Heat shock protein expression and temperature distribution in prostate tumours treated with laser irradiation and nanoshells

Abstract

Purpose: Sub-lethal temperature elevations in the tumour incurred during laser cancer therapy can induce heat shock protein (HSP) expression leading to enhanced tumour survival and recurrence. Nanoshells utilised in combination with laser therapy can potentially enable selective heat deposition, greater thermal injury, and diminished HSP expression in the tumour. The study objective was to measure the distribution of temperature and HSP expression in prostate tumours in response to laser therapy alone or with nanoshells to determine if these combinatorial therapies can minimise HSP expression.

Methods: PC3 cells were inoculated in the backs of CB17-Prkd c SCID/J mice and treated with external laser irradiation (wavelength of 810 nm, irradiance of 5 W/cm², spot size of 5 mm, and heating duration of 3 min) alone or in combination with gold nanoshells (diameter of 55 nm and outer gold shell thickness of 10 nm) introduced into the tumour 24 h prior to laser treatment. Magnetic resonance temperature imaging was used to measure the distribution of temperature elevation in the tumours during laser treatment. Tumours were sectioned 16 h following laser treatment, stained for Hsp27 and Hsp70, imaged with a confocal microscope, and HSP expression levels were quantified as a function of depth in the tumours.

Results: Maximum temperature elevations at the tumour surface were 28°C for laser treatment only and 50°C for laser heating in combination with gold nanoshells. Laser therapy alone caused significant induction of HSP expression in the first few millimeters of the tumour depth, whereas decreasing HSP expression occurred with greater tumour depth. Tumours treated with laser and nanoshells experienced substantial temperatures (73–78°C) at the tumour surface and temperatures greater than 53°C in the first few millimeters which eliminated HSP expression.

Conclusion: Inclusion of nanoshells in laser therapy can provide a mechanism for enhancing heat deposition capable of eliminating HSP expression within a larger tumour region compared to laser heating alone.

• HSP expression
• laser therapy
• cancer
• prostate
• temperature
• MRTI

Introduction

Laser ablative therapies can provide a minimally invasive and potentially more effective treatment than conventional resection procedures for cancer therapy [1–5]. Treatment volumes and durations are dependent upon the temperature distribution generated by the geometry of the fibre tip and the expanding thermal boundary. Therefore, laser therapy effectiveness is limited due to non-specific heating, which can potentially lead to undesirable healthy tissue injury. Treatment of larger tissue volumes with this method requires thermal diffusion from the source, resulting in extended procedural durations and poorly defined lesion boundaries. The efficacy of laser therapies can be further compromised due to induction of molecular chaperones known as heat shock proteins (HSPs) in regions of the tumour where non-lethal temperature elevation occurs resulting from either insufficient or non-selective thermal deposition [6–10]. Temperatures greater than 43°C can induce HSP expression and cause thermal necrosis. Up-regulation of HSP expression by tumour cells can increase tumour recurrence by enhancing cell viability and diminishing effectiveness of chemotherapy and radiation treatments that are employed in conjunction with hyperthermia [11–14]. HSPs have been implicated in multi-drug resistance [15–17], regulation of apoptosis [18–22], and modulation of p53 [18], [23], [24] for a broad range of neoplastic tissues.

Hsp27 and Hsp70 have been demonstrated to be significantly up-regulated in response to thermal stress with the magnitude of expression directly dependent on temperature elevation and heating duration [6–9], [25–32]. Elevated expression of Hsp27 is a poor prognostic marker for invasive prostatic carcinoma [13] and is associated with tumour cell protection by inhibition of apoptosis [19–24]. Increased expression of Hsp70 mitigates the damaging effects of stress [33], [34] with enhanced survival documented for breast and cervical cancers [24], [35]. The cellular protection provided by HSPs may be attributed to their prominent role in regulation of cell proliferation, prognosis, and drug resistance [17], [36], [37]. A study performed by Beckham et al. illustrated that Hsp70 is critical for cell survival by comparing the responses of Hsp70 knockout murine embryonic fibroblasts (MEF) and control MEF to various mild heat treatments [37]. Recent evidence suggests Hsp70 has a role in the control of cell cycling and growth [38–40]. A positive correlation between Hsp70 levels and proliferative activity has been demonstrated in immunohistochemical studies of breast tumours [12]. Hsp27 and Hsp70 regulate reactive oxygen species by a glutathione-dependent pathway [41], [42], providing protection for intracellular proteins and partially explaining their protective effect against chemotherapeutic agents [14], [15], [38], [43], [44].

