Low Intensity Focused Ultrasound (LOFU)-mediated Acoustic Immune Priming and Ablative Radiation Therapy for in situ Tumor Vaccines

Focal ablative therapies have been primarily used for local tumor ablation. However, they often fail to impact systemic disease. Here we propose the use of low intensity focused ultrasound (LOFU), a noninvasive, nontoxic, conformal therapy, to deliver acoustic stress to the tumor for immune priming. We demonstrate that LOFU significantly induces expression and cell surface localization of heat shock proteins in murine breast (4T1) and prostate adenocarcinoma (TPSA23) cancer cell lines. In vivo LOFU followed by ablative radiation therapy (RT) results in primary tumor cure, upregulation of a cytotoxic immune response and induction of immunological memory by inhibiting secondary tumor growth upon re-challenge with tumor cells. We, therefore, describe a regimen of a combination therapy with noninvasive, acoustic immune priming and ablative radiation therapy to generate an in situ tumor vaccine, induce CD8+ T cells against tumor-associated antigens and provide a viable oncologic treatment option for solid tumors.


LofU increases the expression and cell surface localization of heat shock proteins (HSp).
We analyzed the expression of HSP mRNA and protein localization in LOFU-treated, mouse breast and prostate cancer cell lines, 4T1 and TPSA23, respectively. We first determined the effects of varying low intensities (I SATP < 800 W/cm 2 ) of ultrasound on Hsp gene expression in 4T1 cells, a mouse model of triple negative breast cancer. Quantitative RT-PCR (qRT-PCR) analysis using primers for Hsp gene families showed that there were significant increases in mRNA levels across all family members with Hsp70 and Hsp90aa1 RNA displaying the highest expression (13-16 fold over non-treated), when normalized to Gapdh RNA expression, with increasing intensity of LOFU, four hours after treatment (Fig. 1A). To examine whether LOFU treatment increased cytoplasmic HSP70 protein levels, we performed HSP70 ELISA of cell lysates. There was a significant increase from 93.13 ± 27.8 to 255.3 ± 28 pg of cytosolic HSP70 per mg of total protein, four hours after LOFU treatment (Fig. 1B). Since the cell membrane is the first to encounter ultrasound pulses, we therefore examined cell surface localization of HSP70 and HSP90 on 4T1 by flow cytometry as a measure of acoustic stress. The translocation of cytoplasmic HSPs to the cell surface also provides an activation signal for natural killer cells and danger signals for DC activation 14,15 . Cell surface HSP70 increased after treatment with 5 W, 50% duty cycle (7.3% of cells having surface HSP70 compared to 4.8% in non-treated). For HSP90, the surface localization also peaked with 5W, 50% duty cycle (19.2% versus 9.3% non-treated) before reaching a plateau with higher intensity treatments (22.5% and 23.2% with 7W, 50% and 9W, 50% respectively) (Fig. 1C). Lastly, we measured the secretion of HSP70 in the culture supernatant of 4T1 cells by ELISA 4 hours and 24 hours after LOFU treatment. Four hours after LOFU, there was no evidence of HSP70 or HSP90 secretion. However, 24 hours after treatment, there was an increase in HSP70 secretion by LOFU-treated cells, compared to untreated cells (2.5 ng/mL versus 0.476 ng/mL, respectively) ( Fig. 1D).
