Investigating cell death responses associated with histotripsy ablation of canine osteosarcoma

Abstract

Background

Osteosarcoma (OS) is the most frequently occurring primary bone tumor in dogs and people and innovative treatment options are profoundly needed. Histotripsy is an emerging tumor ablation modality, and it is essential for the clinical translation of histotripsy to gain knowledge about the outcome of nonablated tumor cells that could remain postablation. The objective of this study was to characterize the cell death genetic signature and proliferation response of canine OS cells post a near complete histotripsy ablation (96% ± 1.5) and to evaluate genetic cell death signatures associated with histotripsy ablation and OS in vivo.

Methods

In the current study, we ablated three canine OS cell lines with a histotripsy dose that resulted in near complete ablation to allow for a viable tumor cell population for downstream analyses. To assess the in vivo cell death genetic signature, we characterized cell death genetic signature in histotripsy-ablated canine OS tumors collected 24-h postablation.

Results

Differential gene expression changes observed in the 4% viable D17 and D418 cells, and histotripsy-ablated OS tumor samples, but not in Abrams cells, were associated with immunogenic cell death (ICD). The 4% viable OS cells demonstrated significantly reduced proliferation, compared to control OS cells, in vitro.

Conclusion

Histotripsy ablation of OS cell lines leads to direct and potentially indirect cell death as evident by, reduced proliferation in remaining viable OS cells and cell death genetic signatures suggestive of ICD both in vitro and in vivo.

Keywords:

• Focused ultrasound
• tumor ablation
• bone cancer
• cell death
• comparative oncology

Introduction

Osteosarcoma (OS) is the most frequently occurring malignant primary bone tumor in dogs with an estimated incidence of 13.9/100,000 cases diagnosed annually in the United States [1,2]. Despite developments in OS treatment options over the past several decades disease prognosis remains grim, and advancement in treatment options to improve survival outcomes are profoundly needed [3–6]. The current definitive standard of care treatment consists of surgical resection of the primary tumor via limb amputation or limb salvage surgery in select cases followed by adjuvant chemotherapy [7,8]. With the current definitive standard of care treatment, the median survival is 10-12 months with the primary cause of death being metastatic disease [9]. Furthermore, canine OS shares numerous biological and pathological similarities with human OS, allowing the dog to serve as a valuable and informative comparative oncology model [1,3,10]. To improve patient prognosis, innovative treatment options for OS are profoundly needed, and advancing treatment options for canine OS has strong potential to translate into improvements in human OS treatment.

Recently, our group has demonstrated the feasibility of treating canine primary bone tumors ex vivo [11] and in vivo [12,13] with the emerging noninvasive focused ultrasound tumor ablation technique histotripsy. Histotripsy is also a nonthermal, and nonionizing focused ultrasound modality that utilizes high-amplitude ultrasound pulses to generate acoustic cavitation and mechanically disintegrate targeted tissue into its subcellular components [14–16]. Histotripsy has previously been employed in canines to ablate experimentally induced prostate tumors [17], spontaneously occurring primary bone tumors [13,18–20], and soft tissue sarcomas [19,20]. In humans, histotripsy has been used to ablate benign prostate tumors, heart valves, and liver tumors [21–25] with preclinical studies investigating breast, brain, pancreas, and liver tumor ablation [22,26].

The destruction of tissue directly targeted by histotripsy results in immediate and direct cell lysis, effectively homogenizing the tissue into subcellular components [15]. While direct tumor cell death is the primary goal of a viable tumor ablation technique, the indirect mechanism(s) by which an ablation technique initiates cell death in nonablated tumor cells, is also important. Cell death mechanisms after tumor ablation can be associated with immunogenic or nonimmunogenic cell death, with immunogenic cell death (ICD) capable of stimulating an antitumor immune response due to the release of inflammatory mediators such as damage-associated molecular patterns (DAMPs), tumor antigens, and cytokines [27–30]. Immune stimulation has the potential to activate an anti-tumor immune response which could decrease metastatic disease development and increase survival expectations [28,31]. Histotripsy ablation results in the release of proteins and signaling molecules such as DAMPs [26,30,32,33], thus demonstrating its potential to initiate ICD signaling and cell death of surrounding tumor cells, and stimulate an antitumor immune response. However, there remains a gap in the knowledge regarding the cell death response of any remaining viable cells post histotripsy ablation, especially in the context of OS. A key goal of histotripsy ablation is to destroy and kill all tumor cells within a targeted area, and it is essential for the clinical translation of histotripsy that knowledge is gained about the survival outcome of potentially non-ablated cells that remain viable within a targeted area. Therefore, the objective of the current study was to characterize the cell death genetic signature and proliferation response of canine OS cells post a near complete (96% ± 1.5) ablation with histotripsy and to evaluate genetic cell death signature associated with histotripsy ablation and OS in vivo. The goal of targeting complete ablation was selected because in future clinical applications the goal will be to achieve complete tumor ablation. We hypothesized that post near complete histotripsy ablation viable OS cells will demonstrate reduced proliferation compared to untreated and sham-treated control cells and that histotripsy ablation will result in a cell death genetic signature associated with ICD compared to untreated OS cell lines and untreated OS patient tumors. To evaluate our hypotheses, we performed in vitro OS cell ablation, quantified the degree of ablation achieved, characterized proliferation of viable OS cells post histotripsy, and evaluated the cell death genetic signature of in vitro and in vivo ablated OS cells/tumors. In our study, the non-ablated cells post histotripsy ablation are referred to as the surviving cells throughout this manuscript.