Photoabsorbing nanoparticles targeted to tumour cells and used in combination with laser therapy may enable selective control of heat deposition, thermal injury, and HSP expression in the tumour and healthy tissue. Ultimately this control can permit reduced injury to normal surrounding tissue and mitigate the protective effects of HSPs, leading to reduced tumour recurrence [1], [7], [8], [45–52]. Nanoparticles such as gold nanoshells [1], [45–50], single and multi-walled carbon nanotubes [7], [8], [51–56], and single walled carbon nanohorns [56–60] have been shown to possess significant optical absorption in the near infrared wavelength range of 800–1200 nm where light penetration in tissue is in the order of 1 centimeter. When these particles are included in vitro and in vivo and coupled with laser irradiation, greater temperature elevation and tumour cell death is achieved with a lower laser dose [1], [7], [8], [45–60].

Metal nanoshells that consist of a spherical dielectric core (e.g. silica) surrounded by a metal shell (e.g. gold) have been demonstrated as promising photoabsorbers for laser based cancer treatment [1], [45–50]. Nanoshells possess a highly tunable plasmon resonance that determines the particle's scattering and absorbing properties. The plasmon resonance, and in turn the nanoshell's optical properties, can be tuned across a broad range of the optical spectrum from the ultraviolet to the infrared by controlling the relative core and shell thickness [1], [45–50], [61–63]. Previous studies by O’Neal and Hirsch et al. have demonstrated significantly increased temperature elevation measured with magnetic resonance temperature imaging (MRTI) (ΔT = 37°C for nanoshells + laser irradiation and ΔT = 10°C for laser irradiation only) and tumour regression following treatment with a continuous wave diode laser (wavelengths of 808 or 820 nm at an irradiance of 4 W/cm² for durations of 3–6 min) in combination with gold nanoshells compared to laser heating alone [47], [48]. Other groups have compared the heating kinetics, spatial temperature distribution, and morphological alterations in tissue following hypodermic and intramuscular injections of gold nanoshells in combination with irradiation with a continuous wave (wavelength of 810 nm, power of 2 W, duration of 10–360 s) or pulsed diode laser (wavelength of 810 nm, peak power of 8 W, on–off time ratio of 0.25, and pulse duration of 1 ms) [49], [64]. These studies demonstrated the capability of nanoshells in combination with irradiation by a continuous wave or pulsed laser to achieve a temperature rise of 46–50°C within a duration of 40 s and cause targeted tumour destruction at a depth of 4 mm while preserving overlying skin. Gold nanoshells combined with silica rattles and drugs can permit multifunctional photothermal treatment and chemotherapy delivery [65]. Computational models have been created to predict temperature elevation and thermal injury associated with nanoshell inclusion in therapies using continuous wave or pulsed lasers, and confirmed their capability for enhanced heat generation and tumour destruction [64], [66], [67].

Although previous studies have provided knowledge of the relationship of nanoshell properties and laser parameters with optical absorption, temperature elevation, and thermal injury [1], [45–50], the spatial distribution of HSP expression in tumours following treatment with laser irradiation and gold nanoshells has not been measured. This information is critical for accurately predicting the likelihood of tumour survival and recurrence following hyperthermia therapies. We have characterised the thermally induced kinetics of Hsp27 and Hsp70 levels following hyperthermia applied with water bath heating of normal (RWPE-1 cells) and cancerous prostate cells (PC3 cells) for elevated temperatures of 44–60°C and periods of 1–30 min [26]. We have also determined the HSP expression and cell viability of kidney and prostate cancer cells in response to laser heating and multi-walled carbon nanotubes [7]. In order to accurately predict the response of in vivo kidney tumours to laser therapy and multi-walled carbon nanotubes, we have measured the spatial distribution of Hsp27, Hsp70, and Hsp90 and correlated this with MRTI measured temperature and tumour regression [8]. Other studies have utilised a bioluminescent reporter gene to measure Hsp70 expression, determine the zone of thermal damage [68], and provide more comprehensive characterisation of cell death kinetics after heat shock [8], [9]. This technique can be adapted to in vivo models, which is imperative in determining effective thermal therapy protocols [10].