TPSA23 is a TRAMPC1-derived tumor cell line, which expresses human prostate specific antigen (PSA) 16 . In TPSA23 cells, Hsp RNA expression was increased across all families of Hsp genes ( Fig. 2A) within 4 hours of LOFU. Hspa1b (HSP70) mRNA showed the greatest increase (624.5 ± 121.3-folds) over non-treated group (p = 0.02). Hspa1a (HSP72) closely followed with a 458.3 ± 152-fold change (p = 0.03) in RNA expression. LOFU treatment of TPSA23 cells significantly increased the expression of mRNAs of Hspb1 (HSP27) by 222.8 ± 92.8-folds (p = 0.03), Hsp90aa1 (HSP90α) by 28.9 ± 5.1-folds (p = 0.04), and Hsph1 (HSP110) by 4.9 ± 1.3-folds (p = 0.007) over non-treated cells, four hours after LOFU treatment. The gene expression changes also translated into a 2.5-fold increase in HSP70 protein expression at 24 hours, as demonstrated by an ELISA of HSP70/HSPA1A (Fig. 2B). LOFU alone also increased the HSP70 concentration in lysates of in vitro treated TPSA23 cells, 4 hours, 8 hours and 24 hours after treatment (Fig. 2B). We, then analyzed the surface localization of HSP60, 70 and 90, four hours after LOFU treatment of cells, using flow cytometry, (Fig. 2C). Surface HSP90 increased the most after LOFU treatment (15.2% ± 8.9, untreated to 56.7% ± 1.2, LOFU-treated), followed by HSP70 (7.4% ± 5, untreated to 38.7% ± 1.1, LOFU-treated), and HSP60 (0.45% ± 0.19, untreated to 22.4% ± 4, LOFU-treated). Together these results indicate that a short pulse (1.5 sec) of LOFU induced a substantial acoustic stress response with modulation of HSP RNA expression and protein localization in 4T1 and TPSA23 tumor cells. The cell surface translocation of HSPs are more pronounced in TPSA23 cells, compared to 4T1 cells indicating that differences in the intrinsic biology of tumor types may determine the extent of cell surface translocation of HSPs. The cell surface HSPs post-LOFU treatment could provide danger signals for DC activation and induction of tumor-specific immune response 15 . LOFU immune priming of radiation therapy (RT) for TPSA23 tumors. As demonstrated in Figs 1 and 2, both 4T1 and TPSA23 cell lines showed increased expression of HSPs at the mRNA and protein levels after LOFU treatment. However, TPSA23 showed a more robust response, hence we chose TPSA23 as a tumor model for further in vivo studies. We selected a LOFU treatment of 1 MHz frequency, 100% duty factor, 5 W and 1.5 second treatment time per focal spot for optimal immunomodulation without increasing the cytotoxicity at higher LOFU intensities. Therapeutic ultrasound has also been shown to inhibit phosphorylation and activation of www.nature.com/scientificreports www.nature.com/scientificreports/ STAT3 in prostate cancer cells, transiently for 6 hours [17][18][19] . Since phopho-STAT3 can induce radio-resistance 20-23 , we chose to deliver RT fractions, 2-4 hours after LOFU treatment of TPSA23 tumors. Thus, we hypothesized that LOFU will transiently make TPSA23 cells radiosensitive and induce a heat shock response in TPSA23 tumors with an increase in the expression, cell surface localization and secretion of HSP proteins. Ablative RT after LOFU treatment would increase the release of tumor derived peptide-HSP complexes for cross-presentation by antigen presenting cells, thereby, increasing the immunogenicity of tumor cells and tumor growth retardation. In order to study the effect of combining LOFU treatment to ablative RT, we first determined whether LOFU inhibits the clonogenic capacity of irradiated TPS23 cells. Figure 3A shows the results of a clonogenic assay after LOFU alone, RT (2 Gy), and LOFU + RT. Combination therapy resulted in a reduction in the surviving fraction of cells, when compared to each individual treatment and radiation alone, indicating that LOFU enhances the tumoricidal effects of RT by multiple mechanisms, including inhibition of STAT3 activation.
To confirm our in vitro results of an increased cell surface HSPs after LOFU treatment, we further evaluated the cell surface expression of HSP60 and 90 by flow cytometry in LOFU-treated TPSA23 tumors in vivo (Fig. 3B). LOFU alone and LOFU + RT treatment showed a higher percentage of HSP60 and HSP90 expressing tumor cells, www.nature.com/scientificreports www.nature.com/scientificreports/ consistent with the in vitro results. Finally, we tested whether LOFU increases the secretion of HSPs from irradiated tumor cells. Figure 3C shows that only after combination treatment of LOFU + RT was there an increase in the plasma concentration of HSP90α (1.285 ± 0.652 pg/mL) when compared to monotherapy (LOFU alone: 0.386 ± 0.309 pg/mL; RT alone: 0.466 ± 0.254 pg/mL) or nontreated (0.628 ± 0.224 pg/mL) control groups. These results raise the possibility that acoustic immune priming by LOFU could increase the immunogenic potential of RT by providing HSP-chaperoned peptide antigens and danger signals released from dying irradiated tumor cells, to activate both the innate and adaptive immune response against tumors.