Methods

Cell culture

The canine OS cell lines D17 (CCL-183, ATCC, Manassas, VA, USA), D418, and Abrams were cultured in DMEM media (ATCC, Manassas, VA, USA), supplemented with 10% FBS and 1% penicillin-streptomycin at 37 °C and 5% CO₂. Cell cultures were passaged every 2–3 days at 70–80% confluency. Passages 5–9 were used for ablation experiments. The D418 and Abrams cell lines were graciously donated by Dr. Jason Somarelli from the Duke Cancer Institute. We chose to evaluate three different OS cell lines instead of a single cell line in efforts to represent the heterogeneous nature of spontaneously occurring OS.

OS patient tumor samples

Primary canine OS tumor samples utilized for the current study were archived formalin-fixed paraffin-embedded (FFPE) tumor samples from canine patients with spontaneously occurring disease (n = 4) collected between September 2020 and March 2021 as previously reported [12]. Briefly, the canine OS patients enrolled in this previous study received histotripsy ablation to a portion of their primary tumor (∼2-cm spherical volume) 24 h before standard-of-care limb amputation surgery. The canine treatments were conducted under the Virginia Tech Institutional Animal Care and Use Committee approval as previously reported [12].

In vitro histotripsy ablation

Prior to histotripsy ablation, OS cells were harvested from cell culture flasks with dissociation reagent TrypLE Express (Thermo Fisher Scientific, Waltham, MA, USA) and resuspended in un-supplemented DMEM at 5x10⁶ cells/mL for histotripsy ablation. A custom-built histotripsy system composed of a 1 MHz 16-element transducer, driven by a high voltage pulser delivering single-cycle pulses, controlled by a field programmable gate array board (Altera DE0-Nano Terasic Technology, Dover, DE, USA), and powered by a high voltage DC power supply (GENH750W, TDK-Lambda, National City, CA, USA) was utilized for all experiments. The system was operated via a custom MATLAB script (The MathWorks, Natick, MA, USA). A 1 ml volume of cells was treated in custom-designed sample treatment tubes composed of acoustically transparent plastic Tygon tubing (Saint-Gobain, Courbevoie, France, cat# B-44-4X) printed elastic resin caps, which were secured to the treatment system with a custom 3D-printed PLA tube interface mount (Figure 1(A,B)). A previously established in vitro histotripsy treatment method was followed [34] with the exception of the pulse repetition frequency (PRF) used, the total number of pulses delivered per treatment, and the addition of lasers to monitor bubble cloud activity during treatments (Figure 1(C)). Single-point treatments were delivered at a PRF of 500 Hz for 5 min, resulting in a total of 150,000 pulses applied per treatment at a pressure level of ∼33MPa. Previously conducted optimization experiments were conducted to determine the total number of pulses that would need to be delivered at a PRF of 500 Hz to achieve 90% or greater cell ablation in a cell suspension of 5 × 10⁶ cells/mL. The fluid flow introduced by histotripsy-induced acoustic cavitation results in the circulation of the cells throughout the sample tube, resulting in cell-bubble cloud interactions. During histotripsy, sample tubes were submerged in degassed water maintained at 37 °C (Figure 1(A–C)). In the current study, an average of 96% ( ± 1.5) of total cells were targeted for ablation to allow for an average 4% surviving cell population to evaluate the cell death and proliferation responses post histotripsy ablation. The percentage of immediate ablated and surviving cell populations were determined by trypan blue viability staining (Figure 1(D)).