This is the first study measuring and quantifying HSP expression associated with laser irradiation of prostate tumours in combination with nanoshells and correlating this information with temperature distribution measured with MRTI. This paper will explore whether inclusion of nanoshells in laser therapy can effectively minimise HSP expression in the tumour compared to laser heating alone.

Methods

Tumour preparation

Human prostate cancer cells (PC-3) (CRL-1435, American Type Culture Collection (ATCC), Manassas, VA) were cultured with Ham's F12 medium (30-2004, ATCC) supplemented with 10% FBS (30-2020, ATCC) and 5% penicillin-streptomycin (15140-122, Gibco, Carlsbad, CA). On the day of tumour inoculation, PC-3 cells were harvested from culture and counted in a haemocytometer with trypan blue to determine cell number and viability. Cells were centrifuged at 1500 rpm for 7 min. Supernatant was removed and 3 × 10⁷ PC3 cells were resuspended in 1400 µl ice cold BD Matrigel™ Matrix to promote tumour cell growth. Mice were anaesthetised with isoflurane. PC3 cells were inoculated into the backs of 4–6 week old female CB17-Prkd c SCID/J mice with an injection volume of 0.2 mL containing 5 × 10⁶ PC3 cells in Matrigel for each mouse. Prostate tumours were grown to a tumour burden of 1.0 cm in diameter (∼3 weeks). Twenty mice were employed: five controls with no treatment, five with introduction of nanoshells, five with laser heating only, and five with nanoshells and laser therapy.

Laser heating protocols

External laser irradiation was employed rather than intratumoural heating to prevent perturbation of the tissue associated with probe insertion, which could elicit HSP expression due to tissue trauma and inflammation as described previously [48]. The laser parameters employed were laser wavelength of 810 nm, laser irradiance of 5 W/cm², laser spot size of 5 mm, and heating duration of 3 min. The chosen laser parameters enabled induction of HSP expression gradients and varying levels of necrosis in tumours exposed to laser alone or in combination with nanoshells. For treatments utilising nanoshells and laser heating, nanoshells were injected into the mouse tail vein 24 h prior to laser irradiation to enable adequate nanoshell accumulation in the tumour volume as confirmed by prior studies with laser treatment with nanoshells and multi-walled carbon nanotubes [8], [47], [48]. Nanoshells furnished by Nanospectra Biosciences (Houston, TX) and composed of a silica core with a diameter of 55 nm and an outer gold shell thickness of 10 nm were utilised due to their capability for absorption at 808 nm and significant temperature elevation and tumour regression when laser irradiated [48].

