thermally induced apoptosis, necrosis, and heat shock...
TRANSCRIPT
Alfred S. SongDepartment of Biomedical Engineering,
The University of Texas,
1 University Station, BME4.202A,
Austin, TX 78712;
Baylor College of Medicine,
1 Baylor Plaza, MS BCM 368,
Houston, TX 77030
e-mail: [email protected]
Amer M. NajjarDepartment of Cancer Systems Imaging,
The University of Texas MD
Anderson Cancer Center,
Unit 059, 1515 Holcombe Boulevard,
Houston, TX 77030
e-mail: [email protected]
Kenneth R. Diller1
Department of Biomedical Engineering,
The University of Texas,
1 University Station, BME4.202A,
Austin, TX 78712
e-mail: [email protected]
Thermally Induced Apoptosis,Necrosis, and Heat ShockProtein Expression in 3D CultureThis study was conducted to compare the heat shock responses of cells grown in 2D and3D culture environments as indicated by the level of heat shock protein 70 expressionand the incidence of apoptosis and necrosis of prostate cancer cell lines in response tograded hyperthermia. PC3 cells were stably transduced with a dual reporter systemcomposed of two tandem expression cassettes—a conditional heat shock proteinpromoter driving the expression of green fluorescent protein (HSPp-GFP) and acytomegalovirus (CMV) promoter controlling the constitutive expression of a “beacon”red fluorescent protein (CMVp-RFP). Two-dimensional and three-dimensional culturesof PC3 prostate cancer cells were grown in 96-well plates for evaluation of their time-dependent response to supraphysiological temperature. To induce controlled hyperther-mia, culture plates were placed on a flat copper surface of a circulating water manifoldthat maintained the specimens within 60.1 �C of a target temperature. Hyperthermiaprotocols included various combinations of temperature, ranging from 37 �C to 57 �C,and exposure times of up to 2 h. The majority of protocols were focused on temperatureand time permutations, where the response gradient was greatest. Post-treatment analy-sis by flow cytometry analysis was used to measure the incidences of apoptosis (annexinV-FITC stain), necrosis (propidium iodide (PI) stain), and HSP70 transcription (GFPexpression). Cells grown in 3D compared with 2D culture showed reduced incidence ofapoptosis and necrosis and a higher level of HSP70 expression in response to heat shockat the temperatures tested. Cells responded differently to hyperthermia when grown in2D and 3D cultures. Three-dimensional culture appears to enhance survival plausibly byactivating protective processes related to enhanced-HSP70 expression. These differenceshighlight the importance of selecting physiologically relevant 3D models in assessing cel-lular responses to hyperthermia in experimental settings. [DOI: 10.1115/1.4027272]
Keywords: heat shock protein, thermal therapy, 3D culture, apoptosis, necrosis
Introduction
Thermoablative cancer therapy has become a widely practicedtherapeutic option. Forms of thermoablative therapy include highintensity focused ultrasound [1], photothermal therapy [2], and ra-dio frequency ablation [3,4]. Understanding the dynamics of heattransfer and tumor tissue response to elevated temperature at thecellular level is critical to develop and optimize this therapeuticmodality. Efficient and effective therapy is defined by maximalablation of malignant tissue (by coagulation necrosis [5,6]) whilesparing surrounding normal tissue as much as possible. Thus,accurate multimodal prediction of the cellular response to hyper-thermia will potentially enable effective planning and control oftherapies to achieve desired outcomes [7,8]. Cell-constitutive andtissue-constitutive data requisite to support this approach are lim-ited, and most useable design data come from in vitro studiesusing traditional 2D cell culture, which may not accurately repli-cate the relevant cellular response in a physiologically relevant3D environment. The objective of this study is to measure the dif-ferences in response of cells grown in 2D and 3D culture tohyperthermia.
The cellular stress response to hyperthermia in virtually all liv-ing organisms is manifested by the expression of HSPs in an orch-estrated fashion to salvage denatured proteins and promote cellsurvival. HSPs are upregulated along with complementary pro-teins when cells experience sublethal stress [9]. As a familyof proteins, they are designated by molecular weight and are
overexpressed in response to various stressful stimuli, includingtemperature extremes [10], hypoxia [11], osmotic stress [12], me-chanical and fluid shear stress [13], and ionizing radiation [14]. Ofthe various HSPs, the 70 kDa family (HSP70) has a central role inthe stress response in conjunction with its greater level of induc-tion [15].
The characteristic preferential transcription and translation ofHSPs are part of a greater stress response by which cellular mech-anisms are activated to promote survival during a window of timeafter an initial insult [16]. The resulting initial and subsequentwindows of tolerance are inferred by the molecular chaperone ac-tivity of HSPs, i.e., the ability to stabilize protein structure and aidin the recovery of other proteins from a denatured state [17]. Howthis response is distinguished in cells grown in 3D culture hasonly seen limited investigation to date [18,19].
Cells in a 3D environment are known to be more tolerant ofchemical stress that is lethal to cells in a 2D environment [20].Modulation of heat and mass transfer processes in a thicker 3Dspecimen may account for significant physiological and environ-mental differences. For example, cells within the interior of anabnormally perfused tumor can experience varying degrees of hy-poxia which may directly induce a low extracellular pH [21,22].Thick specimens may also demonstrate complex thermal proper-ties during simulated ablation [23]. These stresses in turn inducethe expression of stress proteins, such as HSPs that facilitate theirsurvival [24–27]. But the differences in stress response and celldeath kinetics due to thermal insult of 3D and 2D cultures are notyet identified. Candidate sources of disparities include distinctionsin transport processes and intercellular signaling within the cul-ture. The literature that presents data for 3D effects on the viabil-ity of cells after hyperthermia is limited [28]. Based on these
1Corresponding author.Manuscript received January 14, 2014; final manuscript received March 17, 2014;
accepted manuscript posted March 24, 2014; published online May 12, 2014. Assoc.Editor: Ram Devireddy.