LofU + Rt cures primary prostate tumor in a t cell dependent manner. In order to study whether LOFU-induced immune priming followed by ablative RT can induce anti-tumoral immunity and improve tumor control, we used a murine model of prostate cancer, TPSA23, grown either in wild-type male C57BL/6 mice, prostate specific antigen (PSA) transgenic mice 24 , or in immunodeficient athymic nude mice. We used PSA as a tumor antigen because PSA vaccines have been designed for immunotherapy for prostate cancer 25 . Since PSA is a foreign protein in C57BL/6 mice, we are also studying the effects of LOFU in PSA transgenic mice to determine whether LOFU can induce an immune response in animals tolerant to PSA as a self-antigen. Tumor growth in wild-type male C57BL/6 mice was significantly inhibited in the LOFU + RT combination therapy group, with 46% of mice (n = 28) never reaching 5 times the tumor volume at the start of treatment (V 0 ) ( Fig. 4B-top, C-left). Tumor volumes of wild-type mice in the RT alone and LOFU + RT combination group were significantly reduced compared to the non-treated group at day 47 post-inoculation, the latest time point at which all mice were still alive (p = 0.03, and 0.0007 respectively, Dunn's multiple comparisons test). Additionally, with the combination of www.nature.com/scientificreports www.nature.com/scientificreports/ LOFU and RT resulted in a significant number of primary tumor cures (7 out of 11 mice), whereas RT alone did not cure any mice (Fisher's exact test, p = 0.0039). These results suggest that ultrasound treatment prior to RT is required for successful local tumor control.
To determine if our LOFU + RT combination therapy could overcome tolerance to self-antigen, we used the TPSA23 cell line in transgenic mice expressing human PSA in the prostate 24 . Our results demonstrated that combination therapy of LOFU and radiation was able to overcome or bypass tolerance and successfully cure the primary tumor in over 50% of the mice treated ( Fig. 4B-middle, C-center). When compared to RT alone, LOFU + RT achieved more complete response of tumors (0/17 cures in RT versus 9/16 cures in LOFU + RT groups, Fisher's exact test p = 0.0003). On day 35 post tumor inoculation, only LOFU + RT and no LOFU or RT alone had tumor growth retardation compared to non-treated (p = 0.001 by Dunn's multiple comparisons test). To determine whether the efficacy of LOFU + RT required T lymphocytes, identical experiments were conducted in athymic nude mice. The combination treatment of LOFU + RT demonstrated no enhancement in the tumor growth retardation over radiation alone (194.1 ± 106.7 mm 3 and 104.4 ± 36.6 mm 3 , respectively, p = 0.13 Mann-Whitney U on day 41) indicating involvement of T lymphocyte mediated antitumor immune response elicited by LOFU + RT combination (Fig. 4B-bottom, C-right).
LofU + Rt augments anti-tumoral cytotoxic t cell immunity. To determine the impact of combination therapy on systemic CD8+ T cell responses, we analyzed tumor antigen specific T cells in splenocytes 7 days after the completion of primary tumor treatments, using a PSA-specific MHC Class I pentamer. There was an increase in the PSA-specific, activated (CD62L-) CD8+ T cells across all the treatment groups in WT mice, with LOFU + RT having the highest percentage of CD62L-/pentamer + CD8 T cells (1.557 ± 0.127%) compared to LOFU alone (0.497 ± 0.064%) and RT alone (0.923 ± 0.387%) groups (Fig. 5A). Similar experiments were performed in PSA-Tg mice, which did not show a significant increase in PSA-specific CD8+ T cells with LOFU + RT (data not shown). These results indicate that the tolerance to PSA was most likely circumvented as indicated by the LDH release to tumor lysate. An LDH release assay indicative of immune-mediate cytotoxicity also confirmed that splenocytes from PSA-transgenic mice treated with the combination therapy were able to kill more naïve tumor cells (NT: 4.186 ± 3.1%, LOFU: 5.1 ± 4.3%, RT: 2.91 ± 1.7%, LOFU + RT: 11.613 ± 3.98%; Fig. 5B). Additionally, reduction of TIM3 expression on the surface of activated (CD62L-CD69-) and effector memory  www.nature.com/scientificreports www.nature.com/scientificreports/ complete tumor growth inhibition (Fig. 6C). Re-challenge response in WT mice (Fig. 6B) was different, at day 43 post-inoculation, only 1 out of 4 showed tumor growth inhibition and persisted in its tumor inhibition response until the end of the experiment.