Figure 1. Images of the in vitro histotripsy system. (A) Depicts the entire system with essential pieces of equipment labeled. (B) Depicts the custom sample tube submerged in degassed 37 °C water in preparation for histotripsy. (C) Depicts the laser aligned to identify and monitor the cavitation bubble cloud within the sample tube. (D) A representative image of trypan blue-stained treated canine OS cells immediately post-treatment. Cell debris, trypan blue-stained dead cells, and live whole cells are labeled.

The sham-treated cells (5 x10⁶ cells/mL) were subjected to the same treatment setup process as histotripsy-ablated cells, but histotripsy was not delivered. The histotripsy ablated, sham-treated, and untreated OS cells, were maintained at 37 °C and 5% CO₂ for 24 to 72 h. The following downstream experiments were conducted to evaluate ablative outcomes of the three canine OS cell lines post histotripsy ablation: 1) cell viability assessment, 2) morphological changes, 3) proliferation, and 4) cell death genetic signatures (Figure 2.), as detailed in the following sections. The OS cells targeted by histotripsy that were directly ablated by histotripsy are referred to as the ‘histotripsy ablated cells’ and the cells that remained viable post histotripsy ablation are referred to as the ‘surviving cells’.

Figure 2. Schematic of in vitro experimental design. The figure depicts the downstream in vitro cell death experiments carried out for all three canine OS cell lines. Created with Biorender.

Qualitative and quantitative in vitro cell death assessment posthistotripsy

At 24-h post ablation, the histotripsy-ablated, sham-treated, and untreated cells were qualitatively evaluated and examined with an inverted microscope (Olympus IX37) for assessment of cell morphology and the presence of cell debris. Cells were evaluated for condensed rounded appearance, indicative of loss of viability. Cells were also qualitatively evaluated for their degree of adherence to tissue culture-treated surfaces of the cell culture plate. Cellular debris was classified as microscopically observable non-adherent particles which appeared smaller than cells present in culture wells.

Quantitative assessment of cell viability for histotripsy-ablated, sham-treated, and untreated OS cells was performed using the LIVE/DEAD Viability/Cytotoxicity Assay (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer protocol. For analysis, live cells were calcein AM positive and ethidium homodimer-1 negative, dead cells were calcein AM negative and ethidium homodimer-1 positive and dying cells were positive for both calcein AM and ethidium homodimer-1. Double positive cells were classified as dying cells because positive staining for calcein AM indicates metabolic activity and positive staining for ethidium homodimer-1 indicates a compromised cell membrane, both of which occur in dying but not dead cells. Cellular staining was analyzed via flow cytometry with a CytoFLEX (Beckman Coulter, Brea, CA, USA), and the software Kaluza (Beckman Coulter, Brea, CA, USA). Experiments were repeated in triplicates.

Cell proliferation assessment post histotripsy ablation

Cell proliferation of all histotripsy-ablated cells was assessed for 72 h, to allow time for cellular division [35] with a commercial CFSE Cell Division Tracker Kit (BioLegend, San Diego, CA, USA) following the manufacturer protocol. For experimental controls, sham-treated and untreated OS cells were evaluated, and for data analysis controls untreated, sham, and histotripsy-ablated OS cells not stained with CFSE dye were considered. A CytoFLEX flow cytometer was used for data acquisition, and data were analyzed with Kaluza analysis software. For data analysis, cell divisions (divisions 1–4) were quantified via fluorescence intensity of CFSE at 72-h post histotripsy ablation. Undivided (analyzed 5-h post-CFSE staining) and unstained OS cells were used to determine the maximum and minimum fluorescent intensities, respectively. Each half-decade of the log fluorescent intensity scale was gated as a division (divisions 1-4). Percentages represent the percentage of total cells acquired for flow cytometry analysis (30,000 cells for all experiments). For this experiment cells were plated in triplicate at 250,000 cells/mL, and experiments were repeated in duplicates.

In vitro gene expression analysis

At 24 h post-histotripsy ablation, the surviving cells (cells adhered to the cell culture dish), the sham-treated cells, and untreated cells were harvested for RNA extraction for all three cell lines with TRI Reagent (Zymo Research, Orange, CA, USA), and the Direct-zol RNA extraction kit (Zymo Research, Orange, CA, USA) was utilized for RNA extraction following the manufacturer protocol. Sample quantity and quality were evaluated with the NanoDrop One/One^c (Thermo Fisher Scientific, Waltham, MA, USA). All samples had a 260/280 ratio value of 1.8-2.0. Qiagen RT² First Strand Kit (Qiagen, Hilden, Germany) was utilized to synthesize 200 ng of cDNA following the manufacturer’s protocol.