Journal of Biomechanical Engineering JULY 2014, Vol. 136 / 071006-1Copyright VC 2014 by ASME
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
documented observations, we hypothesized that cells growing in3D culture, which more closely mimics an in vivo environment,will be more resistant to heat-induced stress response than cellscultured in 2D. The work presented herein provides specific evi-dence of dissimilarities between thermally induced cell death andHSP70 expression in 2D and 3D cultures of prostate cancer cells,including data relating to a potential mechanism of action.
Materials and Methods
Cells. Human prostate cancer cells (American Type TissueCollection, Cat No. CRL-1435TM) were maintained in a tissueculture incubator at 37 �C with 95% relative humidity and 5%CO2 in F12K Kaighn’s Modification (Gibco, Cat No. 21127) sup-plemented with 10% certified fetal bovine serum (Gibco LOT:776620), and 1% Penicillin–Streptomycin. In other experiments,PC3 cells were transduced with a retrovirus to stably express adual reporter system composed of two tandem expression cas-settes—a conditional HSP70 promoter driving the expression ofgreen fluorescent protein (HSPp-GFP) and a CMV promoter con-trolling the constitutive expression of a “beacon” tdTomato(CMVp-tdTomato). These cells, designated PC3-HSPp-GFP/CMVp-tdTomato, were sorted three times to enrich the transgene-expressing population.
Two-Dimensional Cell Culture. Cells to be analyzed by flowcytometry were maintained in polystyrene culture vessels. Thegrowth media was changed every third day, and cells were testedwhen approximately 70% confluent. The cells were passed asneeded, or when plates reached approximately 90% confluence,by washing with phosphate buffered saline, release with a solutionof 0.25% trypsin, and plating into the desired vessel.
3D Cell Culture. Phenol red free high concentration Matrigel(BD Biosciences Cat No. 354262) was used in a 1:2 dilution withcomplete media to create a gelled layer at least 1 mm thick, result-ing in a volume of 35–50 ll in each well of a 96 well plate. Thecoated plate was kept at 4 �C overnight to dissipate air bubbles andto level the surface of the Matrigel. Coated plates were placed in anincubator for at least 2 h to solidify the Matrigel prior to seeding.Control 2D cultures were seeded on the same plates in wells adjacentto Matrigel. After seeding, the cells were allowed to attach overnight(at least 8 h) prior to heating and analysis.
Cell Viability Measurement by Colorimetric Assay. Water-soluble tetrazolium, WST-1, was utilized to determine the thermaldose range at which transition to irreversible cell injury occurs.Cells were cultured in 2D culture until 70% confluent, harvestedwith 0.25% trypsin, counted using a hemocytometer, diluted withcomplete media, and aliquoted into thin walled PCR tubes at adensity of 15,000 cells in 100 ll of growth media. Twelve sam-ples were then heated simultaneously in an Eppendorf thermalcycler across a thermal gradient with temperatures correspondingto 37, 37.4, 38.7, 40.6, 42.9, 45.5, 48.2, 50.8, 53.2, 55.2, 56.5, and57.2 �C for durations of 5, 10, 15, 20, 30, 60, 90, and 120 min. Allsamples were subsequently transferred to an optical 96 well platesafter heating and placed in an incubator. Ten microliters of RocheCell Proliferation Reagent WST-1 (Cat No. 05015944001) wasadded to each well 1 h prior to analysis on a Tecan Safire micro-plate reader 24 h after heat shock.
Heating and Recovery of Cells for Flow Cytometry. A cus-tom designed and fabricated controlled temperature water circula-tion chamber was utilized to heat cells in 96 well culture platesfor subsequent analysis by flow cytometry. The chamber consistedof a flat copper upper surface onto which culture plates could beplaced and a lower volume with inlet and outlet ports throughwhich water from a bath was circulated to cause uniform forcedconvection heating on the bottom of the copper plate. The
chamber maintained the solution within the culture plate wellswithin 60.1 �C of a target temperature in both space and time.The chamber and water bath assembly were contained in an insu-lated housing with air circulated and heated by a proportional-in-tegral-derivative (PID) control system to match the targettemperature. The apparatus was preheated for 30 min prior to useto minimize the thermal transients. Culture plates were placed onthe heated copper surface with a minimal layer of water from thebath to minimize thermal contact resistance between the two solidsurfaces. Understanding that experimental apparatus can affectthe thermal history of specimen [29], this apparatus and protocolrepresent a best effort to reduce thermal transients as well as con-tamination of the cell cultures.
Immediately following heat shock, plates were placed in an in-cubator at 37 �C, 95% relative humidity, and 5% CO2 until 2 hprior to analysis, which amounted to 10 or 22 h of recovery at37 �C followed by 3–4 h of preparation and analysis at 4 �C. Incu-bation of the cells at 4 �C during the preparation time intervalresults in a 4-fold to 6-fold decrease in reaction rate, which indi-cates that cellular kinetics are not significantly altered during thesample preparation time from the experimental end-point untilanalysis [30].
Harvesting of Cells From 2D and 3D Culture for FlowCytometry. Cells were harvested by first liquefying the Matrigelon ice for at least 30 min and transferring the overlaying growthmedia to a 15 ml conical tube prechilled on ice. Ice cold BD cellrecovery solution (Cat No. 354253) and 0.25% trypsin weremixed at a 1:1 ratio, and 200 ll was added to the culture vesselsto further release the cells within the Matrigel layer. An additional100 ll of cold trypsin was added to the culture vessels which werethen incubated at 37 �C for 15 min before the contents were trans-ferred to a 15 ml conical tube. Trypsin was neutralized by adding1 ml of cold media with serum. Cells were centrifuged at 400 g,the supernatant aspirated, and the pellet resuspended in 100 ll ofmedia, and prepared for analysis with either annexin V-FITC andPI or green fluorescence (EGFP).