At day 56, the WT mice cured of primary tumor by LOFU + RT combination had a tumor growth delay of 52% and the PSA-transgenic mice cured of primary tumor by LOFU + RT combination had a tumor growth delay of 74% compared to naïve mice (Fig. 6D) indicative of immune memory generated by LOFU + RT treatment of the previously cured tumor. While both C57BL/6 (WT) and PSA-transgenic mice (C57BL/6 background) primary tumor-cured groups showed tumor growth retardation after rechallenge, only PSA-transgenic mice showed statistically significantly reduction in the tumor volume (p = 0.008, Dunn's multiple comparisons) compared to the naïve group. While there was no statistically significant difference between the rechallenge tumor volume in wild-type rechallenged mice and naïve or PSA-transgenic mice, 1 out of 4 mice demonstrated complete immunological memory, while 3 out of 4 mice had a growth delay compared to non-treated. These experiments indicate that combination treatment of LOFU + RT not only resulted in primary tumor cure but also enhanced immunological memory, which resulted in significant inhibition of the secondary tumor growth.

Discussion
This study highlights the significance of physical energy-based immune priming in combination with ablation for optimal induction of anti-tumoral immunity. Focal ablative therapies have been used primarily for local tumor ablation. However, they often fail to impact systemic disease. Our goal is to design focal therapies to modulate the tumor microenvironment to drive systemic anti-tumoral immunity. In this study, we demonstrate that LOFU-mediated immune priming stimulated a robust heat shock response with an increase in RNA, protein expression, and cell surface localization of HSPs in murine breast and prostate cancer cell lines. While LOFU alone was responsible for increased HSP mRNA and protein expression along with redistribution of HSPs to the cell surface, plasma HSP90α was only increased after LOFU + RT combination therapy. The addition of a non-ablative LOFU treatment enhanced the tumoricidal effects of RT with a significant increase in tumor growth inhibition and complete response in immune competent mice, due to activation of anti-tumoral immunity. The clinical significance of the prostate cancer-specific anti-tumoral immunity is highlighted by the induction of immunity and complete response in the PSA-transgenic mice, where PSA is a self-antigen. Surprisingly, LOFU enhanced the tumoricidal effects of radiation in vitro as evident by the decreased clonogenicity of the combination treatment when compared to the individual treatment. The mechanism of this enhancement is multi-faceted. www.nature.com/scientificreports www.nature.com/scientificreports/ One possible explanation is a LOFU-induced reduction in activated STAT3, which is known to mediate resistance to radiation-induced cell death. LOFU + RT induced PSA-specific activated CD62L-CD8+ T cells, reduced the levels of exhausted TIM3 + effector memory and activated T cells, and had increased tumor-specific CTL activity in splenocytes. Finally, mice that were cured after LOFU + RT exhibited immune memory and tumor growth inhibition upon re-challenge of TPSA23 cells.
The basis of LOFU-mediated immune priming lies with the heat shock response, which is an evolutionarily conserved cellular defense mechanism for promoting protein folding. Stress signals in the tumor microenvironment, including oxidative stress and heat can initiate a heat shock response with induced expression of molecular chaperones of the Hsp gene families. In both 4T1 and TPSA23, LOFU induced Hsp mRNA expression indicating that the acoustic stress pathway has a heat shock response component in tumor cells, possibly by inducing protein misfolding after LOFU treatment. An alternative stressor mechanism could be the direct effect of ultrasound on the plasma membrane. It is expected that the cell membrane would initially "sense" the ultrasound pulses. Since the thermal effects of a 1.5 second LOFU pulse is much lower than traditional hyperthermia treatment which lasts for 30-60 minutes, it is possible that LOFU could induce changes in the fluidity of membrane lipids and induce a heat shock-like response without significant protein denaturation. Mild hyperthermia has previously been shown to initiate a remodeling of lipids, redistribution of stress proteins, while maintaining membrane integrity 26 . A role for "membrane thermosensors" with a cell membrane-initiated heat shock response has been previously described in cancer cells 27 . A critical element for this process is the translocation of HSP proteins, such as, HSP70 and HSP90 on the cell surface of tumor cells 28 . Cell surface HSPs can be a target for natural killer cell-mediated cytolysis 29,30 and can increase radiation sensitivity in tumor cells 31 . Thus, LOFU treatment can not only increase the tumoricidal effects of radiation but also can be viewed as an immune priming therapy that re-engineers the tumor microenvironment to drive systemic immunity.