Temperature measurement

Magnetic resonance imaging (MRI) was used for planning, localisation, and monitoring of the temperature distribution during laser treatment. The distribution of temperature increase was measured by magnetic resonance temperature imaging (MRTI) via the proton-resonance frequency-shift method [48], [69–72] to facilitate correlation of thermal dose with HSP expression distribution. Imaging was performed in a 1.5 Tesla MR scanner (Signa Echospeed, General Electric Medical Systems, Milwaukee, WI) using a 3-inch (7.62 cm) receive-only surface coil specially designed for small animal imaging (courtesy of R. Giaquinto, General Electric Corporate Research and Development, Schenectady, NY). T1- and T2-weighted images were used to plan and localise the treatment by verifying the position of the laser fibre (via a fiducial marker placed on the fibre) relative to the imaged region prior to irradiation. MRTI was performed by using a complex phase-difference technique applied to a fast, 2-D radiofrequency-spoiled gradient-recalled echo sequence (TR/TE = 74 ms/10 ms, flip angle = 30°, bandwidth = 9.62 kHz). To achieve a 5-s per image scan time, a rectilinear field of view (4 × 2 cm²) was used with a prescribed 256 × 128 matrix and acquired voxel size was 0.16 × 0.31 × 3 mm³. Change in temperature from baseline after N images (ΔTN) was extrapolated from the complex-valued MRTI data (S) by using the temperature dependence of the proton resonance frequency shift [72] and a temperature sensitivity coefficient (α) of −0.01 ppm/°C [72] according to equation 1:where δT_i is the temperature difference measured between the i and (i − 1) images, γB₀ is the resonance frequency (63.87 MHz), and TE is the sequence echo time [72]. The images were smoothed with a 3 × 3 Wiener filter. The temperature resolution for MRTI measured temperature difference is less than 1°C.

HSP expression measurement

Mice were sacrificed 16 hours following laser treatment based on our earlier work, which demonstrated that maximum HSP expression was induced following thermal stress at this time [7], [8], [26]. This time-frame also enabled the maximum HSP expression signal to be measured which is critical to allow sufficient comparison between tumours treated under varying conditions and between different locations within a single tumour. Subsequently, tumours were flash-frozen and embedded into OCT cryo-matrix. At every 1 mm increment within the tumour, five sections with a thickness of 5 microns were obtained. Tumour sections were adhered to poly-lysine coated, wide format glass slides and fixed in cold acetone for 10 min. The location of the laser fibre was denoted on each section as a reference for analysis to enable correlation of HSP expression with probe position and corresponding temperature distribution. Tissue sections were enumerated in a grid-format (e.g. 1.1, 1.2, 1.3, etc.) to ensure that the spatial location of each tissue fragment was known in relation to the distribution of the immunofluorescence staining in each corresponding slice.

Immunofluorescence staining of frozen tissue samples was employed to localise and quantify HSP expression following heating. Tissue sections were incubated with Triton X-100 solution for 15 min to permeabilise cell membranes to HSP antibodies and fluorochromes. All incubations were conducted in an incubator (T = 37°C and 5% CO₂). Primary monoclonal isotype-specific antibodies were employed to increase specificity of epitope binding and decrease background fluorescence. Next, tissue sections underwent blocking in 1.5% goat anti-mouse sera (SC-2043, Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h to decrease non-specific binding. Tissues were then incubated with Hsp70 primary mouse monoclonal IgG_2A antibody (SC-24, Santa Cruz Biotechnology) for 1 h followed by incubation with its secondary antibody Rhodamine Red-X (IgG_2A) (115-295-206, Jackson ImmunoResearch, West Grove, PA) for 1 h. Next, tissue sections were incubated with biotinylated Hsp27 mouse monoclonal IgG₁ primary antibody (ADI-SPA-800B, Assay Designs, Ann Arbor, MI) for 1 h followed by incubation with its secondary antibody, Cy-2 conjugated streptavidin antibody (016-220-084, Jackson ImmunoResearch) for 1 h. Mounting medium was added, tissues were enclosed in cover glass, and samples were stored at 8°C.

Observation of HSP expression levels was accomplished with a 3D laser scanning confocal microscope (Leica SP2 AOBS). An argon laser (λ = 488 nm) was employed for excitation of the Cy2-streptavidin conjugate to enable visualisation and quantification of Hsp27 expression. A helium-neon laser (λ = 543 nm) was utilised for excitation of the Rhodamine Red-X dye permitting visualisation and quantification of Hsp70 expression. HSP expression was quantified using the Leica Lite image processing software. Histograms of fluorescence in the field of view from five different regions in five separate samples were generated and the mean fluorescence of each image was averaged for all five images. Normalised HSP expression concentration, HSPNorm, was calculated with the following equation:where HSP_Final is the HSP concentration of the tumour following laser heating with or without nanoshells, HSP_Control is the HSP concentration for the unheated tumour, FI_Fm is the mean fluorescence intensity following laser heating with or without nanoshells, and FI_Cm is the mean fluorescence intensity for the untreated tumour. FI_Fm and FI_Cm were measured using confocal microscopy and Leica Lite software.