Annexin V-FITC/PI Staining and EGFP Analysis for FlowCytometry. Flow cytometry analysis was performed on a FACSCalibur flow cytometer (BD Biosciences). The proportion of apo-ptotic and necrotic events within a cell population was determinedby flow cytometry analysis based on annexin V-FITC and PI stain-ing (BD Biosciences Cat No. 556547). Cells which were annexinV-FITC positive but PI negative were considered to have under-gone apoptosis. Cells which were double positive for both fluoro-phores were counted as necrotic. Cultures were plated overnight,heated, and harvested approximately 2 h prior to analysis asdescribed above. Cells were resuspended in 110 ll of media con-taining 5 ll annexin V-FITC and 5 ll PI. Each sample was mixedby gentle pipetting and incubated for 15 min at room temperaturein the dark. Cells were diluted by the addition of 400 ll of mediaand then transferred to plastic flow cytometry tubes. Samples wereplaced on ice until analysis on a BD Biosciences FACS Caliburflow cytometer. Populations were not gated in apoptosis and necro-sis experiments. Thresholds to delineate apoptotic from necroticpopulations were set based on control specimen for each individualexperiment, such that extreme outlying events were excluded.
Expression of HSP70 in PC3-HSPp-GFP/CMVp-tdTomato cellswas determined by measuring GFP expression by flow cytometry.Trypsinized cells were suspended in 400 ll of growth media andGFP expression was determined in the gated viable cell population.
Inhibition of Caspases for Flow Cytometry. Caspase-3 wasinhibited by pre-incubating PC3 cells with Ac-DEVD-CHO (EnzoLife Sciences Cat No. ALX-260-030-M001) at a final concentra-tion of 100 nM for 3 h immediately prior to heat shock.Pan-caspase inhibition was accomplished by pre-incubatingcells with 20 lM Z-VAD-FMK (Enzo Life Sciences Cat No.
071006-2 / Vol. 136, JULY 2014 Transactions of the ASME
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
ALX-260-020-M001) for 30 min prior to heat shock. Heat shockwas carried out on the custom designed heating apparatus at 45 �Cfor 120 min followed by a 22-h recovery time prior to analysis.
Measurement of Caspase Activity. Caspase-3 enzymatic ac-tivity in cell extracts was measured using the substrate Ac-DEVD-AMC (Enzo Life Sciences Cat No. ALX-260-031-M001),which is fluorescently activated upon catalysis by caspase-3. Cellswere grown to 70% confluence in 25 cm2 tissue culture dishes,heated with the water bath chamber, and allowed to recover at37 �C until analyzed at 6, 12, or 24 h. At the time of assay,100–200 ll of lysis buffer was added (5x Lysis Buffer: 250 mMHEPES pH 7.4, 25 mM CHAPS, 25 mM DTT). Cell lysate wascentrifuged at 1000xg to pellet nuclei and debris, and the superna-tant was saved and stored at �20 �C until analysis. Cell lysate (10ll) was added to 90 ll of reaction buffer containing 5 lM Ac-DEVD-AMC 20 mM HEPES pH 7.4, 0.1% CHAPS, 5 mM DDT,2 mM EDTA. Staurosporine-treated Jurkat cell extract served asthe positive control. A Tecan Safire plate reader was used to mea-sure substrate catalysis at excitation and emission wavelengths of360 6 5 nm and 440 6 10 nm, respectively.
Results
High Throughput Identification of Peak Metabolic Activityand Viability Threshold. In order to examine the cellular viabilityduring hyperthermia on a broad scope, a wide spectrum of thermaldoses was applied to PC3 cells grown in 2D culture. Cellular viabil-ity was determined by WST-1 conversion. The objective was toidentify the transition conditions at which hyperthermia results inirreversible cellular injury and cell death. This threshold regioncould be evaluated in greater detail and resolution.
Heating duration ranged from 5 to 120 min at temperaturesranging from 37 �C to 57.2 �C (Fig. 1). The results demonstratetwo salient features. First, a peak in metabolic activity (WST-1indirectly reflects mitochondrial activity) at 42.9 �C for 20 min,and second, a clear temperature and duration threshold beginningat 50.8 �C for 5 min and gradually declining in temperature as du-ration increases beyond which cells are fated to undergo cell deathby apoptosis or necrosis (black region), and below which cellssurvive. This threshold region between cell death and cell viabilitywas selected for further detailed examination to determine therelationship between HSP expression and apoptotic events.
Thermal Dose to Specifically Evoke Apoptosis or Necrosis24 h Poststress in 2D Cell Culture. Based on the identified cellviability threshold region, two hyperthermia temperature/time
conditions were selected to more comprehensively study cell via-bility by the flow cytometry. Apoptosis and necrosis as deter-mined by annexin-V FITC and PI staining, respectively, wereused to more precisely define the thermal dose threshold of viabil-ity (Fig. 2) for a 10 min heating period in 2D cultures. The thresh-old between viability and apoptosis appears to occur between48 �C and 52 �C, and increasing the thermal dose from 52 �C to60 �C for 10 min results in an increasing fraction of cells under-going necrosis as denoted by the percentage of cells which areannexin�/PIþ or annexinþ/PIþ.
Lower temperatures in combination with longer heating dura-tions were tested to characterize the effect of more mild but pro-longed thermal protocols. Figure 3 shows the results of heating at44 �C for 10, 30, 60, or 120 min. All protocols lasting less than120 min resulted in minimal cell death by either apoptosis or ne-crosis. Heating for 120 min caused a more selective shift of theviable population into the apoptotic state.
Data for thermal doses 1 �C higher at 45 �C for 30, 60, 120, or180 min are shown in Fig. 4. The results demonstrate that a 1 �Cincrease above the threshold temperature of 43 �C was sufficientto cause cells to preferentially undergo necrotic cell death. Heat-ing durations of 120 min or more induce a greater portion of cellsto undergo necrosis.
Based on these experiments, the two thermal doses of 44 �Cand 45 �C for 120 min, which straddle the observed thresholdbetween apoptosis and necrosis, were selected for a more refinedanalysis of heat-induced cell injury.