Quantitative real-time PCRs were performed with 200 ng of cDNA, utilizing a custom 96-well RT² PCR Canine Cell Death Array (Qiagen, Hilden, Germany). Each array contained 84 canine genes relating to cell death, five housekeeping genes, and seven internal quality control samples to evaluate genomic DNA contamination, reverse transcription, and assay efficiency (see Supplemental Table 1 for full gene list). Arrays were run on an Applied Biosystems 7500 Fast Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). The cycle threshold (Ct) values acquired from the arrays were analyzed using the GeneGlobe RT² Profiler PCR Data Analysis tool (Qiagen, Hilden, Germany) which calculated fold change relative to untreated cells for each evaluated OS cell line using standard ΔΔCt methodology. The mean of the Ct values for the two housekeeping genes GAPDH and RPLP1 was utilized for normalization because the expression level variation was limited between histotripsy-ablated, sham-treated, and untreated cells, which is desirable for housekeeping gene expression. For all three cell lines, array results are representative of one histotripsy ablation experiment for surviving cells, sham-treated cells, and untreated cells. No statistical analysis was computed due to the pilot characterization nature of this study.

Genes with an up- or downregulation fold change of 2-fold or greater relative to the untreated control group and a fold change in the treatment group that was at least 2x greater than what was observed in the sham treatment group were selected for functional gene enrichment analysis. The public gene ontology (GO) web server, gProfiler [36,37] was utilized for functional enrichment analysis. The time point of 24 h post histotripsy ablation was selected to parallel our clinical trial timepoint [12,13].

In vivo canine cell death genetic signature analysis

RNA was extracted from 20 µm FFPE tissue scrolls from paired histotripsy-ablated and untreated regions of patient tumor samples with a Zymo Quick-RNA FFPE extraction kit (Zymo, California, USA). The manufacturer’s protocol was followed except for the tissue digestion step which was completed with a 12-h incubation period. Qiagen RT² First Strand kit was utilized to synthesize 300 ng cDNA following the manufacturer’s protocol (Qiagen, Hilden, Germany). Quantitative Real-Time PCRs were performed with 300 ng of cDNA, utilizing a custom 96-well RT² PCR Canine Cell Death array as described above (Qiagen, Hilden, Germany). The cycle threshold (Ct) values acquired from array analysis were analyzed using the GeneGlobe RT² Profiler PCR Data Analysis Tool (Qiagen) which calculates fold change with the standard ΔΔCt methodology based on group means. The gene RPLP1 was selected as a reference gene for normalization. The public gene ontology (GO) web server, gProfiler [36,37], was utilized for functional enrichment analysis.

Statistical analysis

Graphing and statistical analyses were performed using GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA). Two-way ANOVAs with Tukey’s multiple comparisons tests were used for the analysis of LIVE/DEAD cell viability results and CFSE proliferation results. For the in vivo canine gene array, significant gene expression changes were determined using a parametric, unpaired, two-sample equal variance, two-tailed distribution Student’s t-test. Statistical analysis was performed by Qiagen Gene Globe analysis software for the in vivo gene arrays. P-value calculation is based on a significant cutoff value of p ≤ 0.05 for all analyses.

Results

Evaluation of histotripsy-induced cell death

Histotripsy was used to generate an average of 96% (±1.5) cell ablation, and we observed evident cell debris immediately post-histotripsy (Table 1., Figure 1(D)). At 24-h posthistotripsy ablation, the viability of histotripsy-ablated, sham-treated, and untreated canine OS cells was assessed. For all three OS cell lines, histotripsy ablation resulted in significantly higher levels of cell death compared to sham-treated and untreated cells, as demonstrated by a significantly larger (p < 0.0001) dead cell population within the histotripsy-ablated cell populations compared to both sham-treated and untreated populations for all three cell lines (Figure 3(A–C)). There were significantly larger live cell populations in sham-treated and untreated cells compared to histotripsy-ablated cell populations, but there were no significant differences between sham-treated and untreated live or dead cell populations. Additionally, there were no significant intra- or inter-group differences for the dying cell population for the three cell lines. At 24-h post histotripsy ablation, within the histotripsy-ablated cell groups, the average live cell population differed amongst the three cell lines, but these differences were not statistically significant (Figure 3(A–C)).

Figure 3. 24-h assessment of histotripsy ablation. (A–C) Graphs of live, dying, and dead OS cells within the histotripsy-ablated, sham-treated, and untreated cell groups. Error bars represent standard deviation (n = 3). Significance was determined with 2-way ANOVA and Tukey’s multiple comparisons tests. **** indicates p < 0.0001.

Table 1. Viability of canine OS cells immediately post histotripsy ablation.

