of Q = 65 elementary charges and consequently a surface charge density of 3.2 × 10 -3 µC/cm2 . Before going on, a brief comment on the surface charge density of lipoparticles is compelled. This charge density is very small when compared to the stoichiometric values that might be estimated on the basis of a full ionization of the polystyrene particle charged groups, thus suggesting that, in this case, as in other similar situations,31,32 a very strong charge renormalization occurs. For example, as pointed out by Haro-Pe´rez (30) Ohshima, H. J. Colloid Interface Sci. 2002, 247, 18–23. (31) Crocker, J. C.; Grier, D. G. Phys. ReV. Lett. 1994, 73, 352–355. (32) Aubouy, M.; Trizac, E.; Bocquet, L. J. Phys. A: Math. Gen. 2003, 36, 5835–5840. (33) Haro-Pe´rez, C.; Quesada-Pe´rez, M.; Callejas-Ferna´ndez, J.; Casals, E.; ´ lvarez, R. J. Chem. Phys. 2003, 118, 5167–5173. Estelrich, J.; Hidalgo-A

Polyion-Induced Aggregation of Lipidic Spheres

Figure 3. ζ-Potential of PAA-induced lipoparticle aggregates as a function of the PAA polyion concentration. The charge inVersion effect changes the overall charge of the aggregates from positive (lipoparticles in the absence of PAA) to negative, after adsorption in excess of PAA chains. The inset shows the ratio R/R0 of the radius of the aggregates normalized to the radius of the bare lipoparticle as a function of the PAA polyion concentration. This behavior is typical of the reentrant condensation effect.

et al.,33 the structure factor of latex particles 80 nm in size determined from static light scattering experiments has been analyzed using as the adjustable parameter an effective charge of Q ) 330 e, considerably smaller than the surface charge estimated by acid-base titration. Coming back to the lipoparticle coating, we have followed, by means of dynamic light scattering methods, the whole process resulting in the progressive lipidic coverage of the polystyrene particles, by varying the ratio Av/Ap. These measurements indicate that, in the condition of the experiment, at Av/Ap = 18, the process is practically ended. From this condition on (from Av/Ap ) 18 to Av/Ap ) 45), further addition of DOTAP liposomes to the polystyrene suspension does not produce further lipid adsorption with the consequence that, in this range of concentration, isolated liposomes coexist with DOTAPcoated polystyrene spheres. This condition is clearly shown in Figure 2, where the size distributions of the different diffusing objects are reported for typical values of the ratio Av/Ap larger than Av/Ap ) 18. As can be seen, the peak due to DOTAP-coated polystyrene particles is accompanied by a complementary peak whose relevance increases together with the increase of the ratio Av/Ap, due to the DOTAP component. The DOTAP-coated polystyrene particles are stable against thermal or mechanical stress. Once DOTAP adsorption occurred, the DOTAPcoated polystyrene particle suspension at Av/Ap ) 18 was thermostatted at 50 °C for 1 h or else sonicated at pulse power mode for 15 min and then thermostatted again at 50 °C for 1 h. In both cases, no significant changes in ζ-potential and size were observed. Contrary to what happened in experiments with different protocols22,23,25 where suspensions were centrifuged to separate particles from the remaining part of vesicles, we have ascertained by means of DLS measurements that at Av/Ap ) 18 complete adsorption was reached and that, in these conditions, practically no free DOTAP liposomes were present. From here onward, we will employ particles with Av/Ap ) 18.

3. Results and Discussion Polyion-induced coated polystyrene particle complex formation has been investigated by measuring the average hydrodynamic radius and the electrophoretic mobility of the diffusing complexes in the solution by means of dynamic light scattering measurements and laser Doppler electrophoresis measurements. The combined use of these two techniques allows us to study both of the two typical phenomena occurring in these systems, that is, the reentrant condensation and the charge inVersion effect. We are going to present and discuss separately the results in the aggregates induced by the two polyions employed, polyacrylate sodium salt [PAA] and DNA. 3.1. Aggregation Induced by PAA Polyions. In this case, the DOTAP bilayer covered polystyrene particle volume fraction

Langmuir, Vol. 24, No. 12, 2008 6047

was maintained fixed to the value of Φ ) 3.3 × 10 -3, corresponding to a concentration of 7.5 × 10 11 particles/mL, and the concentration of PAA polyion was varied from CPAA ) 0.02 mg/ml (2 × 10 -4 monoMol/L) to CPAA ) 2.0 mg/ml (2 × 10 -2 monoMol/L). The two typical effects, reentrant condensation and charge inVersion, are shown in Figure 3, where the ζ-potential and the average size aggregates are shown as a function of the PAA polyion concentration. In this case, the concentration of DOTAP was constant to the value CDOTAP ) 1 mg/mL, corresponding to the one employed for the sample Av/Ap ) 18. If at this Av/Ap ratio the whole DOTAP is involved in the lipoparticle formation, the molar charge ratio ξ ) N-/N+ defined as