Caspases are Involved in Thermally Induced Apoptosis. Inorder to quantify the involvement of caspases in apoptosis, asindicated by annexin V-FITC staining, PC3 cells grown in 2D cul-ture were treated with pan-caspase inhibitor Z-VAD-FMK prior toheating at 45 �C-120 min (Figs. 5(a)–5(c)). Analysis was by flowcytometry. In the 45 �C-120 min group Z-VAD-FMKþ (Fig. 5(c)),a distinct apoptotic population (annexinþ/PI�) is observed thatdoes not appear in the corresponding 45 �C-120 min, without Z-VAD-FMK-group (Fig. 5(b)). Heating at 45 �C for 180 min didnot yield similar results, and no outstanding difference between Z-VAD-FMK treated and untreated groups was seen.
In an attempt to identify the apoptotic mechanism with an indi-vidual caspase, caspase 3 inhibitor Ac-DEVD-CHO was appliedprior to heating at 45 �C for 180 min and flow cytometry analysis.Figures 5(e) and 5(f) demonstrate no obvious difference betweenthe heated groups that were and were not treated with Ac-DEVD-CHO. This result was further confirmed by using the fluorescentcaspase-3-specific substrate, Ac-DEVD-AMC, on cellular extractsfrom cells heated at various thermal doses. The results of theseexperiments in Fig. 6 demonstrate that none of the thermal dosesused induced any measureable Ac-DEVD-AMC fluorescent signalover the broad range of hyperthermia time and temperature. Inone case, heating at 40 �C for 22 min yielded an elevated back-ground level of fluorescence. However, no time-dependentincrease in signal was detected (as observed in the positive con-trol), dismissing the presence of any positive caspase-3 activity.The positive control demonstrated the predicted ramp up in cap-sase activity. A complementary experiment at 45 �C for extendeddurations of heating up to 180 min also demonstrated the absenceof caspase 3 activity (Fig. 7).
Apoptosis and Necrosis are Mitigated and HSP70 isEnhanced in 3D Culture. To assess differences in cell viabilitybetween 2D and 3D cultures, two identical thermal doses of 44 �Cand 45 �C for 120 min were applied to cells grown in 2D and 3Dcultures. These doses define the threshold conditions for injury in2D cultured cells. The cells were assayed for apoptosis and necro-sis 24 h following hyperthermia. The results reported in Fig. 8show a significantly greater percentage of cells undergoing apo-ptosis and necrosis in 2D culture than in 3D culture for cellsheated at 45 �C for 120 min (n¼ 3, t-test, p< 0.05). There was no
Fig. 1 Normalized WST-1 absorbance of PC3 cells for longdurations of heating at various temperatures at approximately12 h later. Each point is normalized to an average value ofWST1 absorbance for cells heated at 37 �C, as evidenced by theuniform shade in the 37 �C row along the bottom of the figure(n 5 3). Note that the time scale is not linear.
Journal of Biomechanical Engineering JULY 2014, Vol. 136 / 071006-3
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
statistically significant difference between the 2D and 3D data forheating at 44 �C.
HSP70 Expression. PC3-p4rep100 cells, which express EGFPunder the control of the human HSP70 promoter, were heated at44 �C and 45 �C for 10 min and assayed by flow cytometry 12 hlater. The data for HSP70 expression in 2D and 3D cultures arereported in Fig. 9. These results show that 3D cultures heated at45 �C for 120 min, but not 44 �C for 120 min, have a statistically
higher level of HSP70 expression in comparison with 2D cultures(n¼ 3, t-test p< 0.05). This differential is consistent with that forapoptosis and necrosis for the same hyperthermic conditions.
Discussion
In order to investigate the effect of a 3D culture environmenton thermally induced cell death and HSP70 expression, theauthors analyzed cellular heat shock responses across a broad
Fig. 2 Apoptosis and necrosis of PC3 cells heated at various temperatures for 10 min. (a) Unheated control, (b) 44 �C, (c) 48 �C,(d) 52 �C, (e) 60 �C. Cells were analyzed by flow cytometry using annexin-V FITC and PI staining 24 h after heat shock.
071006-4 / Vol. 136, JULY 2014 Transactions of the ASME
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
range of temperatures and heating times using a high-throughputapproach. The array of thermal doses used to generate the viabilitydata in Fig. 1 demonstrates two important phenomena. First, apeak of metabolic activity (maximal WST-1 staining) exists in thelandscape of thermal doses which we theorize corresponds tomaximum cellular activity to repair thermal damage [16]. Second,there is a clear and relatively sharp threshold between viabilityand cell death. The region straddling this threshold between via-bility and cell death was chosen to further investigate apoptosis
and necrosis in 2D and 3D cultures as it provides the greatestresponse contrast as indicated by cellular viability and HSP70expression. Thus, the parameters identified in Fig. 1 were used toguide subsequent flow cytometry experiments.
Analysis of cellular viability by flow cytometry required thereevaluation of heat shock responses across an array of tempera-ture doses as demonstrated in Figures 2–4. These results illustratehow the temperature and time characteristics of an applied ther-mal dose dictate the pathways to cell survival or death. Previous
Fig. 3 Apoptosis and necrosis of PC3 cells heated at 44 �C for various durations. (a) Unheated control, (b) 10 min, (c) 30 min,(d) 60 min, (e) 120 min. Cells were analyzed by flow cytometry using annexin-V FITC and PI staining 24 h after heat shock.
Journal of Biomechanical Engineering JULY 2014, Vol. 136 / 071006-5
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
work in our lab [31,32] and by others [33–35] has demonstrated athreshold of cell viability at a well-defined thermal dose. Thistype of constitutive data may be useful for treatment planning andreal-time estimates of the efficacy of localized thermal therapies.Some types of thermal therapy for cancer may cause sites in thefield of thermal stimulation to exceed 98 �C and peripheral areasto reach 50 �C via ablative methods, which temperatures farexceed threshold injury conditions [36].
Caspase inhibition experiments confirmed the involvement ofcaspases in thermally induced apoptosis of PC3s, although caspase-3 is unlikely to play a direct role as evidenced by specific inhibitionvia Ac-DEVD-CHO. This finding is corroborated by other worksthat demonstrate the pan-caspase inhibitor Z-VAD-FMK affectsthermal, but not radiation, induced apoptosis in a leukemia cellline, HL60 [37]. Others have demonstrated the selectivity of cas-pase 9 in thermally induced apoptosis of Jurkat cells [38].
