Anti-HER2 PLGA-PEG polymer nanoparticle containing gold nanorods and paclitaxel for laser-activated breast cancer detection and therapy.

Phase-transition nanoparticles have been identified as effective theragnostic, anti-cancer agents. However, non-selective delivery of these agents results in inaccurate diagnosis and insufficient treatment. In this study, we report on the development of targeted phase-transition polymeric nanoparticles (NPs) for the imaging and treatment of breast cancer cell lines over-expressing human epidermal growth factor receptor 2 (HER2). These NPs contain a perfluorohexane liquid interior and gold nanorods (GNRs) stabilized by biodegradable and biocompatible copolymer PLGA-PEG. Water-insoluble therapeutic drug Paclitaxel (PAC) and fluorescent dye were encapsulated into the PLGA shell. The NP surfaces were conjugated to HER2-binding agent, Herceptin, to actively target HER2-positive cancer cells. We evaluated the potential of using these NPs as a photoacoustic contrast agent. The efficacy of cancer cell treatment by laser-induced vaporization and stimulated drug release were also investigated. The results showed that our synthesized PLGA-PEG-GNRs (mean diameter 285 ± 29 nm) actively targeted HER2 positive cells with high efficacy. The laser-induced vaporization caused more damage to the targeted cells versus PAC-only and negative controls. This agent may provide better diagnostic imaging and therapeutic potential than current methods for treating HER2-positive breast cancer.


Introduction
Breast cancer is the most frequent cancer among women, affecting millions of women's lives each year [1]. In order to improve breast cancer outcomes and survival, early detection is critical. Various detection methods, including mammography, ultrasound (US), magnetic resonance (MR) imaging and other imaging modalities have been routinely used to detect breast cancer. However, deficient resources and the limited sensitivity of the existing methods results in the majority of women being diagnosed in the later stages of neoplastic development [2]. A proposed alternative imaging method that may offer increased sensitivity while remaining non-invasive and less expensive is photoacoustic (PA) imaging. This hybrid imaging modality converts non-ionizing optical energy into ultrasound waves in biological tissues based on tissue optical absorption properties [3][4][5]. Thus, it can provide high quality images with rich optical contrast and US image resolution [6,7]. Abundant endogenous chromophores including hemoglobin, lipid, and melanin provide strong photoacoustic contrast due to their relatively-greater light-absorbance in the relevant optical window (near-infrared (NIR) window in biological tissue). Exogenous contrast agents, such as metal [8,9], carbon [10,11], and organic dye [12] are further used to enhance PA contrast and for applications such as targeted molecular imaging [13,14] and therapy [15].
Approximately 20-30% of all human breast cancers overexpress the human epidermal growth factor receptor 2 (HER2) that contributes to tumorigenesis [16,17]. Herceptin, a recombinant, humanized IgG monoclonal antibody against the extracellular domain of the HER2, is the first line treatment for HER2 positive breast cancers. Literature suggests that HER2 treatment results in improved patient survival [18,19]. Paclitaxel (PAC) is a common chemotherapeutic drug used for breast cancer treatment [20]. However, its non-selectivity and rapid systemic clearance remain major limitations [21,22]. Drug loading into nanocarriers has proven to enhance efficacy and reduce side-effects of chemotherapeutic drugs. Specifically, active targeting of HER2-positive breast cancer cells can be achieved by polymeric NPs coated with antibody Herceptin [23]. However, those NPs deliver the therapeutic payload without providing any imaging contrast to verify that they have accumulated in the target tissue. Combining imaging contrast agents with therapeutic agents in a single formulation would achieve this goal.