Another characteristic of LOFU-mediated immune priming is the release of HSP-peptide complexes after tumor ablation. Cytosolic proteins are released into the extracellular compartment with eventual increase in plasma levels of HSPs after focal ablation of tumor. Tumor-infiltrating phagocytes, including DCs, can engulf these extracellular antigens and typically process and present antigenic peptides via MHC class II for CD4 + T helper cell activation through the endosomal pathway. For successful activation of CD8 + cytotoxic T lymphocytes (CTL), engulfed extracellular antigens need to transfer from the endosomes to the cytoplasm for proteasomal degradation for eventual cross-presentation to class I MHC via the endoplasmic reticulum antigen presentation pathway. Upon engulfment of extracellular antigens by DCs, HSP90 has been shown to promote the translocation of endosomal proteins into the cytosol, thus contributing to cross-presentation of extracellular antigens for CTL activation 32,33 . Therefore, it is possible that the increased release of HSP90-peptide complexes from LOFU + RT-treated tumor cells induced tumor-specific CD8 + CTL activation seen in our studies (Fig. 5C). While we have focused on extracellular HSP90 in our studies, tumor-derived HSP70-peptide complexes can also be cross-presented by human DCs 34 . www.nature.com/scientificreports www.nature.com/scientificreports/ In summary, we describe a novel paradigm of immune priming ablation therapy using LOFU in combination with ablative RT for generating in situ tumor vaccine. With minimal side effects, non-invasive and non-ionizing LOFU treatment can be administered repeatedly with any ablative therapy, such as, RT, hormonal ablation, chemotherapy or immunotherapy during the course of prostate cancer therapy with the goal of inducing a systemic anti-tumoral immunity, thereby converting focal tumor ablation into systemic cure. In vitro treatment plan. Prior to in vitro treatment, cells were trypsinized with 0.05% trypsin, quantified with a hemocytometer and 2 × 10 6 cells were distributed into 0.2 mL PCR tubes with 200ul of media. Cells were pelleted in the PCR tubes, submerged under degassed water over an ultrasound absorber. Treatment was performed on Philips' Therapy and Imaging Probe System (TIPS, Philips Healthcare, Briarcliff, NY) in a grid fashion with a raster pattern to cover the entire pellet. All treatments were performed at 1 MHz frequency, 100 Hz pulse repetition frequency and a 1.5 second treatment time at each focal spot. Table 1 shows the other ultrasound parameters used. After treatment, cells were replated and incubated at 37 °C with 5% CO 2 until the desired timepoint. All cells were routinely tested for mycoplasma contamination with the MycoAlert Kit (Lonza, Walkersville, MD).

Methods
RnA isolation, qRt-pcR. For RNA isolation, at the desired time point, cells were harvested with 1 mL of TRIZol (Invitrogen, Carlsbad, CA) and stored at −20 °C until isolation. RNA isolation was performed according to manufacturer's instructions (Invitrogen, Carlsbad, CA). cDNA was generated according to manufacturer's instructions with the Verso cDNA synthesis kit (Thermo Fisher Scientific, Waltham, MA). qPCR was performed in 384-well plates with SYBR green as the marker on the Applied Biosystems 7900HT PCR System at the Genomics Core at Albert Einstein College of Medicine (Thermo Fisher Scientific, Waltham, MA). Primers (Table 2)  HSP70 and HSP90α eLiSA. Four, eight and twenty-four hours after LOFU or sham treatment cell lysates were generated using RIPA buffer (MilliporeSigma, Burlington, MA) with protease/phosphatase inhibitors and stored at −80 °C until ready for use. The cell culture supernatant was also stored at −80 °C for future use. Total protein concentration was measured using spectrophotometry (SpectraDrop, SpectraMax M3, Molecular Devices, Sunnyvale, CA). HSP70/HSPA1A ELISA was performed following manufacturer's instructions using 50 µL of lysate or cell culture supernatant.
