**3. Results**

#### *3.1. Characterization of Thermosensitive Liposomes (TSLs)*

TSLs composed of DPPG were prepared as is described in the Methods Section (Section 2.2.1). Their size and stability were characterized by DLS and Zeta Potential, as shown in Table 1. The lipid was selected because its high ability to associate with cationic polielectrolytes as well as for its transition temperature in the mild hypertermia range. The experiments were done at 24 ◦C, where DPPG is in the gel phase. The vesicles exhibited hydrodynamic diameters around 130 nm and a high negative Zeta Potential, which should guarantee adequate suspension stability, since it was previously reported that Zeta Potential values higher than |20| mV are sufficient to prevent vesicle coalescence [52]. The polydispersity index value was 0.11, evidencing a narrow distribution of particle size. The morphology of the freshly prepared TSLs

was observed under transmission electron microscopy. Results showed well-dispersed spherical-shaped vesicles, with a particle size compatible with the one estimated by DLS (Figure 1).


**Table 1.** Hydrodynamic diameter (d) and Zeta Potential (ZP) of TLSs and blue, green and red fluorescent nanoparticles.

**Figure 1.** Transmission electron microscopy (TEM) images (**<sup>a</sup>**,**b**) of thermosensitive liposomes (TSLs) at different magnifications.

The thermal behavior of the TSLs was explored using light scattering measurements. We analyzed the light scattered by the TSLs suspension as a function of temperature, taking into account that gel phase bilayers scatter more light than fluid membranes, a feature which has been attributed to the higher refractive index of gel membranes, as compared with fluid ones [53]. This experiment was directly made in the spectrofluorometer, by selecting the same wavelength for both excitation and emission monochromators with the smallest slit. The scattered light (430 nm) was collected at an angle of 90◦ of the incident light. As is shown in Figure 2a, a sharp drop of the scattered light was observed around 41–42 ◦C, which coincides with the transition temperature reported for DPPG. These results were confirmed by the fluorescence anisotropy measurements using the fluorescent probe DPH incorporated into the TSLs bilayer. This is a commonly used tool for thermotropic characterization of liposomes [54]. The plot of the steady-state fluorescence anisotropy, <*r*>, of DPH versus temperature is shown in the Supplementary Material (Figure S1), and was a perfect sinusoidal, displaying a sharp transition of anisotropy values around 42 ◦C.

**Figure 2.** Effect of temperature on (**a**) the scattered light by TSLs and on (**b**) area of the emission spectrum of carboxyfluorescein (CF) encapsulated in TSLs in sodium phosphate buffer. (**c**) CF released in % as function of time (0–900 s) at different temperatures: 30 ◦C (black), 35 ◦C (red), 37 ◦C (green), 40 ◦C (blue) and 45 ◦C (magenta).3.2. Encapsulation and Release Assays.

#### *3.2. Encapsulation and Release Assays*

The dye carboxyfluorescein (CF) was entrapped in the aqueous cavity of TSLs, as was described in the Methods Section (Section 2.2.3). This marker was chosen for the release assays because of its interesting photophysical properties. When CF is highly concentrated, most of its fluorescence is quenched due to dimerization to a nonfluorescent compound as well as the Förster Resonance Energy Transfer (FRET) process between monomers and dimers [55]. Therefore, it is possible to monitor its release from the TSLs measuring the increase in fluorescence as a function of temperature (Figure 2b) or time (Figure 2c). Figure 2b shows an abrupt increase in the fluorescence of CF above 40 ◦C, which evidences the release of the dye from the liposome close to the transition temperature. Samples were also exposed to temperatures ranging from 30 to 45 ◦C, and the time release profile was recorded at each temperature (Figure 2c). Every sample was preheated at 30 ◦C before being placed in the thermostatized fluorimeter holder. Percent release was calculated from the change in fluorescence intensity, as detailed in the Methods Section (Section 2.2.3), after addition of Triton X-100 10%. Results show that below 37 ◦C, the dye remains entrapped into the liposome, at least within the experimental time period. At 40 ◦C, it starts to be slowly removed from the TSLs, probably due to the coexistence of gel and fluid domains at temperatures just below T m, as has been reported in previous works for DPPG [56]. Finally, a rapid release of CF takes place at 45 ◦C, especially in the first three minutes, evidencing the ability of the DPPG-TSLs to carry and release hydrophilic compounds triggered by hyperthermia. The fact that only a 40% of CF is released at this temperature after 15 min of incubation could be attributed to dye interaction with the lipid membrane [56].