For the past decade, researchers have been synthesizing polymer-based multifunctional theragnostic agents for specific breast tumor imaging and therapy in union [24][25][26]. Dong et al. have developed a targeted theragnostic platform which combines US/MR imaging and photothermal therapy (PTT) of breast cancer [27]. The agent was fabricated by coating a gold nanoshell around a Poly (lactide-co-glycolic acid) (PLGA) particle, co-loaded with perfluorooctyl bromide and superparamagnetic iron oxide nanoparticles, followed by conjugation to anti-HER2 antibodies. Cell-targeting studies demonstrated good efficiency of receptor-mediated specific binding of HER2-positive human breast cancer cells and good enhancement of US/MR molecular imaging and cytotoxicity effect of targeted cells. However, the stability of gold nanoshells in vivo required further investigation. Feng et al. assessed multifunctional theragnostic polymer nanoparticles containing red dye for tumor targeting. Their results showed promising fluorescence detection, photodynamic therapy, and PTT in vivo [28]. The limitation of this platform is the use of fluorescence dye which would cause photobleaching, and compromise its effective use for PA imaging in vivo.
In recent years, liquid fluorocarbon phase-change nanodroplets have drawn attention due to their considerable permeability and retention, allowing them to exert their therapeutic effect from deep within a tumor. The phase transition from liquid to gas of these droplets may be achieved by applying low-frequency ultrasound (< 1 MHz) or high intensity focused ultrasound to them [29]. However, the elevated Laplace pressure of these nanodroplets requires the use of significant ultrasound pressure, thereby increasing the risk of unwanted side-effects to the surrounding tissue [30][31][32]. Light triggered optical droplet vaporization (ODV) is a much safer alternative [33,34]. When applied with relatively low-energy laser irradiation, the temperature of optical absorbers within the nanodroplets rises instantaneously, resulting in thermoelastic expansion and the generation of a photoacoustic signal. Furthermore, when laser energy exceeds a certain threshold, the local temperature rise triggers a liquid-gas phase transition which can induce proximal cell membrane damage and cell death [35,36].
In this study, we developed PLGA-PEG polymer nanoparticles loaded with silica-coated gold nanorods (GNRs) and chemotherapeutic drug PAC as theranostic agents for anti-HER2 tumor therapy. We used GNRs as light absorbers, as they strongly absorb NIR light and have strong light-to-heat conversion among most commonly used metallic NPs [37,38]. Silica coating layer stabilizes the GNRs under laser irradiation in addition to amplifying the PA signals [39], making them strong candidates for PA and ODV applications. The silica-coated GNRs were designed to be encapsulated in the core with a perfluorohexane (PFH) liquid interior [36] so that the PLGA shell could be loaded with water-insoluble chemotherapeutics, leaving sufficient surface area for functional and targeting purposes. We used PLGA-PEG copolymers as shell material because hydrophilic polymer PEG has been shown to protect solid particulates from interaction with different solutes [40,41], thus increasing the circulation time of these theragnostic agents in the blood stream. The NP surfaces were conjugated to Herceptin to actively target HER2-positive cancer cells. Particles were internalized through receptor-mediated endocytosis, allowing for the localized delivery of PAC either by eventual degradation of the particle shell or by laser-induced vaporization. We also conjugated fluorescent dye to the shell to enable fluorescence imaging. The efficacy of cancer cell targeting and treatment by laser-induced vaporization and stimulated drug release was investigated. Our theragnostic nanoplatform demonstrated strong potential as a PA imaging contrast and therapeutic agent for selective breast cancer detection and therapy.