Plasma from treated mice was isolated the day after the end of treatment and frozen at −80 °C until ready for HSP90α ELISA. Manufacturer's instructions (MyBioSource, San Diego, CA) were followed using 50 µL of plasma. clonogenic assay. After LOFU treatment or sham treatment, 500 cells were plated in 6-well plates and incubated for 2 hours. The indicated plates were irradiated at 2 Gy using a Shepard Mark I Cesium 137 irradiator and incubated for 7 days. Media was removed from the plates which were then stained with 0.1% crystal violet in www.nature.com/scientificreports www.nature.com/scientificreports/ 10% phosphate buffered formalin for 2 minutes before rinsing with water. Once the plates were dried, colonies with greater than 50 cells were counted, and the survival fraction was calculated using the plating efficiency from non-treated cells.
In Vivo model & treatment. All mouse protocols were performed in accordance with policies previously approved by the Institutional Animal Care and Use Committee of Albert Einstein College of Medicine. Wild-type C57BL/6 male mice and athymic nude (BALB/c background) male mice were purchased from Charles River Laboratories (Wilmington, MA). A breeding pair of PSA-transgenic mice (PSA-Tg), which exhibits prostate-specific expression of the entire human PSA gene, was gift from Dr. Frelinger 24 , and the colony was maintained at the Albert Einstein College of Medicine.
For tumor inoculation, TPSA23 cells were harvested, counted and resuspended in PBS at 20 × 10 6 cells/mL. Fifty microliters of the cell suspension was injected into the right flank of shaved male C57BL/6 mice, anesthetized under continuous inhaled isoflurane. When the tumors reach 4-6 mm in diameter, treatment began. The area around the tumor was epilated with Nair (Church & Dwight, Ewing Township, NJ) and tumor dimensions measured with calipers prior to ultrasound treatment. Mice were anesthetized with isoflurane for treatments. Tumors were pulled away from the mouse's body and acoustically coupled to a gel pad (Aquasonic, Parker Laboratories, Fairfield, NJ) on top of an ultrasound absorber. Ultrasound treatments were performed with the same ultrasound device as in vitro treatments at 1 MHz frequency, 100% duty cycle, 5W and 1.5 second treatment time per focal spot in a grid pattern with a maximal temperature of 45 °C achieved. Radiation treatment (10 Gy) was performed on the Small Animal Radiation Research Platform (XStrahl Medical, Suwanee, GA) 2-3 hours following ultrasound treatment due to our prior studies using this treatment scheme 13 and other studies demonstrating reduction in phosphorylated-STAT3 after focused ultrasound treatment 17,18 . Treatments were repeated after about 48 hours to allow time for normal tissue recovery. Tumors were measured 1-2 times per week and tumor volume was calculated with the following formula = × × × . V length width height 3 1456 6 35 . If used for tumor growth and survival studies, mice were euthanized when tumor volume reached 1000 mm 3 in accordance with IACUC protocols. tissue digestion. Tissues were harvested after mice were humanly euthanized. Spleens were digested by mechanical force, filtered through a 40 µm strainer and RBC were lysed with ACK lysing buffer. Tumors were chopped into small pieces, added to 100 U/mL each of collagenase I & IV (Gibco, Thermo Fisher Scientific, Waltham, MA) and Dnase I (Roche, Basel, Switzerland), heated at 37 °C, with mechanical digestion. The cell suspension was filtered through 70 µm strainer prior to staining. immune Studies. Seven days after the end of treatment, at the end of peak T cell activation range (typically 5-7 days), splenocytes were isolated from treated mice and used for several assays. Flow cytometry for PSA-specific (MHC Class I restricted pentamer, ProImmune, Oxford, UK) and exhausted T cells (TIM3+) was performed. An LDH cytotoxicity assay was performed after overnight incubation of 20,000 mitomycin C treated tumor cells with the same number of splenocyte according to manufacturer's instructions (Pierce LDH Cytotoxicity, Pierce Biotechnology, Waltham, MA).