Fig. 4 Flow cytometry scatter plots of PC3 cells heated at 45 �C for various durations. (a) Unheated control, (b) 30 min, (c) 60min, (d) 120 min, (e) 180 min. Cells were analyzed by flow cytometry using annexin-V FITC and PI staining 24 h after heat shock.
071006-6 / Vol. 136, JULY 2014 Transactions of the ASME
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
The apoptotic population distribution that appears in Fig. 5(c)may be explained by an in vitro artifact referred to as late apopto-sis [39], a process by which cells that have undergone apoptosisas a primary means of cell death progress further into necrotic celldeath or secondary necrosis and become PI permeable. The pro-cess of secondary necrosis is not thought to be observed in vivodue to the presence of phagocytic cells of the reticuloendothelial
system, i.e., tissue macrophages, which clear apoptotic and ne-crotic cells and debris before secondary necrosis can take place.In light of this theory of late apoptosis, it seems that inhibition ofthe caspases may delay the kinetics of this process, thus implicat-ing their role in thermally induced apoptosis. If secondary necro-sis is indeed occurring in these experiments, then the observationtime points for apoptosis and necrosis experiments should be
Fig. 5 Apoptosis and necrosis of PC3 cells heated in the presence of caspase inhibitors. (a)–(c) Pan-caspase inhibition (Z-VAD-FMK) with thermal dose of 45 �C for 120 min. (d)–(f) Caspase 3 inhibition (Ac-DEVD-CHO) with thermal dose of 45 �C for 180min. (a) Unheated control for Z-VAD-FMK, (b) heated without Z-VAD-FMK, (c) heated with Z-VAD-FMK, (d) unheated control forAc-DEVD-CHO, (e) heated without Ac-DEVD-CHO, and (f) heated with Ac-DEVD-CHO. A delay in the progression to late apopto-sis or necrosis is observed in the presence of inhibitor Z-VAD-FMK. No difference is detected in the presence of caspase 3 in-hibitor. Cells were analyzed by flow cytometry using annexin-V FITC and PI staining 24 h after heat shock.
Journal of Biomechanical Engineering JULY 2014, Vol. 136 / 071006-7
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
carefully considered when utilizing the results for data-driventherapeutic systems [7,8]. One option would be to interpolate thekinetics of apoptosis and necrosis separately using the existingdata. Alternatively, the percentages of apoptosis and necrosis canbe summed to generate a predictive algorithm until further experi-mental improvements are made in differentiating between celldeath kinetics.
Having established thermal doses which preferentially lead toapoptosis or necrosis (120 min at 44 �C or 45 �C, respectively),experiments were repeated to investigate the effect that 3D culturehad on the two modes of cell death. Relative apoptotic and ne-crotic events were significantly depressed in 3D culture comparedto 2D culture at a thermal dose of 45 �C for 120 min, whichdefined the threshold region of hyperthermia for induction of ther-mal injury, indicating a role for protective mechanisms specific tothe 3D culture environment. It should be noted that the baselinelevel of cells which undergo apoptosis or necrosis in 3D culturesis higher than in 2D environments. In these studies, the resultswere normalized to untreated control to specifically determine theincrease or decrease of cellular response, as shown in Fig. 8. Theexistence of elevated background levels of apoptotic and necroticcells is consistent with other studies using different 3D matrices,
Fig. 6 Caspase 3 activity of PC3 cells heated at various temperatures and durations assessedby Ac-DEVD-AMC fluorometric assay. Cells heated for 3 or 5 min (blue markers) and 10 or 22 min(red markers) at 37 �C (1), 40 �C (�), 44 �C (w), 50 �C (�), 55 �C (*), 56 �C (�), 57 �C (D), or caspaseextract positive control (3). Cells were processed and analyzed at 6, 12, and 24 h after heatshock. Assay duration is approximately 3 h. Y axis units are arbitrary. No caspase 3 activity fromPC3 cells was detected.
Fig. 7 Caspase 3 activity as assessed by Ac-DEVD-AMC fluo-rometric assay in PC3 Cells. Cells were heated at 45 �C at theindicated durations. Negative and positive controls were waterand apoptotic Jurkat cell extract, respectively. Cells were proc-essed and analyzed 24 h after heat shock. No significant cas-pase 3 activity was detected (n 5 3).
Fig. 8 Apoptosis and necrosis of 2D and 3D cultured PC3 cellsafter heating. Cells were heated at the indicated temperaturesfor 120 min and observed 24 h after heat shock. Samples heatedat 45 �C demonstrate a significant difference (n 5 3, t-testp < 0.05) between 2D and 3D cultures. Populations have beennormalized to respective 37 �C control groups.
071006-8 / Vol. 136, JULY 2014 Transactions of the ASME
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
such as viscose-rayon/styrene butadiene copolymer or agar[19,20].
Upregulation of HSP70 was evaluated as the protective mecha-nism that reduces the incidence of apoptosis and necrosis in 3Dculture. The data in Fig. 9 demonstrate a greater induction ofHSP70 expression in 3D cultures at the same thermal dose (45 �Cfor 120 min) for which there was reduced induction of apoptosisand necrosis in Fig. 8. The reason for increased HSP70 expressionin 3D culture is not well understood. Khoei et al. have reported asimilar observation [19] and attribute the decrease of apoptoticand necrotic events in DU145 spheroids to HSP70 expression.
The role of HSP70 in modulating temperature-induced celldeath is well-established, and its upregulated expression is likelyto be a component of the mechanism by which apoptosis and ne-crosis are reduced in 3D culture. It is well-established that a highlevel of heat shock protein expression and HSP70, in particular,correlates with protection from hyperthermia [9,15,16,40]. How-ever, a direct causal relationship between enhanced thermoprotec-tion and elevated HSP70 expression in 3D culture has yet to beestablished. Other factors may play a role. Extracellular adhesionmolecules or intercellular signaling pathways may plausibly playa role in regulating HSP70 expression in 3D cultures. No endoge-nous intracellular molecules, such as growth factors, are known toaffect HSP70 regulatory pathways, although the inverse paradigmhas been observed [41–43]. No published work was found to sup-port or refute a direct interaction between cell adhesion moleculesand HSP70, but evidence was found for a convergence betweenvarious HSPs and integrins upon the pathway of apoptosis or,more precisely, detachment induced apoptosis (anoikis) [44–54].