PLGA-GNR preparation
Gold nanorods were prepared by seed-mediated growth method [42]. Silica coating and fluorination of the GNRs were accomplished according to modified methods [39,43]. PLGA particles containing PFH, GNRs and fluorescent dyes were prepared using a double emulsion solvent evaporation process [35,36,44]. Briefly, PLGA-PEG polymer (25 mg) and DiI (100 µg) were dissolved in chloroform (1 mL). GNRs in PFH solution (0.5 mL) were mixed with PLGA-PEG solution and emulsified for 45 seconds, with 2-second-on, 1-second-off, 10 W pulses using a tip sonifier (BRANSO, USA). Then the emulsion was homogenized with polyvinyl alcohol solution (6 mL, 4% v/v) for 30 seconds. The final emulsion was mixed with isopropanol solution (10 mL, 2% v/v) and stirred for 3 hours at room temperature to evaporate organic solvents. The final products were collected and stored at 4°C for future use. To load the therapeutic drug PAC into the PLGA-PEG particles, PAC (5 mg) was dissolved in the PLGA-PEG in chloroform solution, and then the double emulsion method was followed.

Particle characterization
The physical and optical properties of the GNRs and PEGylated PLGA-GNRs were characterized. A transmission electron microscope (TEM, FEI Tecnai 20, Philips, Japan) and a scanning electron microscope (SEM, FEI XL30 ESEM, Philips, Japan) were used to measure the NP structure and morphology. For the TEM measurement, a 10 µL diluted solution of particles was placed on a carbon-coated copper grid. After the sample was dried, the images were taken at 300 keV. For the SEM measurement, the sample was coated with a thin layer of gold film and scanned at 20 keV. The particle absorption spectral were measured using a spectrophotometer (Shimadzu UV-3600, Japan) with particles suspended in an aqueous solution in a 1 cm thick covet. The NP concentration and size distribution were measured using an Archimedes device (Malvern Panalytical, UK).

HER2 positive test
The human breast cancer BT474 cells (HER2 positive) and MDA-MB-231 cells (HER2 negative) were purchased from American Type Culture Collection (ATCC) and maintained in a humidified cell incubator at 37°C and 5% CO 2 with Dulbecco's Modified Eagle Media with 10% v/v fetal bovine serum and 1% v/v penicillin-streptomycin. Cancer cells (2 × 10 4 ) were seeded in a 35 mm petri dish for 20 hours. Bovine serum albumin (4 µL, 4% v/v) were added to the dish and incubated with Herceptin (2.2 mg) for 3 hours. The cell dish was rinsed with PBS for three times, and secondary antibodies Alexa Fluor-488 goat anti-human IgG (20 µL) (Thermo Fisher Scientific, Canada) were added to the cell culture dish for one hour. Cells were rinsed with PBS, and imaged using a fluorescence microscope (Olympus, Japan).

Cancer cell targeting
Monoclonal antibody Herceptin was conjugated to the PLGA-GNR particle surface using a carbodiimide technique [45]. Briefly, PLGA particles (5 mg) were incubated with MES buffer (0.1 M, pH 5.5). A mixture of EDC (14.2 mg) and NHS (2.3 mg) were added and agitated at room temperature for 1 hour. After washing with MES buffer (0.1 M, pH 5.5), the sediment was re-dispersed in MES buffer (0.1 M, pH 8). Then excess Herceptin (2.5 mg) was added and kept at 4°C overnight. The resulting solution was washed using PBS and the supernatant was discarded. The final PLGA-GNR-Herceptin nanoparticles were collected via centrifuge (700 x g, 5 min).
To test cell targeting by PLGA-GNR-Herceptin, BT474 and MDA-MB-231 cells were seeded in 35 mm glass-bottom petri dishes (2 × 10 4 cells per dish) for 20 hours, then incubated with fluorescent dye DiI labeled PLGA-GNP-Herceptin in an incubator (37°C, 5% CO 2 ) for 1 hour. Cells were rinsed with PBS for three times to remove loosely attached and free particles in the medium. Fluorescent dye DiO and Hoechst were added to the cell culture dish and incubated for 30 minutes and imaged using a photoacoustic/fluorescence microscope. The targeting efficacy was tested by quantifying the DiI positive fluorescence areas in each sample using the ImageJ software. The experiments were repeated 3 times.

Photoacoustic and fluorescence imaging
The photoacoustic imaging of monolayer cells targeted by PLGA-GNRs were conducted using a SASAM (Saarland Scanning Acoustic Microscope) photoacoustic microscope (Kibero GmbH, Germany) which consisted of an optical microscope (IX81 Olympus, Japan), a 375 MHz single element transducer and a 532 nm focused laser (Teem Photonics, France). The transducer had a −6 dB bandwidth of 180 MHz and lateral/axial resolution of 4 µm/7.5 µm. The laser (330 ps pulse width, 4 kHz repetition rate) was directed through an optical neutral density (ND) filter and coupled into a single-mode fiber (Coastal Connections, USA). The collimated laser beam at the fiber output was directed through the right side port of the SASAM and reflected off a dichroic mirror (Chroma Technology Corp., USA) housed in the IX-81 fluorescence cube turret and focused through a 4X optical objective with a NA of 0.1 onto the sample [46]. The laser pulse fluence after the objective was approximately 20 mJ/cm 2 for imaging, and 90 mJ/cm 2 for therapy. Fluorescence images of cells were acquired by rotating the turret to a fluorescence cube with excitation and emission wavelengths of 480 nm and 520 nm, respectively. A CCD camera was affixed to the IX-81 left side port for image acquisition. The cell dish was placed on a motorized sample stage (Marzhauser Wetzlar, Germany). PA signals of the cells were generated by raster scanning through the overlapping laser/transducer focal spots with a step size of 0.5 µm, and digitized at a rate of 8 Gs/s using a 10 bit digitizer (Acquiris, USA), and averaged 100 times to increase the signal-to-noise ratio.