ξ ) N- ⁄ N+ )

CPAA ⁄ MWPAA CDOTAP ⁄ MWDOTAP

(3)

varied from ξ ) 0.5 to 25. Here, CPAA and CDOTAP are the concentrations of PAA polyions and DOTAP lipids and MWPAA and MWDOTAP are their monomer molecular weights, respectively. As can be seen, with the increase of the PAA content (i.e., with the increase of ξ), complexation begins and the size of the complexes gradually increases until a maximum is reached. Further increase of the PAA content determines the formation of decreasing-size complexes until the size of the original lipoparticles is approximately reached again (reentrant condensation). Besides this average size evolution, aggregates undergo the charge inVersion effect, documented by the ζ-potential values whose sign changes at the isoelectric condition, differentiating positive charge aggregates before the isoelectric condition from negative charge aggregates above the isoelectric condition. It must be noted that, in this case, contrary to what happens in the case of liposome vesicles,13,14,19 the maximum size reached by the stable, equilibrium aggregates is relatively lower. This can be justified by the fact that the surface charge density of lipoparticles is lower than the one of liposomal vesicles (we measured a ζ-potential of 23 mV against a value of 48 mV in the case of liposomes) so that aggregation is favored and, consequently, larger aggregates easily flocculate and precipitate, leaving in the suspension only aggregates of smaller size, which are exactly the ones we effectively observe by dynamic light scattering measurements. As a matter of fact, close to the point of charge inversion, we observed, by visual inspection, a restricted amount of the sample forming large flocculating aggregates. Moreover, the ionic strength of these lipoparticle suspensions is relatively higher than the one of the simple liposome suspensions employed in our previous investigations,15–17,19,34 favoring aggregate flocculation and precipitation. As far as the overall ionic strength is concerned, its increase is due to the ionic contribution derived from the polystyrene particle suspension, whose ionic electrical conductivity, at the concentration Φ ) 0.0008, is σ ) 0.00464 mho/m. Within the effective medium approximation theory,35 the conductivity σ of a collection of spheroidal particles of conductivity σp, uniformly dispersed in a continuous medium of conductivity σm, is given by

σ ) σm

2σm + σp - 2Φ(σm - σp) 2σm + σp + Φ(σm - σp)

(4)

Since in our case σp,σm, σ ) σm(2(1 - Φ))/(2 + Φ) and, for Φ = 0.8 × 10 -3, we expect that the conductivity σ of the suspension does not differ appreciably from the conductivity of (34) Sennato, S.; Bordi, F.; Cametti, C.; Di Biasio, A.; Diociaiuti, M. Colloids Surf., A 2005, 270, 138–147. (35) Hasted, J. B. Aqueous Dielectrics; Chapman and Hall: London, U.K., 1973.

6048 Langmuir, Vol. 24, No. 12, 2008

Zuzzi et al.

The main results are summarized in Figure 4, where we show, in analogy with the data reported in Figure 3, the ζ-potential and the average size of the DNA-induced lipoparticle aggregates as a function of the DNA polyion concentration. Both the reentrant condensation and the charge inVersion are clearly recognizable. In analogy with the previous results, the molar charge ratio ξ, defined as

ξ ) N- ⁄ N+ )

Figure 4. (A) Ratio R/R0 of the radius of the DNA-induced aggregates normalized to the radius of the bare lipoparticles as a function of the DNA polyion concentration. This behavior is typical of the reentrant condensation effect. (B) ζ-Potential of DNA-induced lipoparticle aggregates as a function of the DNA polyion concentration. The charge inVersion effect changes the overall charge of the aggregates from positive (lipoparticles in the absence of DNA) to negative, after adsorption in excess of DNA chains.