Conclusions
Understanding the role of the extracellular environment and itsinfluence on HSP70 expression will facilitate refinement of ther-moablative therapy parameters leading to more effective treat-ments. The potential for advancing thermoablative therapy, asillustrated by this study, highlights the need to more fully eluci-date the role of the extracellular environment in the cellularresponse to hyperthermia.
It can be concluded from this study that the 3D environment ofreconstituted basement membrane reduces the induced cell deathby apoptosis or necrosis of prostate cancer cell line, PC3, to athermal dose (45 �C–120 min) that preferentially evokes necrosis24 h postheat shock. HSP70 expression is not simply enhanced in3D culture for all thermal doses, but rather it is enhanced in thehigher thermal dose of 45 �C–120 min. This finding may allude toa mechanism of HSP70 expression that is affected by both 3Denvironment and thermal dose. Future investigation will focus on
the details of the underlying mechanism of the observed results,which when clearly understood should inform the treatment plan-ning and clinical management of patients who receive localizedthermal therapies.
Acknowledgement
This research was sponsored by National Science FoundationGrants CNS 0540033, CBET 0828131, and CBET 096998, andthe Robert and Prudie Leibrock Professorship in Engineering atthe University of Texas at Austin. The authors also would like toacknowledge Andrew Mark in his contribution to the program-ming of the PID algorithm for the temperature control software ofthe heating apparatus.
References[1] Kennedy, J. E., Ter Haar, G. R., and Cranston, D., 2003, “High Intensity
Focused Ultrasound: Surgery of the Future?,” Br. J. Radiol., 76, pp. 590–599.[2] Lal, S., Clare, S. E., and Halas, N. J., 2008, “Nanoshell-Enabled Photothermal
Cancer Therapy: Impending Clinical Impact,” Acc. Chem. Res., 41(12), pp.1842–1851.
[3] Goldberg, S. N., and Dupuy, D. E., 2001, “Image-Guided Radiofrequency Tu-mor Ablation: Challenges and Opportunities—Part I,” J. Vasc. Interv. Radiol.,12(9), pp. 1021–1032.
[4] Dupuy, D. E., and Goldberg, S. N., 2001, “Image-Guided Radiofrequency Tu-mor Ablation: Challenges and Opportunities—Part II,” J. Vasc. Interv. Radiol.,12(10), pp. 1135–1148.
[5] Brausi, M., Castagnetti, G., Gavioli, M., Peracchia, G., de Luca, G., and Olmi,R., 2004, “Radio Frequency (RF) Ablation of Renal Tumours Does Not Pro-duce Complete Tumour Destruction: Results of a Phase II Study,” Eur. Urol.Suppl., 3(3), pp. 14–17.
[6] Solbiati. L., Ierace, T., Goldberg, S. N., Sironi, S., Livraghi, T., Fiocca, R., Ser-vadio, G., Rizzatto, G., Mueller, P. R., Del Maschio, A., and Gazelle, G. S.,1997, “Percutaneous US-Guided Radio-Frequency Tissue Ablation of LiverMetastases: Treatment and Follow-up in 16 Patients,” Radiology, 202, pp.195–203.
[7] Oden, J. T., Diller, K. R., Bajaj, C., Browne, J. C., Hazle, J. D., Babu�ska, I.,Bass, J., Biduat, L., Demkowicz, L., Elliott, A., Feng, Y., Fuentes, D., Prud-homme, S., Rylander, M. N., Stafford, R. J., and Zhang, Y., 2007, “DynamicData-Driven Finite Element Models for Laser Treatment of Cancer,” Numer.Methods Partial Differ. Equation, 23(4), pp. 904–922.
[8] Fuentes, D., Oden, J. T., Diller, K. R., Hazle, J. D., Elliott, A., Shetty, A., andStafford, R. J., 2009, “Computational Modeling and Real-Time Control ofPatient-Specific Laser Treatment of Cancer,” Ann. Biomed. Eng., 37, pp.763–782.
[9] Lindquist, S., and Craig, E. A., 1988, “The Heat-Shock Proteins,” Annu. Rev.Genet., 22, pp. 631–677.
[10] Lindquist, S., 1986, “The Heat-Shock Response,” Annu. Rev. Biochem., 55, pp.1151–1191.
[11] Yeh, C., Hsu, S., Yang, C., Chien, C., and Wang, N., 2010, “Hypoxic Precondi-tioning Reinforces HIF-Alpha-Dependent HSP70 Signaling to Reduce IschemicRenal Failure-Induced Renal Tubular Apoptosis and Autophagy.” Life Sci.,86(3–4), pp. 115–123.
[12] Kurz, A. K., Schliess, F., and H€aussinger, D., 1998, “Osmotic Regulation of theHeat Shock Response in Primary Rat Hepatocytes,” Hepatology, 28(3), pp.774–781.
[13] Schett, G., Redlich, K., Xu, Q., Bizan, P., Gr€oger, M., and Tohidast-Akrad, M.,1998, “Enhanced Expression of Heat Shock Protein 70 (Hsp70) and Heat ShockFactor 1 (HSF1) Activation in Rheumatoid Arthritis Synovial Tissue. Differen-tial Regulation of Hsp70 Expression and HSF1 Activation in Synovial Fibro-blasts by Proinflammatory Cytokines, Shear Stress, and AntiinflammatoryDrugs,” J. Clin. Invest., 102(2), pp. 302–311.
[14] Calini, V., Urani, C., and Camatini, M., 2003, “Overexpression of HSP70 isInduced by Ionizing Radiation in C3H 10T1/2 Cells and Protects From DNADamage,” Toxicol. in Vitro, 17(5–6), pp. 561–566.