Cancer cell treatment with paclitaxel and laser
Initially, BT474 cells were seeded in a glass-bottom, 96-well plate (8000 cells per well), 6 wells per group and left overnight in the incubator (37°C, 5% CO 2 ). Targeted and non-targeted PLGA-GNR (100 µL, 375 µg/mL) and PLGA-PAC (100 µL, 75 µg PAC/mL) NPs were added to each well incubated for 1 hour. Then each well was rinsed with PBS (pH = 7.4). After all the treatments were completed, an Annexin V apoptosis assay (Annexin V-FITC and PI) was applied to evaluate cell viability. Fluorescence images were taken at 2-hour and 24-hour post-treatment using an ImagXpress Micro Confocal system (MolecularDevices, USA). Blue fluorescence Hoechst was observed with laser of 405 nm and an emission channel of 413-472 nm. Green fluorescence FITC was observed with a 488 nm laser and an emission channel of 503-588 nm. Red fluorescence PI was observed with a 535 nm laser and an emission channel of 550-650 nm. Early apoptotic quantification was evaluated by measuring the relative positive green fluorescence area on each selected field of view using the ImagXpress analysis system which is based on the area associated with pixels above a fluorescence threshold. For late apoptotic evaluation, cell survival was calculated by counting the number of red fluorescence-stained cells. The average values and standard deviations were based on 6 randomly selected fields of views per well from three repeated treatments.
For laser irradiation on cell damage test, BT474 cells were seeded in a glass-bottom, 96-well plate (8000 cells per well) and were incubated with either targeted/non-targeted PLGA-GNP or PLGA-GNP-PAC, and treated with laser pulses (720 nm, 20 mJ/cm 2 ) (OPOTEK, RADIANT, HE 355 LD) for 5 minutes. The pulse width of the laser was 9 ns with beam diameter of 9 mm. The laser pulse repetition frequency was 10 Hz. After laser treatment, cells were incubated in an incubator (37°C, 5% CO 2 ). At 24-hour post-treatment, cells were incubated with PI and imaged using a fluorescence microscope. Cell damage was evaluated by counting the number of cells stained positively with PI from 6 randomly selected fields of view from two independent repeated treatments.

Statistical analysis
Quantitative data were expressed as the mean ± standard deviation (mean ± SD). Statistical differences among multiple groups were determined by analysis of variance (ANOVA), and data from two independent samples were compared using Student's t test. P < 0.05 was considered statistically significant.

Particle characterization
The PLGA-GNRs were prepared using a double emulsion evaporation process and characterized using the Archimedes device, a TEM, and a SEM. Figure 1(A) illustrates a PLGA-GNR particle active targeting and vaporization route. Herceptin conjugated PLGA particles attach to HER2 through receptor/ligand interaction. Particles are internalized through endocytosis. Upon laser irradiation, the particle undergo vaporization which can cause cell membrane integrity loss. PAC is then released to cytoplasm to cause further cell death. Particle vaporization and drug release work synergistically on cell killing. The particle preparation process is demonstrated in Fig. 1(B). Particle characterization results are shown in Fig. 2. A TEM image of silica-coated GNRs (Fig. 2(A)) exhibits the black cores as the GNRs and the gray shells as the silica-coating layers. Figure 2(B) is a representative TEM image of a single PLGA-GNR particle. The gray portion is the PLGA-PEG polymer shell and the black dots are the GNRs. Figure 2(C) is a representative SEM image of PLGA-GNRs showcasing the spherical shape of the particles and their wide size distribution. Figure 2(D) displays the optical absorption spectral of GNRs, silica-coated GNRs and PLGA-GNRs. The absorption peaks are at 780 nm, 791nm and 720 nm for GNR, silica-coated GNR, and PLGA-GNR, respectively. The red-shift of the peak for silica-coated GNR was due to the NP size increase. The spectrum of PLGA-GNR exhibits a broad peak and blue shift compared with the GNRs. This phenomenon is due to the coalescence of GNRs inside PLGA. Figure 2(E) shows a size distribution of PLGA-GNPs with a mean diameter of around 285 ± 29 nm. Figure 2(F) is a fluorescence image of DiI-labeled PLGA-GNR particles.