the external medium, σm. Assuming Na+ ions derived from the ionization of the surface charged groups of polystyrene particles, with an equivalent conductance of λNa+ ) 50.1 cm2 mho/equiv, the measured electrical conductivity results in a Na+ ion concentration of 0.84 × 10-3 mol/L. This value justifies the appearance of precipitated aggregates close to the charge neutralization condition. Elsewhere, we have investigated in detail the influence of added salt (with the consequence of an increase of the ionic strength) on the aggregation behavior of these systems and we refer to the related papers for the observed phenomenology.19,36 Finally, it is worth noting that, in these experimental conditions, the formation of DOTAP-coated polystyrene clusters (lipoparticles) occurs in obedience to the additivity of the ionic contributions. For clusters at ξ ) N-/N+ = 1, the electrical conductivities of PAA polymer (at the concentration CPAA ) 0.036 mg/mL), of DOTAP liposomes (at the concentration CDOTAP ) 0.25 mg/mL), and of the polystyrene particle suspension (at a concentration Φ ) 0.0008) are 0.00240, 0.00221, and 0.00464 mho/m, respectively. When mixed together, the overall electrical conductivity of the resulting aggregates is 0.00918 mho/m, very close to the value derived from an additivity rule of the single components (i.e., 0.00925 mho/m). The above stated phenomenology is totally similar to the one previously observed in the case of soft particles, that is, simple DOTAP lipidic vesicles (liposomes), instead of solid particles, as in the present case of lipoparticles. For a detailed comparison, see refs .13–18 We will complete our parallel considering the influence of a different aggregation agent, such as DNA, that, in the case of soft particles (liposomes), could cause their rupture and, as a consequence, their different rearrangement, as claimed, here and there, in the literature.1,6 3.2. Aggregation Induced by DNA Polyions. In the case of DNA polyions, the aggregation phenomenology is essentially the same as the one observed in the case of PAA polyions in the presence of DOTAP-coated polystyrene particle suspensions and, moreover, similar to the one observed in the case of liposome suspensions, that is, in the presence of soft particles, where a rupture of the lipidic bilayer might be, in principle, possible. (36) Bordi, F.; Cametti, C.; Sennato, S. Chem. Phys. Lett. 2005, 409, 134–138.

CDNA ⁄ MWDNA CDOTAP ⁄ MWDOTAP

(5)

varied from ξ ) 0.1 to 5. In this case, CDNA is the concentration of DNA and MWDNA is the nucleotide molecular weight. As previously stated, the concentration of DOTAP is CDOTAP ) 1 mg/mL, corresponding to lipoparticles at a ratio Av/Ap ) 18. It could be noted that charge inversion and the maximum in the aggregate size do not occur at a molar charge ratio ξ ) N-/N+ close to the stoichiometric point of charge inversion. This is due to the fact that the effective charge density of the DNA chain, as well as that of the PAA chain, differs from its stoichiometric charge (full ionization of each base pair) because of the Manning counterion condensation effect.37,38 These details, which are not really relevant in the present context, do not modify the overall scenario of the polyion-induced charge particle cluster formation.

4. Conclusions Our results offer further support to the existence of relatively large equilibrium clusters of mesoscopic colloidal particles in low-density colloidal suspensions, stuck together by a oppositely charged polyion that acts as an electrostatic glue. Whereas this picture has found a favorable acceptance in the case of synthetic polyelectrolytes, and by now various forms of experimental evidence are available, even from our laboratory, when we have to deal with liposomes and DNA (lipoplexes), the possible existence of a cluster phase that maintains the integrity of the liposomal vesicles inside the cluster finds more obstacles and its acceptance is today still a little controversial. In the case of liposome-DNA complexes (lipoplexes), different structures have been evidenced. This implies the liposome rupture and the consequent rearrangement of the lipidic component, favored by the DNA. However, in different experimental conditions, with different environmental parameters and with different mixing protocols and in different concentration ranges, liposomal clusters stuck together by DNA chains have been observed at the same extent as well. The results presented here would put forward further evidence for the existence of an equilibrium cluster phase as a general class of low-density colloids which represent one of the many facets of the complex formation driven by electrostatic forces. We have observed the same basic phenomenology (charge inversion and reentrant condensation) that characterizes this class of colloids both in soft particles (liposomes in the presence of synthetic polyions and DNA) and, presented in this work, in solid particles (lipoparticles, in the presence of synthetic polyions, such as PAA, and natural polyions, such as DNA). In this latter case, the structural reorganization is inhibited by the presence of solid particles, but, nonetheless, the same phenomenology occurs. Within the experimental conditions employed, in the presence of low-density colloidal particle suspensions, the formation of particle clusters is essentially governed by the surface (37) Manning, G. S. J. Chem. Phys. 1969, 51, 924–933. (38) Manning, G. S. Q. ReV. Biophys. 1978, 11, 179–246.

Polyion-Induced Aggregation of Lipidic Spheres

electrical properties in connection with the oppositely charged linear polyions which, in this context, act as an electrostatic glue. This class of colloids is characterized by the balance between short-range attractive charge patch attraction and long-range repulsive interaction giving rise to an equilibrium cluster phase, where particles form reversible, relatively large aggregates. In these colloidal systems, the peculiar mechanism that accounts for the short-range attraction is due to the correlated adsorption of polyions on the oppositely charged surface, forming a two-

Langmuir, Vol. 24, No. 12, 2008 6049

dimensional strongly correlated short-range-order structure. Once this particle surface is formed, the particle aggregation proceeds independently of the bulk structure of the particle and the same phenomenology is observed for both solid and soft particles. These aspects, which are particularly characteristic of softmatter physics, have been addressed in a series of recent papers to which we refer for further details.13–18 LA8005458