[15] Kregel, K. C., 2002, “Heat Shock Proteins: Modifying Factors in PhysiologicalStress Responses and Acquired Thermotolerance,” J. Appl. Physiol., 92(5), pp.2177–2186.
[16] Diller, K. R., 2006, “Stress Protein Expression Kinetics,” Annu. Rev. Biomed.Eng., 8, pp. 403–424.
[17] Calderwood, S. K., Khaleque, M. A., Sawyer, D. B., and Ciocca, D. R., 2006,“Heat Shock Proteins in Cancer: Chaperones of Tumorigenesis,” Trends Bio-chem. Sci., 31(3), pp. 164–172.
[18] Khoei, S., Fazeli, G., Amerizadeh, A., Eslimi, D., and Goliaei, B., 2010,“Elimination of Enhanced Thermal Resistance of Spheroid Culture Model ofProstate Carcinoma Cell Line by Inhibitors of Hsp70 Induction,” YakhtehMed. J., 12(1), pp. 105–112. Available at: http://celljournal.org/components4.php?rQV==wHQwAkO0JXY0N3XmxHQwYDQ6QWS05WZyFGcfZGfAhjN1EDQ6QWStVGdp9lZ8BEMApDZJxWY0J3bQxWYuJXdvp2XmxHQyATNApDZJ52bpR3Yh9lZ
[19] Khoei, S., Goliaei, B., Neshasteh-Riz, A., and Deizadji, A., 2004, “The Role ofHeat Shock Protein 70 in the Thermoresistance of Prostate Cancer Cell LineSpheroids,” FEBS Lett., 561(1–3), pp. 144–148.
Fig. 9 HSP70 induction in 2D and 3D cultures of PC3 cells at44 �C and 45 �C for 120 min. The difference in HSP70 inductionbetween the 2D and 3D culture was found to be statistically sig-nificant for heating at 45 �C for 120 min (n 5 3, t-test p < 0.001).
Journal of Biomechanical Engineering JULY 2014, Vol. 136 / 071006-9
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms
[20] Sun, T., Jackson, S., Haycock, J. W., and Macneil, S., 2006, “Culture of SkinCells in 3D Rather Than 2D Improves Their Ability to Survive Exposure to Cy-totoxic Agents,” J. Biotech., 122(3), pp. 372–381.
[21] Jain, R. K., 2005, “Normalization of Tumor Vasculature: An Emerging Conceptin Antiangiogenic Therapy,” Science, 307, pp. 58–62.
[22] Harris, A. L., 2002, “Hypoxia—A Key Regulatory Factor in Tumor Growth,”Nature, 2, pp. 38–47.
[23] Santos, I., Haemmerich, D., Pinheiro, C., and Rocha, A., 2008, “Effect of Vari-able Heat Transfer Coefficient on Tissue Temperature Next to a Large VesselDuring Radio Frequency Tumor Ablation,” Biomed. Eng. Online, 7(1), p. 21.
[24] Gomer, C. J., Ryter, S. W., Ferrario, A., Rucker, N., Wong, S., and Fisher, A.M. R., 1996, “Photodynamic Therapy-Mediated Oxidative Stress can InduceExpression of Heat Shock Proteins,” Cancer Res., 56(10), pp. 2355–2360.Available at: http://cancerres.aacrjournals.org/content/56/10/2355.abstract
[25] Luk, J. M., Lam, C.-T., Siu, A. F. M., Lam, B. Y., Ng, I. O. L., and Ju, M.-Y.,2006, “Proteomic Profiling of Hepatocellular Carcinoma in Chinese CohortReveals Heat-Shock Proteins (Hsp27, Hsp70, GRP78) Up-Regulation and TheirAssociated Prognostic Values,” Proteomics, 6(3), pp. 1049–1057.
[26] Heacock, C. S., and Sutherland, R. M., 1990, “Enhanced Synthesis of StressProteins Caused by Hypoxia and Relation to Altered Cell Growth and Metabo-lism,” Br. J. Cancer, 62(2), pp. 217–225.
[27] Li, G. C., and Shrieve, D. C., 1982, “Thermal Tolerance and Specific ProteinSynthesis in Chinese Hamster Fibroblasts Exposed to Prolonged Hypoxia,”Exp. Cell Res., 142(2), pp. 464–468.
[28] Khoei, S., Goliaei, B., and Neshasteh-Riz, A., 2004, “Differential Thermo-Resistance of Multicellular Tumor Spheroids,” Iran. J. Sci. Technol., 28, pp.107–116.
[29] He, X., and Bischof, J. C., 2005, “The Kinetics of Thermal Injury in Human Re-nal Carcinoma Cells,” Ann. Biomed. Eng., 33(4), pp. 502–510.
[30] Connors, K. A., 1990, Chemical Kinetics: The Study of Reaction Rates in Solu-tion, VCH Publishers, Inc., New York.
[31] Rylander, M. N., Diller, K. R., Wang, S., and Aggarwal, S. J., 2005,“Correlation of HSP70 Expression and Cell Viability Following Thermal Stim-ulation of Bovine Aortic Endothelial Cells,” ASME J. Biomech. Eng., 127(5),pp. 751–757.
[32] Wang, S., Xie, W., Rylander, M. N., Tucke, P. W., Aggarwal, S. J., and Diller,K. R., 2008, “HSP70 Kinetics Study by Continuous Observation of HSP-GFPFusion Protein Expression on a Perfusion Heating Stage,” Biotechnol. Bioeng.,99, pp. 146–154.
[33] Samali, A., Holmberg, C. I., Sistonen, L., and Orrenius, S., 1999,“Thermotolerance and Cell Death are Distinct Cellular Responses to Stress: De-pendence on Heat Shock Proteins,” FEBS Lett., 461(3), pp. 306–310.
[34] Huang, C.-T., Lu, Y.-H., and Jen, C.-P., 2010, “Investigation on Supraphysio-logical Thermal Injury in Two Well-Differentiated Human Hepatoma CellLines, Hepg2 and Hep3B,” J. Therm. Biol., 35(8), pp. 411–416.