HER2 positive test and cancer cell targeting
Cancer cell surfaces were conjugated with Herceptin and were verified by fluorescence labeled goat anti-human antibodies using a fluorescence microscope. In Fig. 3(A), on BT474 cell surfaces, HER2 receptors were conjugated with Herceptin which fluorescence strongly in green due to the AlexaFluor 488 labelling on the secondary antibodies. It indicates the over-expression of HER2 on cell membrane. In comparison, MDA-MB-231 cells show minimal fluorescence in green due to insufficient HER2 receptors present on the cell surfaces.
The appropriateness of using BT474 versus MDA-MB-231 cells as models of HER2-positivity and negativity, respectively, was assessed through a Western Blot analysis. Normalized HER2 protein expression of the positive cell line, BT474, and the negative cell line, MDA-MB-231, were plotted in Fig. 3(B). Two distinct bands could be visualized at 70 kDa and 135 kDa, corresponding to the monomer and dimer forms of HER2, respectively. As indicated by pixel intensity quantification, BT474 cell expressed HER2 by nearly an order greater than MDA-MB-231, confirming that the cell models selected were appropriate for further assessing the contrast between HER2-targetting nanoparticle treatment on cancer lines with versus without the target of interest.
Cancer cell targeting was tested by incubating DiI labeled, Herceptin conjugated PLGA (PLGA-HER) particles with cells for a half hour and imaged using a fluorescence microscope. The results show that after PLGA-HER particle incubation with cells, the BT474 cells demonstrated strong staining on the cell surface due to the presence of Herceptin/HER2 binding, which resulted in a visible increase in amount of attached DiI-labelled PLGA-HER particles (Fig. 3(C) (p)). By comparison, there was negligible staining on the MDA-MB-231 cell surfaces, indicating the absence of Herceptin/HER2 bonding (Fig. 3(C) (l)). The targeting efficiency was also quantified by comparing the DiI fluorescent positive area between Fig. 3(C) (i) and (p), and was plotted in Fig. 3(D). As evidenced in Fig. 3(D), BT474 cell line with DiI red fluorescence was quantified and corresponded to 14-fold increase for the BT474 cell lines compared to the MDA-MB-231 cell lines, confirming the higher uptake due to functionalization for targeted delivery.

Photoacoustic measurements
A monolayer of BT474 cells was targeted by PLGA-GNRs and imaged using a photoacoustic/fluorescence microscope (SASAM). Figure 4(A) shows the schematic PA imaging system and a display of the experimental setup. The transducer and lasers were focused on the sample plane. A bright field image and fluorescence image were taken prior to PA imaging. PA imaging was performed by a raster scan over the sample. A PA image was reconstructed using all radio frequency (RF) signals collected. Laser fluence was low so that no phase transition occurred during the imaging. Figure 4(B) shows a bright field (a), fluorescence (b) and PA image (c) of cells targeted by PLGA-GNR-HER particles. In the fluorescence image (Fig. 4(B) (b), particles, membrane, and cell nuclei were stained with DiI (red), DiO (green), and Hoechst (blue), respectively. In the corresponding spatially co-registered PA image (Fig. 4(B) (c)), the bright spots are the locations of the particles. An RF line was sampled from one of the bright spots in PA image, showing a signal-to-noise ratio to be approximately 20 (Fig. 4(A) (c)). When comparing the fluorescence image and PA image, there is good agreement of the particle locations and morphology between the two images. However, the size and number of the particles do not exactly match, as the sources of signal between the two imaging modalities differs. The signal in the fluorescence images comes from the evenly-distributed DiI in the particle's shell whereas the PA signal originates from the unevenly-distributed GNRs (TEM image of Fig. 2(B)). Furthermore, the resolution of two imaging methods in this instance is different, with spatial resolutions of 1 µm and 4 µm for fluorescence microscopy and PA, respectively. Overall, these images show potential of DiI labeled PLGA-GNRs as dual model imaging contrast agents for PA and fluorescence imaging.  Figure S1).
When the laser fluence was increased to above 90 mJ/cm 2 , PLGA-GNR particles within the cell were vaporized, resulting in bubble formation which was observed under the SASAM microscope. In Fig. 4(B) (d-g), the bubble remained trapped within the cell, slowly expanding over time and eventually damaging the cell membrane and exiting cell. Membrane integrity was lost upon bubble formation, and PI fluorescence was observed within 6 seconds of the membrane damage ( Fig. 4(B) (i)). A video clip of this phenomenon has been included in the supplemental materials (Visualization 1). These results suggest that vaporization inside the cancer cells can cause cell membrane integrity loss, resulting mechanical damage to the cell which further causes cell necrosis. The vaporization alone can achieve cancer cell death in the absence of therapeutic drugs which could potentially reduce the systemic cytotoxicity. Moreover, mechanical disruption of the cell in vivo could also potentially promote abscopal immune responses at untreated tumor sites [47]. The cell membrane disruption might preserve the integrity of cellular components and therefore preserve immunogenic or proinflammatory responses seen in mechanical disruption approaches such as histotripsy [47].