Results

Laser heating only case

Temperature distribution

Figure 1 shows the distribution of maximum temperature elevation measured with MRTI (top right) and maximum temperature increase as a function of time at the tumour surface (bottom centre) during laser treatment (wavelength of 810 nm, irradiance of 5 W/cm², and pulse duration of 3 min) and HSP expression measured 16 h post-treatment (top left) for varying tumour depths. All temperature measurements represent the mean value for five mice (top right). The initial tumour temperature was measured with a thermocouple to be in the range of 23–28°C and the maximum temperature rise was determined by Equation 1. Significant levels of temperature elevation were induced in the tumour surface (ΔT = 25–28°C) with maximum measured temperatures in the range of 48–51°C for all mice. With increasing distance from the tumour surface and incident laser probe, the temperature elevation decreased with temperatures at the base of the tumour of 36–39°C. The average change in temperature versus time at the surface of the tumour (bottom centre) demonstrated sub-lethal temperature elevations.

Figure 1. Hsp27 and Hsp70 expression measured 16 h following laser irradiation for three tumour depths with distinct temperature elevation zones (top left) and corresponding mean MRTI measured maximum temperature elevation during laser irradiation of five mice (top right). Mean measured temperature elevation as a function of time at the tumour surface (bottom centre).

HSP quantification

Three distinct temperature elevation zones were identified as indicated by the red, yellow, and blue circles representing 1 mm incremental distances in tumour depth (Figure 1). Hsp27 and Hsp70 expression was measured and quantified in these regions 16 h following laser treatment using immunofluorescence. Green and red fluorescence denote Hsp27 and Hsp70 respectively. Significant levels of Hsp27 and Hsp70 expression were induced in the tumour surface (denoted by the red circle) where maximum temperatures of 48–51°C (temperature elevations of 25–27°C) occurred for sufficient time. With increasing depth, the level of HSP expression diminished due to decreasing temperature elevation which was less efficient in inducing expression. The 1 mm depth (yellow region) experienced a maximum temperature of 47°C which is capable of inducing HSP expression. The 2 mm region (denoted by blue circle) only reached a maximum temperature of 39°C which is below the threshold of 43°C necessary to elicit HSP expression and therefore this level exhibits only basal HSP expression.

Figure 2 shows histograms of fluorescence counts (i.e. number of cells with a given fluorescence intensity) as a function of fluorescence intensity for varying depth from the tumour surface. For both HSPs, the tumour surface (0 mm) possesses the broadest fluorescence distribution which encompasses the greatest number of fluorescent counts with high fluorescence magnitudes. With increasing depth, the distribution narrows causing the fluorescence counts to be associated with lower fluorescence intensity. The decreasing fluorescence intensity with depth is expected since the temperature elevation diminishes with distance from the tumour surface which is less conducive to inducing HSP expression. The mean fluorescence (FI_Fm) at each tissue depth for all of the five mice was determined using the histograms and analysis methods in Leica Lite image processing software. The normalised HSP concentration was then determined for each mouse using Equation 2. The measured fluorescence at 2 mm was nearly identical to the basal level of fluorescence so it was employed for the basal value. The FI_Fm values at each tissue depth for both Hsp27 and Hsp70 were averaged for all five mice. The average FI_Fm, the associated standard deviation, and calculated normalised Hsp27 and Hsp70 concentrations are shown in Figure 2. The mean fluorescence level for Hsp27 and Hsp70 at the surface (0 mm) is more than twice the level of fluorescence measured at 1 mm. The mean fluorescence intensity decreased approximately 77% and 72% for Hsp27 and Hsp70 respectively as the measurement region moves from 1 mm to 2 mm.