[35] Marylevitch, N. P., Schuschereba, S. T., Mata, J. R., Gilligan, G. R., Lawlor, D. F.,and Goodwin, C. W., 1998, “Apoptosis and Accidental Cell Death in CulturedHuman Keratinocytes After Thermal Injury,” Am. J. Pathol., 153(2), pp. 567–577.
[36] Madersbacher, S., Pedevilla, M., Vingers, L., Susani, M., and Marberger, M.,1995, “Effect of High-Intensity Focused Ultrasound on Human Prostate Cancerin Vivo,” Cancer Res., 55(15), pp. 3346–3351. Available at: http://cancerres.aacrjournals.org/content/55/15/3346.abstract?sid=e8f2c7a4-bcb4-4695-b408-7c40ee713951
[37] Nijhuis, E. H. A., Poot, A. A., Feijen, J., and Vermes, I., 2006, “Induction ofApoptosis by Heat and c-Radiation in a Human Lymphoid Cell Line; Role of
Mitochondrial Changes and Caspase Activation,” Int. J. Hyperthermia, 22(8),pp. 687–698.
[38] Shelton, S. N., Dillard, C. D., and Robertson, J. D., 2010, “Activation ofCaspase-9, but not Caspase-2 or Caspase-8, is Essential for Heat-Induced Apo-ptosis in Jurkat Cells,” J. Biol. Chem., 285(52), pp. 40525–40533.
[39] Wilson, L., and Matsudaira, P. T., 1995, Cell Death, Vol. 46, Academic Press,New York.
[40] Ritossa, F., 1962, “A New Puffing Pattern Induced by Temperature Shock andDNP in Drosophila,” Cell Mol. Life Sci., 18(12), pp. 571–573.
[41] Todryk, S., Melcher, A. A., Hardwick, N., Linardakis, E., Batemen, A., andColombo, M. P., 1999, “Heat Shock Protein 70 Induced During Tumor CellKilling Induces Th1 Cytokines and Targets Immature Dendritic Cell Precursorsto Enhance Antigen Uptake,” J. Immunol., 163(3), pp. 1398–1408. Availableat: http://www.jimmunol.org/content/163/3/1398.abstract
[42] Yoo, J. C., and Mayman, M. J., 2006, “HSP70 Binds to SHP2 and Has Effectson the SHP2-Related EGFR/GAB1 Signaling Pathway,” Biochem. Biophys.Res. Commun., 351(4), pp. 979–985.
[43] Th�eriault, J. R., Mambula, S. S., Sawamura, T., Stevenson, M. A., and Calder-wood, S. K., 2005, “Extracellular HSP70 Binding to Surface Receptors Presenton Antigen Presenting Cells and Endothelial/Epithelial Cells,” FEBS Lett.,579(9), pp. 1951–1960.
[44] Beere, H. M., Wolf, B. B., Cain, K., Mosser, D. D., Mahboubi, A., and Kuwana,T., 2000, “Heat-Shock Protein 70 Inhibits Apoptosis by Preventing Recruitmentof Procaspase-9 to the Apaf-1 Apoptosome,” Nature Cell Biol., 2(8), pp.469–475.
[45] Ravagnan, L., Gurbuxani, S., Susin, S. A., Maisse, C., Daugas, E., and Zam-zami, N., 2001, “Heat-Shock Protein 70 Antagonizes Apoptosis-InducingFactor,” Nature Cell Biol., 3(9), pp. 839–843.
[46] Kim, Y.-M., de Vera, M. E., Watkins, S. C., and Billiar, T. R., 1997, “NitricOxide Protects Cultured Rat Hepatocytes From Tumor Necrosis Factor-a-Induced Apoptosis by Inducing Heat Shock Protein 70 Expression,” J. Biol.Chem., 272(2), pp. 1402–1411.
[47] Li, C.-Y., Lee, J.-S., Ko, Y.-G., Kim, J.-I., and Seo, J.-S., 2000, “HeatShock Protein 70 Inhibits Apoptosis Downstream of Cytochrome C Releaseand Upstream of Caspase-3 Activation,” J. Biol. Chem., 275(33), pp.25665–25671.
[48] Komarova, E. Y., Afanasyeva, E. A., Bulatova, M. M., Cheetham, M. E., Mar-gulis, B. A., and Guzhova, I. V., 2004, “Downstream Caspases are Novel Tar-gets for the Antiapoptotic Activity of the Molecular Chaperone HSP70,” CellStress Chaperones, 9(3), pp. 265–271.
[49] Creagh, E. M., Carmody, R. J., and Cotter, T. G., 2000, “Heat Shock Protein 70Inhibits Caspace-Dependent and -Independent Apoptosis in Jurkat T Cells,”Exp. Cell Res., 257(1), pp. 58–66.
[50] Bretland, A. J., Lawry, J., and Sharrard, R. M., 2001, “A Study of Death byAnoikis in Cultured Epithelial Cells,” Cell Prolif., 34(4), pp. 199–210.
[51] Frisch, S. M., and Screaton, R. A., 2001, “Anoikis Mechanisms,” Curr. Opin.Cell Biol., 13(5), pp. 555–562.
[52] Grossmann, J., 2002, “Molecular Mechanisms of ‘Detachment-Induced Apo-ptosis—Anoikis’,” Apoptosis, 7(3), pp. 247–260.
[53] Kamarajan, P., and Kapila, Y. L., 2007, “An Altered Fibronectin Matrix Indu-ces Anoikis of Human Squamous Cell Carcinoma Cells by Suppressing IntegrinAlpha V Levels and Phosphorylation of FAK and ERK,” Apoptosis, 12, pp.2221–2231.
[54] Bunek, J., Kamarajan, P., and Kapila, Y. L., 2011, “Anoikis Mediators in OralSquamous Cell Carcinoma,” Oral Dis., 17(4), pp. 355–361.
071006-10 / Vol. 136, JULY 2014 Transactions of the ASME
Downloaded From: http://biomechanical.asmedigitalcollection.asme.org/ on 05/20/2014 Terms of Use: http://asme.org/terms