#### *3.3. Preparation and Characterization of Fluorescent Nanoparticles*

Once the physical properties and thermal behavior of the TSLs were characterized, the next step was to obtain the fluorescent nanoparticles. They were prepared by incorporation of the three polyelectrolytes: blue (HTMA-PFP), green (HTMA-PFBT) and red (HTMA-PFNT) in the bilayer of TSLs, at 25 ◦C (gel phase), as is shown in Scheme 2. This temperature was selected to prevent the release of the encapsulated drug from the TSLs before heating. In previous works, we demonstrated the ability of these CPEs to interact and insert into fluid phase lipid bilayers, by determining their a ffinity and membrane location [42–44]. But, taking into account the higher lipid packing of the gel phase, we cannot assume that the amount of polyelectrolyte bound to the fluid-phase membrane was the same as that bound to the gel-phase membrane. Therefore, the first experiments were focused to estimate the affinity of the CPEs for the TSLs, by determining their partition coe fficient, Kp, defined in Equation (2). With this end, three series of samples containing increasing concentrations of DPPG-TSLs with final lipid concentrations ranging from 0 to 1 mM were prepared in bu ffer, and a constant concentration (3 μM) of HTMA-PFP, HTMA-PFBT and HTMA-PFNT was added to each series of samples respectively, which were incubated for 30 min at room temperature.

The emission spectra for the three polyelectrolytes, recorded at the di fferent lipid concentrations, are shown in Figure 3a–c. A low fluorescence emission signal was detected for the CPEs in bu ffer, as a consequence of the formation of metastable aggregates [42–44]. The increase of the lipid concentration induced an enhancement in fluorescence intensity and a blue-shift of the emission spectra (insets in Figure 3), which suggests the breaking of aggregates as a consequence of the interaction of the three polyelectrolytes with TSLs-bilayer. Plotting the area of each spectrum versus lipid concentration (Figure S2) and using Equation (3), it was possible to determine the Kp values for the three CPEs. These values are summarized in Table 2 and indicate that, although the bilayer is in the gel-phase, the three polymers have high a ffinity for the TSLs. However, the Kp values are one order of magnitude lower than those obtained for anionic membranes in the fluid-phase [42–44]. This result suggests that the mode and nature of the interaction between CPEs and lipid membranes is di fferent in the two phases. Probably, in the gel-phase, the interaction is mainly electrostatic between the quaternary amine groups of CPEs and the negative charge of the lipid head groups. In contrast, in the fluid-phase, the decrease of the lipid packing allows the insertion of the polymer chains and the hydrophobic forces

contribute to a better solubilization. The fact that the fluorescence intensity of the CPEs incorporated in the gel-phase was lower than that registered in the fluid-phase at the same conditions (data not shown), supports this hypothesis.

**Figure 3.** Fluorescence emission spectra of (**a**) HTMA-PFP (3 μM), (**b**) HTMA-PFBT (3 μM) and (**c**) HTMA-PFNT (3 μM) in buffer with increasing concentrations of TSLs. Insets: Normalized fluorescence emission spectra of CPEs in sodium phosphate buffer (black) and incorporated in TSLs (color).

**Table 2.** Partition coefficient, Kp, values, and Stern–Volmer constants, KSV, for HTMA-PFP, HTMA-PFBT and HTMA-PFNT in TSLs by using 9,10-anthraquinone-2,6-disulfonic acid (AQS) as a quencher.


The Kp values were used to optimize the concentration of each component for the fabrication of the fluorescent nanoparticles. The lipid concentration was fixed to 0.5 mM to limit the turbidity of the samples, which becomes an obstacle for fluorescence measurements. The concentration of the CPEs was 3 μM in order to obtain a good fluorescent signal and to ensure that more than 90% of the polyelectrolyte was bound to the TSLs. The stability of the nanoparticles was assessed by monitoring the fluorescence intensity of the sample in the maximum of the emission spectrum as a function of time, after addition of the CPEs. The signal stabilized in the first seconds for HTMA-PFP and HTMA-PFNT and in ~5 min for HTMA-PFBT and remained stable during the experimental time (Figure S3). In addition, the possibility of simultaneously exciting the fluorescence emission of blue, green and red nanoparticles suspended in the same sample was also studied. The mixture was excited at 335 nm, where the three polyelectrolytes absorb (see Scheme 1). The recorded spectrum, shown in Figure S4, clearly displays the three bands corresponding to the characteristic spectra of HTMA-PFP, HTMA-PFBT and HTMA-PFNT.