Figure 2. Histograms for quantification of Hsp27 and Hsp70 fluorescence and corresponding Hsp27 and Hsp70 mean fluorescence, standard deviation, and normalised concentration at tumour depths of 0, 1, 2 mm from the tumour surface measured 16 h following laser irradiation.

Influence of nanoshell inclusion on HSP expression

Hsp27 and Hsp70 expression in tumours was measured before (Figure 3A and B) and following nanoshell introduction (Figure 3C and D) to determine whether nanoshell inclusion caused sufficient perturbation to induce HSP expression. Tumours with and without nanoshell inclusion showed identical and faint expression of both Hsp27 and Hsp70 expression, confirming nanoshell inclusion did not cause a difference in HSP expression. The sections shown were acquired from the 1 mm tumour depth and are representative of the entire tumour. A histogram showing fluorescence counts for Hsp27 and Hsp70 as a function of fluorescence intensity for the tumour with nanoshell inclusion depicts a very narrow fluorescence distribution which is nearly identical to the basal level of fluorescence shown in the laser heated tissue for the 2 mm level in Figure 2.

Figure 3. Fluorescence images of Hsp27 before (A) and following nanoshell inclusion (B), Hsp70 before (C) and following nanoshell inclusion (D) and histogram of Hsp27 and Hsp70 fluorescence counts as a function of fluorescence intensity following nanoshell inclusion (E).

Nanoshell + laser heating case

Temperature distribution

Figure 4 shows the distribution of maximum temperature elevation measured with MRTI (top right) and maximum change in temperature as a function of time at the tumour surface (bottom centre) during laser treatment (wavelength of 810 nm, irradiance of 5 W/cm², and pulse duration of 3 min) in combination with nanoshells and corresponding HSP expression measured 16 h post-treatment (top left). All temperature measurements represent the mean value for five mice (top right). The initial tumour temperature was measured to be 23–28°C and the maximum temperature increase was determined by Equation 1. Inclusion of nanoshells in the laser therapy caused more significant temperature elevations over a larger tumour region compared to tumours treated with laser only. Temperature elevations at the tumour surface (denoted by red circle) were in the range of 40–50°C with corresponding maximum temperatures of 73–78°C for sufficient time to cause tissue necrosis. The minimum measured temperature at the tumour base was approximately 46°C.

Figure 4. Hsp27 and Hsp70 expression dual stained image for three distinct tumour depths (top left) measured 16 h following laser irradiation in combination with nanoshells and corresponding MRTI measured mean maximum temperature increase for five mice (top right). Average change in temperature versus time at the tumour surface (bottom centre).

HSP quantification

Hsp27 and Hsp70 expression was quantified 16 h following laser treatment for the three regions denoted in the MRT image in Figure 4 where green and red fluorescence denote Hsp27 and Hsp70 respectively. The tumour surface denoted by the red circle experienced severe temperatures of 73–78°C for sufficient time to cause complete cell death, protein denaturation, and no visible HSP expression. The middle 1 mm region, denoted by the yellow circle experienced significant temperature elevation ranging from 53–56°C. This region experienced significant thermal denaturation of proteins and cell damage yielding very little HSP expression. The bottom tumour level experienced temperatures of 46–50°C which is conducive for HSP expression induction over the period of exposure.

Histograms for quantification of average fluorescence of five mice for the three tumour depths were generated and representative histograms are shown in Figure 5. Due to the significant temperature elevation at the tumour surface of approximately 73°C, the tumour surface (0 mm) exhibited no fluorescence, which is expected due to significant protein denaturation which eliminated HSP expression. As a result, the histogram for the tumour surface is not shown. At the tissue depth of 1 mm, the histogram has a narrow distribution with the majority of fluorescence counts attributed to low fluorescence intensity due to thermal denaturation of proteins characteristic of temperatures of 56°C. At the 2 mm depth, the histogram distribution widens such that a greater number of fluorescence counts are associated with higher fluorescence intensity. This is expected since temperatures of 46°C are effective at inducing HSP expression. Hsp27 and Hsp70 mean fluorescence, standard deviation, and concentration were determined for each tumour level for all five mice using the Leica Lite software and Equation 2, and are shown in Figure 5.