Cell line	% Viable cells post ablation
D17	3.5 ( ± 0.4)
D418	4.16 ( ± 2.8)
Abrams	3.5 ( ± 1.3)

The average percentage of viable cells immediately post histotripsy ablation was calculated based on the initial viable 5x10⁶ OS cells prior to ablation and the number of viable cells after ablation. Standard deviation values noted in parentheses. Cell viability was determined by trypan blue staining.

At 24 h post histotripsy ablation, qualitative morphological assessment of histotripsy-ablated cells revealed predominantly cellular debris with rare viable cells, as hypothesized. The adherent and presumably viable D17, D418, and Abrams cells appeared to have a condensed morphology compared to untreated control cells (Figure 4.). For all three cell lines, sham-treated cells resembled untreated control cells at 24 h post-culture (Figure 4.).

Figure 4. Qualitative cell assessment 24-h posthistotripsy ablation. Images of histotripsy-ablated, sham-treated, and untreated OS cells 24 h post histotripsy ablation. (A–C) D17, (D–F) D418, and (G–I) Abrams.

Evaluation of histotripsy-ablated OS cell proliferation

At 72-h post-histotripsy ablation, the cell proliferation response of all three cell lines was evaluated in respect to sham-treated and untreated control cells. We classified four cell divisions based on fluorescent intensity with division 1 having the greatest fluorescent intensity and division 4 the least fluorescent intensity. As hypothesized, there was a greater population of sham-treated and untreated D17 cells in all cell divisions compared to histotripsy-ablated D17 cells. For all three cell lines, 40% or more of sham-treated and untreated cells were categorized as division 3 based on CFSE fluorescent intensity (Figure 5.). For D17 sham-treated and untreated cells, proliferation remained similar from division 3 to division 4 compared to histotripsy-ablated D17 cells (Figure 5(A)). There was a significantly greater population of a sham treated (p < 0.01) and untreated (p < 0.05) D418 cells compared to histotripsy-ablated D418 cells in division 3 (Figure 5(B)). There was a significantly greater population of sham-treated (p < 0.0001) and untreated (p < 0.0001) Abrams cells compared to histotripsy-ablated Abrams cells in division 3, and in division 4 similar significant trends (p < 0.01) were observed (Figure 5(C)). The histotripsy-ablated D17 and D418 cells demonstrated proliferation up to division 3 as evidenced by a larger population of cells in divisions 2 (12%) and 3 (16%) compared to division 1 (7%) (Figure 5(A,B)). However, the histotripsy-ablated Abrams cells demonstrated a differing proliferative capacity compared to the histotripsy-ablated D418 and D17 cells with declining population of cells from division 1 (7%) to division 3 (3.6%) (Figure 5(C)). For all histotripsy-ablated cells, cell proliferation declined after 3 divisions (Figure 5.). The aforementioned comparisons between the histotripsy-ablated cells were statistically insignificant.

Figure 5. Proliferative capacity of histotripsy-ablated OS cells. Cell divisions [1–4] were quantified via decreasing fluorescent intensity of CFSE at 72-h post histotripsy ablation. (A) The proliferation of histotripsy-ablated, sham-treated, and untreated D17 cells. (B) The proliferation of histotripsy-ablated, sham-treated, and untreated D418 cells. (C) The proliferation of histotripsy-ablated, sham-treated, and untreated Abrams cells. Error bars indicate the standard error of the mean (n = 2). Significance was determined with Tukey’s multiple comparison tests. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

Assessment of cell death genetic signatures of in vitro and in vivo histotripsy-ablated OS

At 24-h post histotripsy ablation, for all three OS cell lines, RNA was extracted from surviving cells, sham-treated, and untreated cells. Cell death genetic signatures were then analyzed (see Supplemental Table 1 for full gene list) to evaluate the indirect cell death outcomes associated with histotripsy ablation of OS. As hypothesized, the changes in the cell death genetic signatures of the surviving cells were associated with ICD pathways, but this response was cell line dependent. The greatest differential gene expression changes were observed in the surviving D17 cells, with 25 upregulated genes and 25 downregulated genes. Functional enrichment analysis of the upregulated genes revealed multiple cell death pathways including apoptosis, necroptosis, and autophagy (Figure 6(A–D)), the NFκB signaling pathway (Figure 6(E)), immune response (Figure 6(F)), and inflammatory response (Figure 6(G)) associated with histotripsy ablation.