Cancer cell treatment
PLGA particle drug loading efficacy was tested in our previous study which indicated an approximate 78.5% PAC loading efficiency [36]. In this study, particle cytotoxicity was tested by comparing the anti-tumor effect of four groups: a. free PAC, b. PAC loaded into PLGA particles without conjugation to Herceptin (PLGA-PAC), c. particles loaded with PAC plus conjugation to Herceptin (PLGA-PAC-HER), and d. control group, PBS control (-). 70% v/v alcohol was also included as a positive control (+) for the loss of cell viability. A standard Annexin V-FITC assay kit was used to assess cytotoxicity. Annexin V binding to exposed phosphatidylserine on the extracellular surface indicated that apoptosis had been initiated, and PI binding to nuclear DNA reflected complete membrane compromise -a marker of later stages of apoptosis and necrosis. In combination, these two reagents allowed for the differentiation among early apoptotic cells (FITC positive, PI negative), necrotic as well as late apoptotic cells (FITC positive, PI positive), and viable cells (FITC negative, PI negative).
Cells were incubated with Annexin V (FITC and PI) and imaged at 24-hours post-treatment using an ImagXpress system. In Fig. 5(A), the untargeted group of PLGA-PAC (10 µM) showed an observable increase in green fluorescence and a few red fluorescent cells, indicating some degree of early apoptosis and necrotic damage to cells. In contrast, the targeted PLGA-PAC-HER (10 µM) group featured significantly greater cellular damage, with very few nuclei (in blue) remaining adherent to the dish. The remainder of the cells were detached and washed away. The observed cell damage is equivalent to the group with free PAC (50 µM) treatment. This could suggest that PLGA-PAC-HER may be utilized to achieve a similar anti-cancer effect of PAC while decreasing the required concentration 5-fold. Targeted particles can more efficiently deliver drug to the cancer cell cytoplasm, and release drug gradually as the PLGA shell degrades over time. Non-targeting particles (PLGA-PAC) did not remain attached to cells as were removed during the rinsing process. As such, only a small portion of these non-targeting particles were able to be internalized by the cells or left to release drug in the media. These results demonstrate that targeted particles (PLGA-PAC-HER) are effective means of localizing drug payloads in proximity to or within HER2 expressing cancer cells.
Early apoptosis was quantified by estimating the positive green fluorescence areas on randomly selected fields of views (6 views per well, three repeats). The quantified results are shown in Fig. 5(B). At 2-hour post-treatment, free PAC (50 µM) and PAC (10 µM) groups show the highest levels of apoptosis (17% and 10%) due to the high concentration of free PAC in the solution and quick uptake by the cells. For the targeted PLGA-PAC-HER and non-targeting PLGA-PAC treatment groups, early apoptosis was lower than free PAC groups (<10%) because most of the PAC in PLGA shell is not released [48][49][50]. At 24-hours post-treatment, low levels of early apoptosis were observed due to the progression of the cell death to the latter stages of apoptosis and necrosis.
For viability evaluation, cell survival was calculated by counting the number of red and blue fluorescence-stained cells from randomly selected fields of views. Figure 5(C) shows that, at 2-hour post-treatment, cells in most of the groups are still viable by counting the blue nuclei attached to the dish. The groups of PLGA-PAC (10 µM) and PLGA-PAC-HER (10 µM) have fewer surviving cells (75% and 72%, respectively) while the free PAC (50 µM) have more surviving cells (90%) (*p < 0.05). At 24-hours post-treatment, most of the treated groups show low cell viability. PLGA-PAC-HER (10 µM) show the most cellular damage (4% viability) due to continuous drug release. Overall, this result indicates that PAC was delivered to the cell via the targeted PLGA particle and released through a diffusion or erosion process. The targeting approach can achieve higher therapy efficiency with lower doses.