The localization of the CPEs in the lipid bilayer of TSLs was investigated by quenching experiments using the anionic acceptor AQS as a fluorescence quencher. This molecule has been observed to be an excellent quencher for cationic conjugated polyelectrolytes, and it is soluble in water but not in lipid bilayer, so the fluorescence only will be deactivated if the CPEs are in the buffer or in the membrane surface [43,46]. When increasing concentrations of AQS were added to three different samples, containing the three polyelectrolytes in buffer, a strong decrease in their fluorescence signal was observed. In contrast, when the same experiment was performed in samples containing the fluorescent nanoparticles, the quenching effect was much less efficient (Figure 4). The Stern–Volmer plots (Equation (4)) were linear in all the studied ranges, with Ksv values similar in the three multicolor fluorescent nanoparticles, ranging from 1.00 × 10<sup>3</sup> M−<sup>1</sup> for red fluorescent nanoparticles to 1.32 × 10<sup>3</sup> M−<sup>1</sup> for green fluorescent nanoparticles. These quenching values are lower than in buffer (Table 2), confirming that the polyelectrolytes are bound to the lipid bilayer but close to the surface, because they are relatively accessible to the quencher. Finally, we compared the Ksv values obtained in the fluorescent nanoparticles with those achieved in anionic (PG) lipid vesicles in fluid-phase, which were ∼0, especially for HTMA-PFP and HTMA-PFNT [46]. These differences in Ksv support the hypothesis previously proposed that the lipid packing affects the mode of interaction between the polyelectrolytes and the lipid membrane, as well as their final location in the bilayer.

**Figure 4.** Stern–Volmer plots for quenching of (**a**) HTMA-PFP (3 μM), (**b**) HTMA-PFBT (3 μM) and (**c**) HTMA-PFNT (3 μM) by AQS in sodium phosphate buffer (squares) and in TSLs (circles).

Once optimized and analyzed, the distribution of components of the fluorescent nanoparticles we characterized by their size and colloidal stability, as well as their morphology (Table 1 and Figure 5). DLS results show that the incorporation of the polyelectrolytes slightly increases the size of the TSLs, which is compatible with their location close to the membrane surface. The decrease in the Zeta Potential was minimum and the nanoparticles exhibited good colloidal stability, preserving their spherical shape.

**Figure 5.** Transmission electron microscopy (TEM) images of blue (**a**), green (**b**) and red (**c**) fluorescent nanoparticles.

The evaluation of the thermosensitive properties of the fluorescent nanoparticles were carried out through two kinds of studies. First, we explored if the incorporation of CPEs modified the thermotropic behavior of the TSLs, previously characterized in Figure 2, and second, we analyzed if the fluorescence of the polyelectrolytes was sensitive to the structural modifications which take place in the vesicle bilayers at the lipid phase transition. For the first purpose, light scattering measurements were performed on the nanoparticles' suspension, as a function of temperature. Figure 6 shows the thermograms recorded for the fluorescent nanoparticles, blue, green and red, as well as for the TSLs in the absence of CPEs. The shape of the obtained curves is similar in the four samples: the light scattered by the nanoparticles slightly decreases as temperature rises, with a sharp drop occurring at Tm, whose value can be obtained from the first derivative plot and is coincident with that obtained in the absence of CPEs (inset in Figure 6). These results indicate that the integration of the polyelectrolytes does not affect the Tm and cooperativity of the lipid transition and thus, the fluorescent nanoparticles display the same thermosensitive properties as the DPPG-TSLs.

**Figure 6.** Effect of temperature on the light scattered by TSLs (squares) and by blue (circles), green (down-triangles) and red (triangles) fluorescent nanoparticles. Inset: First derivative of the thermograms.

As for the second purpose, the emission spectra of the three fluorescent nanoparticles were recorded as a function of temperature and the area under each spectrum was plotted between 20 and 60 ◦C (Figure 7a–c). In these plots, it is possible to distinguish three regions with a different behavior. Between 20 and 32 ◦C, the changes in fluorescence intensity are not very noticeable. However, above 30 ◦C, a rise in the fluorescence intensity is observed, reaching a maximum value around 42–45 ◦C. One possible explanation for this behavior could be that the CPEs are sensitive to the lipid pretransition, which occurs in DPPG at temperatures above 30 ◦C. The lipid pretransition is a transition of low enthalpy occurring below Tm, in which a flat bilayer in gel-phase becomes a periodically undulated membrane, called the ripple-phase [56]. In this ripple-phase, the hydrocarbon chains remain mainly in their rigid, extended, all-trans conformation, like in the gel-phase. However, some works have proposed the possible existence of fluid regions coupled with the geometry of the ripples [56].