Figure 6. Differential gene expression patterns of surviving D17 cells. Bar graphs representing the fold changes of differentially expressed genes in the surviving D17 cells and sham-treated cells relative to untreated cells’ expression level. Genes are categorized by relevant pathways identified by functional enrichment analysis. Functional enrichment analysis showed an association between genes associated with (A&B) Apoptosis, (C) necroptosis, (D) autophagy, (E) NFκB signaling (F) immune response, and (G) inflammatory response.

The surviving D418 cells had two upregulated and two downregulated genes. Functional enrichment analysis of the upregulated genes revealed correlation with the apoptosis pathway (Figure 7(A)). Two of the downregulated genes corresponded with apoptosis as well (Figure 7(B)), and both up- and downregulated genes were associated with NFκB signaling (Figure 7(C)).

Figure 7. Differential gene expression patterns of surviving D418 cells. Bar graphs representing the fold changes of differentially expressed genes in the surviving D418 cells, and sham-treated cells relative to untreated D418 cells. Genes are categorized by relevant pathways identified by functional enrichment analysis and their regulation (up or down) pattern. Functional analysis of differential expression changes in the D418 group revealed pathways associated with A, B) Apoptosis and C) NFκB signaling.

The surviving Abrams cells had limited differential gene expression compared to sham-treated and untreated controls with only genes CCDC103 (-2.00), NOL3 (-2.04), and TNFRSF1A (-2.00) exhibiting a fold change which met the described inclusion criteria. These genes were downregulated, and functional enrichment analysis discovered no intersection between these three genes. However, NOL3 and TNFRSF1A were associated with the biological pathways of tumor necrosis factor-mediated signaling and cellular response to tumor necrosis factor (Figure 8.).

Figure 8. Differential gene expression patterns of surviving Abrams cells. Bar graph representing the fold changes of differentially expressed genes in Abrams surviving treated and sham-treated cells relative to untreated cells’ gene expression level. There was limited differential gene expression changes in the surviving Abrams cells.

In the histotripsy-ablated tumor samples (n = 4), there were 11 genes significantly upregulated in the ablated regions of the tumors compared to paired untreated tumor regions: BAX, BCL2, BCL2L1, CASP4, CFLAR, DENND4A, FAS, HTT, NFKB1, PTEN, and SQSTM (Figure 9). As hypothesized, functional gene enrichment analysis indicated changes are most strongly correlated with biological pathways relating to cell death and inflammation including apoptosis, autophagy, and necroptosis (Figure 9).

Figure 9. Ex vivo genetic cell death signature assessment. The 11 significantly upregulated genes in histotripsy-ablated canine OS tumor samples compared to paired untreated tumor regions (n = 4) were associated with (A) Apoptosis, (B) autophagy, and (C) necroptosis.

Discussion

This study was conducted to evaluate the response of the remaining viable OS cells after near complete histotripsy ablation (96% ± 1.5) of the OS cells were immediately ablated with histotripsy. It has previously been reported that the primary means of cell death associated with histotripsy is immediate cell lysis by direct cell-cavitation bubble cloud interaction [15,16], but there is limited knowledge about the indirect cell death outcomes (i.e. outcome of non-ablated cells within a targeted ablation region), especially in the context of histotripsy ablation of OS. Therefore, in the current study, we sought to gain a further understanding of the indirect cell death outcomes associated with histotripsy in OS by evaluating the cell death response of remaining surviving OS cells 24 h after in vitro ablation, genetic signatures at 24-h post in vitro and in vivo ablation, and cell proliferation at 72-h post in vitro ablation. For all three evaluated canine OS cell lines (D17, D418, and Abrams) mechanical destruction was achieved and the average live and dead cell populations within the histotripsy ablated cell group did not significantly differ between the three OS cell lines, suggesting consistent ablative outcomes at 24-h posthistotripsy. However, at 72-h posthistotripsy ablation, the D17 and D418 cell lines demonstrated greater proliferative capacity than the Abrams (Figure 3.). This suggests that Abrams cells were potentially more acutely susceptible to the indirect cell death outcomes associated with histotripsy compared to the D17 and D418 cells which still demonstrated proliferation, although significantly less than healthy control cells. The CFSE proliferation assay utilized in this study provided insightful preliminary data regarding proliferative capacity of non-ablated cells. Future in vitro studies could include survival/proliferation assays with a higher degree of sensitivity such as the clonogenic assay to provide further insight into the potential capability of non-ablated cells to regrow in a monoculture setting.