Laser irradiation on cell damage was also evaluated. Figure 6(A) features images of cells with the various treatments. Figure 6(B) is a schematic of the experimental set-up. Cells containing PLGA-GNR-HER and PLGA-GNR-PAC-HER particles were irradiated with the laser beam while the media temperature was kept at 37°C using a water bath. A power meter was used to measure the laser power. Figure 6(C) shows cell viability assessed by counting the PI positive cells (in red) as compared to the control group. At 24-hour post-treatment, most of the cells in the laser treated PLGA-GNR-HER-laser group stain for necrotic cell death, with only 7% of the cells surviving. In comparison, in the non-targeted PLGA-GNR-laser group and laser treated control group, cell survival is 97% and 99%, respectively.
Laser treatment result demonstrated that the damage to cancer cells was most likely caused by laser-induced particle vaporization. Targeted particles either internalized by the cells or in  close proximity to them vaporized upon laser irradiation, causing cell membrane disruption and necrotic cell death (Fig. 4(B)). For the combination of laser and PAC treatment group (PLGA-GNR-PAC-HER-laser), cell survival is about 9% which is similar as the laser treated group (PLGA-GNR-HER-laser). This result indicates that laser treatment of cells with the PLGA-GNR-HER particles does not further require the addition of PAC to achieve the same effect in this in vitro setting. This promising outcome will be the subject of future investigations assessing whether this finding may be replicated under in vivo conditions.
Overall, our ODV effect on cell damage has demonstrated that laser-induced vaporization could cause serious damage to cancer cells and serve as a potential alternative treatment. With the active targeting strategy, a sufficient number of PLGA-GNRs were attached to the cancer cell surface and entered the cell in a short period of time (see Visualization 2), enabling ODV to occur on, near, or inside the cell. After the particle was internalized and before its degradation or excretion [15,51], it was vaporized by the laser to serve its therapeutic purpose.
Our developed PLGA-GNRs have a few advantages over other published PTT methods [27,28,35,38]. Firstly, the laser fluence (20 mJ/cm 2 ) used in our experiment is within the safety standard limit, and is unlikely to cause side effects to the surrounding tissue when applied in in vivo studies. Secondly, researchers have demonstrated that the magnitude of immunostimulation caused by membrane disruption was stronger than that seen with tumor thermal ablation [47]. And it is more likely to trigger potentially therapeutic immune responses against cancer antigens which could enhance cancer cytotoxicity. Lastly, the formed bubbles can serve as US contrast agents which generate additional signal for tumor imaging. Our particle as a multifunctional platform is suitable for in vivo study. Our future work will focus on xenograft mouse tumor models to investigate the effect of active targeting to HER2+ tumors. In the mouse model, PA imaging can be used for functional molecular imaging by utilizing our targeted PLGA-GNRs as a) PA imaging contrast agents to ensure accumulation of the agent in the target tissue and b) therapy agents based on ODV, followed by PA imaging to assess treatment response [52][53][54].

Conclusion
In recent years, nanomedicine has become a promising avenue in the search for effective anti-tumor treatments. Active targeting and delivery can localize high-dose drug payloads and minimize systematic toxicity. Targeted therapeutic approaches without using cytotoxic drugs could also minimize systematic toxicity. In this work, nano-scaled, multi-model phase-change theragnostic agents for combined PA and fluorescence imaging and therapy were developed. We demonstrated their PA imaging and therapeutic capacity. Our results showed that particles with active-targeting capabilities exhibited the highest efficiency for imaging and drug delivery among all treatment groups. Laser-induced vaporization was more efficient in cancer cell treatment compared to the PAC-only treated group and could be an alternative therapy method to reduce the need to include cytotoxic drugs, thereby lessening systemic burden during treatment regimens. We will test the tumor targeting and therapeutic effects of these agents in vivo to further demonstrate their potential for treating breast cancer.