Figure 5. Histograms for quantification of Hsp27 and Hsp70 fluorescence and corresponding Hsp27 and Hsp70 mean fluorescence, standard deviation, and normalised concentration at tumour depths of 0, 1, 2 mm measured 16 h following laser irradiation with nanoshells.

Discussion

This is the first study measuring and quantifying Hsp27 and Hsp70 expression in prostate tumours following laser therapy alone or in combination with gold nanoshells. HSP expression for varying tumour depths was correlated with corresponding temperature distributions measured with MRTI in response to therapy. Knowledge of the relationship between temperature distribution and spatial HSP expression in vivo is critical for accurately predicting regions of treatment success or failure, since HSP expression is an indirect measure of thermal dose and an important indicator of the likelihood of tumour survival and recurrence [6–10]. Ultimately, determination of the thermal response and HSP expression profile in tumours for a given set of treatment parameters can permit selection of laser regimens (e.g. irradiance, duration, and wavelength) and nanoparticle properties (e.g. concentration, type, and delivery method) that create thermal dose distributions that maximise tumour necrosis and minimise HSP expression within the targeted tumour volume.

MRTI enabled spatial measurement of temperature within the tumour volume and confirmed that the inclusion of gold nanoshells significantly increased temperature elevation compared to laser heating alone. Previous studies have utilised MRTI to measure temperature distribution in prostate and kidney tumours treated with laser irradiation in combination with gold nanoshells [48] and multi-walled carbon nanotubes [8]. In this study a laser wavelength of 810 nm, irradiance of 5 W/cm², spot size of 5 mm, and heating duration of 3 min were employed to achieve maximum temperature elevations of 25–28°C for laser only and 40–50°C for laser heating in combination with gold nanoshells. In an earlier study by Hirsch et al. tumours were treated with a wavelength of 810 nm, irradiance of 4 W/cm², beam diameter of 5 mm for 3 min, alone or in combination with gold nanoshells (identical geometry and concentration as the current study). Temperature elevations of 35°C for nanoshells + laser and 10°C for laser alone were measured with MRTI. The larger temperature elevations achieved in our study are attributed to the higher laser irradiance employed. Temperature elevations attained within the tumours treated with nanoparticles and laser therapy have been shown to cause tumour death for both in vitro [7], [8], [48] and in vivo studies [47], [48] resulting in effective prostate [47], [48] and kidney tumour regression [8]. Temperature elevations measured with MRTI for laser and laser + nanoshell treatments also correspond closely with predictions made by published computational models [64], [66], [67].

HSP expression can enable determination of tumour regions with a high likelihood of treatment failure due to their documented role in enhancing cell viability and promoting tumour recurrence [6–16]. In tumours treated with laser alone, maximal Hsp27 and Hsp70 expression was observed at the surface of the tumour proximal to the incident laser which corresponded with peak temperatures in the range of 48–51°C. With decreasing distance from the tumour surface, HSP expression levels diminished corresponding with lower temperature increases which were less conducive to HSP expression induction. In contrast, tumours treated with laser and nanoshells experienced substantial temperatures of 73–78°C at the tumour surface and 53–56°C in the first few millimeters for sufficient time to cause coagulative necrosis and denaturation of HSPs. Induction of HSPs were observed deeper in the tumour at the interface between the tumour and normal tissue. In a prior study, similar HSP expression trends and expression levels were observed for kidney tumours treated alone or in combination with multi-walled carbon nanotubes and 1064 nm continuous wave laser irradiation [8]. Therefore, it is evident that inclusion of nanoshells in laser therapy can provide a mechanism for enhancing heat generation necessary to eliminate HSP expression within a large tumour region. More precise control of HSP expression within the tissue could be achieved through targeting of nanoshells to the tumour cells.

Acknowledgements

This research was funded by the following sources: National Science Foundation Grants CBET 0955072 and 0731108 and National Institute of Health Grant 1 R21 CA135230-01.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.