Additionally, despite similar response to the same acoustic intensity (i.e. histotripsy ablation parameters) we observed differences in the cell death genetic signature of surviving OS cells 24 h post histotripsy ablation amongst the three cell lines. The surviving D17 cells had 50 genes with differential expression patterns that met the inclusion criteria for functional gene enrichment analysis and the D418 cells had 4 genes (Figures 6,7 and Supplemental Table 2.). We observed limited differential gene expression changes in surviving Abrams cells with only the genes NOL3, CCDC103, and TNFRSF1A being slightly downregulated (Figure 8.). This further exemplifies that while mechanical ablation with histotripsy of all three canine OS cell lines was consistently achieved, the downstream response within the evaluated 24- and 72-h time periods varied between the three cell lines.

The differences in downstream response to histotripsy ablation of canine OS cells could be explained by the differences in behaviors, characteristics, and responses of different canine OS cell lines [38]. Moreover, tumor progression and the metastatic potential of canine OS cell lines are reported to differ in preclinical xenograft rodent models depending on the cell line utilized for experimental tumor induction [38,39]. In addition to differences in metastatic potential, murine models of canine OS have shown differences in responsiveness to chemotherapy treatment depending on the canine OS cell line utilized for the rodent xenograft model [40,41]. Our observed differences between the three OS cell lines also relate to the inherent heterogeneous nature of spontaneously occurring OS in canine patients and suggest that the cell death response to histotripsy ablation may be patient dependent as well. We have demonstrated in vitro that even when near complete cell ablation is achieved with histotripsy, the level of surviving tumor cell aggressiveness and resistance posthistotripsy ablation differs between cell lines, and this may also be the case for individual OS patients. This highlights the potential need to consider personalized cancer treatment approaches to help predict individual patient responses to histotripsy ablation and tailor individual patient treatments accordingly. Future studies investigating the cell death response post histotripsy ablation using patient-specific models such as patient-derived tumor organoids would provide further insight into patient-dependent responses post histotripsy.

In the current study, we also evaluated the cell death genetic signature in histotripsy-ablated canine OS tumor samples to gain preliminary insight into the in vivo cell death response to histotripsy ablation of OS. While we expectedly observed variation between our patients’ samples due to tumor heterogeneity between patients, we also observed significant mean upregulation of 11 genes in histotripsy-ablated portions of tumors compared to paired untreated regions (Figure 7.). The 11 genes were associated with the cell death signaling pathways apoptosis, autophagy, and necroptosis. The cell death genetic signature alterations observed both in vitro in the D17 and D418 cell lines and in vivo suggest that at 24 h post-ablation, the cell death outcomes associated with histotripsy may include ICD mechanisms. However, our 24-h post-ablation evaluation of the cell death genetic signature of Abrams cells does not strongly suggest histotripsy ablation stimulates ICD for this cell line. Additionally, it is worth noting that our in vitro experiments only evaluated the response of canine OS cells (monoculture) vs our in vivo patient tumor sample evaluation consisted of OS tumor cells, and the immune and stromal cells that also compose the tumor microenvironment.

Our in vitro and in vivo functional gene enrichment analysis included multiple cell death pathways that can stimulate ICD, such as apoptosis, necroptosis, autophagy, and the pro-inflammatory NFκB pathway, suggesting that histotripsy ablation can stimulate ICD [28, 29, 42]. Pro-inflammatory signaling pathways such as the NFκB pathway contribute immensely to the immunogenicity of cell death mechanisms such as apoptosis and necroptosis and are essential for ICD and stimulation of an anti-tumor immune response [28,43–45]. In the D17 cells that survived 24 h after histotripsy ablation the majority of differentially expressed genes associated with cell death and damage response were upregulated, but cell death and damage response genes including CASP7, CASP3, and CASP2 were downregulated as well (Figure 6(B)). The inhibition of caspases is associated with immunogenic apoptosis [42] and thus the downregulation of caspase genes in this current study may indicate ICD stimulation. However, further investigation is needed to determine the significance of the observed differential gene expression changes.

In support of our results, a previous investigation has reported differential gene expression patterns associated with ICD in histotripsy-ablated breast cancer cells [30]. In this previous investigation, an upregulation of genes associated with immune response and cell death such as FASLG, TNF, IFNγ, and MAG, was observed these genes were also upregulated in surviving D17 cells in our study [30]. Additional investigations have reported an increase of DAMPs, such as HMGB-1 (30, 33, 34) and DNA [34], both in vivo and in vitro after ablation of tumor cells with histotripsy. The release of DAMPs triggers inflammation [46] and is associated with apoptosis [47], necrosis [48], and ICD [28,29,42], suggesting a role of DAMPs in ICD potentially stimulated by histotripsy. Further investigation is warranted to fully elucidate the indirect cell death outcomes associated with histotripsy and OS and the potential relationship with DAMPs. Our results suggest that histotripsy may be able to induce ICD in OS, but further investigations evaluating beyond cell death genetic signature, such as investigations to characterize apoptotic and necroptotic cellular proteins, DAMP production, and pro-inflammatory cytokines are needed to fully elucidate direct and indirect cell death outcomes.