**Figure 7.** Effect of temperature on the emission spectrum area of (**a**) blue, (**b**) green and (**c**) red fluorescent nanoparticles.

Probably, when the CPEs are added to the DPPG-TSLs at 24 ◦C, they are mainly adsorbed to the surface instead of being incorporated in the lipid bilayer, as is demonstrated from the quenching experiments. The beginning of pretransition, above 30 ◦C, allows the polyelectrolytes to go into the

lipid membrane, which leads to an increase of fluorescence emission intensity that reaches its plateau close to Tm. Once the fluid phase is reached, the polyelectrolye remains embedded in the bilayer and the fluorescence signal stabilizes or tends to decrease because the probability of nonradiative transitions increases with increasing temperature.

#### *3.4. Nanoparticles as Drug Carriers and Bioimaging Probes*

The above experiments indicate that the fluorescent nanoparticles preserve the thermosensitive properties of the TSLs, but they do not inform if the ability to encapsulate and release hydrophilic compounds triggered by hyperthermia is maintained. In fact, the possible internalization of the CPEs in the lipid membrane above 32 ◦C (as previously suggested), could cause a disruption of the lipid packaging, causing the compounds to be released from the nanoparticle before reaching Tm. To check this possibility, CF was firstly entrapped in the aqueous cavity of TSLs and the CPEs were then added to the suspension, as is described in Scheme 2. To monitor the release of CF from the nanoparticles, we recorded the fluorescence intensity of the dye as a function of temperature (Figure 8). The profile of the curves in Figure 8 was very similar for blue, green and red nanoparticles as well as for TSLs in the absence of CPEs. Above 40 ◦C, an abrupt increase in the fluorescence intensity of CF was detected, which suggests that most of the dye is released from the nanoparticles when the phase transition is reached. However, to confirm this conclusion, it is necessary to perform release kinetics of the dye at different temperatures. Nanoparticles were then exposed to temperatures ranging from 30 to 45 ◦C and the time release profile was recorded at each temperature (Figure 9), as was previously performed for the TSLs suspension. Results indicate that the release kinetics from the nanoparticles presents some differences with respect to those obtained in the absence of CPEs (Figure 2c). The most significant difference is that in the case of nanoparticles, an important fraction of CF is released at 40 ◦C, especially from the green nanoparticles. In addition, ∼5% of the dye is released at 37 ◦C, after different incubation periods, depending on the fluorescent nanoparticles. For the blue one, it was necessary to have more than 900 s of incubation, while for the green and red nanoparticles, the release percentage of 5% was reached at 200 and 400 s, respectively. Therefore, although Figure 8 suggests that the presence of CPEs is not affecting the release properties of the TSLs, the internalization of the polyelectrolytes, which takes place above the pretransition temperature, seems to slightly affect the permeability of the membrane, allowing a small fraction of the dye to be slowly released at temperatures below Tm.

**Figure 8.** Area of the emission spectrum of CF encapsulated in TSLs (squares) and blue (circles), green (down-triangles) and red (triangles) fluorescent nanoparticles as function of temperature (20–55 ◦C).

**Figure 9.** CF released in % as function of time (0–900 s) at different temperatures (30, 35, 37, 40 and 45 ◦C) in blue (**a**), green (**b**) and red (**c**) fluorescent nanoparticles multicolor fluorescent nanoparticles.

Finally, we have performed preliminary experiments to test the capability of the fluorescent nanoparticles to be employed as bioimaging probes. With this end, phase contrast and fluorescence microscopy images of HEK293 cells were taken before and 30 min after the addition of blue, green and red nanoparticles. Figure 10 shows the HEK293 cells in the presence of the polyfluorene-based fluorescent nanoparticles, observed by phase contrast and fluorescent microscopy. Phase contrast and fluorescence images correspond to the same field for each type of fluorescent nanoparticle. These images clearly show that nanoparticles are able to interact with cells, allowing for their visualization in three different colors under fluorescence microscopy. This result extends the applications of these new fluorescent nanoparticles, which could be used as probes for bioimaging while transporting and monitoring the pathway of a drug, controlling its release.

**Figure 10.** Microscopy images of HEK293 cells in the presence of (**<sup>a</sup>**, **d**) blue, (**b**, **e**) green and (**c**, **f**) red fluorescent nanoparticles, observed under (**<sup>a</sup>**–**<sup>c</sup>**) phase contrast and (**d**–**f**) visible-light using the Leica DAPI filter (Ex BP 350/50, Em BP 460/50), DsRed filter (Ex BP 555/25, Em BP 620/60) and FITR filter (Ex BP 480/40, Em BP 527/30).