Our in vitro study results can provide insight for optimizing in vivo treatment planning as well. Our results suggest that even when striving for complete tumor ablation, if complete ablation is not achieved, exposure of surviving tumor cells to histotripsy ablated cells may indirectly stimulate cell death. Determining the need to achieve complete ablation of a targeted tumor is especially important for the clinical advancement of histotripsy for tumors which are difficult to identify via ultrasound during treatment, such as bone tumors like OS. Additionally, our results suggest that ablating less than 100% of a tumor volume may over time still lead to complete tumor necrosis potentially due to stimulation of ICD and the resulting immune response. However, future studies investigating partial ablation of cell suspensions such as 25-100% ablation are warranted to gain future insight on the effect of partial evaluation. Previous preclinical rodent model studies have demonstrated an abscopal immune response associated with histotripsy ablation which targets non-ablated untreated tumors [33] or that partial tumor ablation results in full tumor necrosis over time [32,49]. Future preclinical orthotopic murine model studies would be informative to determine the level of tumor ablation required to lead to complete tumor cell death and the eventual elimination of the entire OS tumor. Results from preclinical studies could then inform future canine clinical studies that investigate long-term outcomes of in vivo histotripsy ablation of OS. Additionally, clinical studies that monitor for radiologic evidence of tumor necrosis via imaging such as MRI scans, after delivering histotripsy to planned tumor ablation volumes in canine OS patients are also warranted.

The use of an in vitro treatment model provides an opportunity to investigate physiological mechanisms such as cell death and cell proliferation. The evaluation of OS patient samples allowed us to correlate our in vitro findings with clinical application. Histotripsy ablation for OS is novel, and additional in vitro, in vivo, and ex vivo studies investigating cell death mechanisms, immunomodulatory potential, and ablative outcomes of histotripsy will advance progress toward successful clinical translation of histotripsy ablation for OS. It is essential for future investigations to include both in vitro and in vivo pre-clinical or clinical studies to evaluate the potential response differences of cell suspensions to histotripsy ablation compared to the response of solid tumors. Overall, the results of our study provide evidence that histotripsy ablation likely results in decreased tumor cell proliferation in non-ablated cells of a targeted tumor volume and both in vivo and in vitro up-regulation of cell death genes correlated with ICD. Histotripsy ablation for OS is novel, and additional studies investigating indirect and direct cell death mechanisms, immunomodulatory potential, and ablative outcomes of histotripsy will help advance the clinical translation of histotripsy ablation.

Authors’ contributions

A.H. – Primary contributor to the composition of the manuscript, lead the execution of all experiments, and performed data analysis.

E.R.V- Assisted in histotripsy equipment design, in vitro experiments, and contributed to manuscript writing.

M.P.- Assisted with in vitro experiments. Contributed to manuscript revision.

J.G. and L.R.- Assisted in histotripsy equipment design and experimental design, and trained A. H. to perform in-vitro treatments for the current study. Contributed to manuscript revision.

I.C.A- Contributed to study design and manuscript revision.

E.V.- Oversaw the engineering of the histotripsy equipment and study design. Contributed to manuscript writing and revising.

J.T – Oversaw study design and contributed to manuscript writing and revising.

All authors reviewed the manuscript.

Acknowledgments

The authors would like to acknowledge Vlaisavljevich lab members Alex Simon, Hannah Covell, and former member Alissa Hendricks-Wenger for their efforts in equipment and experimental design for the in vitro histotripsy systems used in this study. Lauren Ruger was supported by an ICTAS Doctoral Scholarship from the Virginia Tech Institute for Critical Technology and Applied Science. Jessica Gannon was supported by an NSF Graduate Research Fellowship.

Disclosure statement

Eli Vlaisavljevich has an ongoing research partnership and financial relationship with HistoSonics, Inc. No other authors have a conflict of interest to report. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Data availability

The data that support the findings of this study are available from the corresponding author, [J.L.T.], upon reasonable request.

Additional information

Funding

NIH funding (Project ID 1R21EB030182-01)(J.T.) and NIH NCI (R01CA269811) (I.C.A), American Kennel Club (Canine Health Foundation No. 02773)(J.T.) and Focused Ultrasound Foundation (FUSF-RAP-823R1)(J.T.) helped support the design and reporting of this study.