Delivery of oncolytic vaccinia virus by matched allogeneic stem cells overcomes critical innate and adaptive immune barriers
Previous studies have identified IFNγ as an important early barrier to oncolytic viruses including vaccinia. The existing innate and adaptive immune barriers restricting oncolytic virotherapy, however, can be overcome using autologous or allogeneic mesenchymal stem cells as carrier cells with unique immunosuppressive properties.
To test the ability of mesenchymal stem cells to overcome innate and adaptive immune barriers and to successfully deliver oncolytic vaccinia virus to tumor cells, we performed flow cytometry and virus plaque assay analysis of ex vivo co-cultures of stem cells infected with vaccinia virus in the presence of peripheral blood mononuclear cells from healthy donors. Comparative analysis was performed to establish statistically significant correlations and to evaluate the effect of stem cells on the activity of key immune cell populations.
Here, we demonstrate that adipose-derived stem cells (ADSCs) have the potential to eradicate resistant tumor cells through a combination of potent virus amplification and sensitization of the tumor cells to virus infection. Moreover, the ADSCs demonstrate ability to function as a virus-amplifying Trojan horse in the presence of both autologous and allogeneic human PBMCs, which can be linked to the intrinsic immunosuppressive properties of stem cells and their unique potential to overcome innate and adaptive immune barriers. The clinical application of ready-to-use ex vivo expanded allogeneic stem cell lines, however, appears significantly restricted by patient-specific allogeneic differences associated with the induction of potent anti-stem cell cytotoxic and IFNγ responses. These allogeneic responses originate from both innate (NK)- and adaptive (T)- immune cells and might compromise therapeutic efficacy through direct elimination of the stem cells or the induction of an anti-viral state, which can block the potential of the Trojan horse to amplify and deliver vaccinia virus to the tumor.
Overall, our findings and data indicate the feasibility to establish simple and informative assays that capture critically important patient-specific differences in the immune responses to the virus and stem cells, which allows for proper patient-stem cell matching and enables the effective use of off-the-shelf allogeneic cell-based delivery platforms, thus providing a more practical and commercially viable alternative to the autologous stem cell approach.
KeywordsVaccinia Cancer Stem Cells Oncolysis Oncolytic virus Virotherapy Immunity Immunotherapy
adipose-derived stem cell
killer-cell immunoglobulin-like receptor
major histocompatibility complex
mesenchymal stem cell
peripheral blood mononuclear cells
stromal vascular fraction
Successful virotherapy of cancer is critically dependent on the ability of oncolytic viruses like vaccinia to overcome multiple defense barriers including the complement/antibody-mediated neutralization , the interferon-induced anti-viral state [2, 3, 4, 5], as well as the innate and adaptive anti-viral immune mechanisms mediated by NK and T cells, respectively. Mesenchymal stem cells (MSC) represent a promising delivery vehicle that protects vaccinia virus from the effects of complement/neutralizing antibodies , while also having the unique ability to home to sites of inflammation and tumor growth . This therapeutic platform also takes advantage of the immunosuppressive properties of MSC [8, 9, 10], in particular, their ability to survive undetected in allogeneic settings and to transiently counteract the innate and adaptive anti-viral immunity , thus enabling rapid virus spread and colonization of the tumor [12, 13].
We therefore investigated the potential and feasibility of using ex vivo expanded mesenchymal stem cells as a delivery vehicle for oncolytic vaccinia virus. Mesenchymal stem cells can be expanded from various sources including adult bone marrow , adipose tissue , blood, and dental pulp or neonatal umbilical cord, placenta etc. [16, 17, 18, 19], and are known to possess potent immunosuppressive properties mediated by IDO [20, 21, 22], PGE [23, 24, 25], Adenosine , TGFβ [10, 27, 28], VEGF [29, 30], HGF [31, 32, 33, 34], iNOS [35, 36, 37, 38], IL-10 , HLA-G5 , and Galectins , regardless of the source of origin [16, 29, 41]. Adipose-derived stem cells (ADSCs) isolated from the stromal vascular fraction (SVF) of liposuction aspirates are particularly useful because they are easily available and can be efficiently expanded in culture for different applications [42, 43, 44, 45]. Our studies indicate that these carrier cells possess impressive virus amplification potential and the capacity to immunosuppress both the innate and adaptive arms of anti-viral immunity.
The immunosuppressive properties of mesenchymal stem cells have been investigated extensively and supported by clinical data demonstrating their ability to avoid immune recognition and survive even in MHC-mismatched allogeneic recipients [11, 46, 47, 48]. What remains still unclear, however, is how the immunosuppressive properties of MSCs will be affected by the progression of vaccinia virus infection especially in allogeneic MHC-mismatched settings. Importantly, MSC-mediated immunosuppression is subject to differential modulation by both type I [36, 49] and II interferons [22, 50, 51, 52, 53, 54, 55], which are similarly involved in the control of vaccinia virus infection [2, 3, 56, 57]. While type I interferons are secreted by any cell type in response to viral infection, including the MSC, Interferon gamma (IFNγ) is a type II interferon that is produced by a limited number of immune cells like NK, NKT and T cells in response to virus infection  or, as we demonstrate, the carrier stem cells, particularly in some unfavorable allogeneic MHC-mismatch settings. These conflicting roles of IFNγ to limit virus spread through the induction of an anti-viral state but at the same time to stimulate MSC-mediated immunosuppression counteracting anti-viral immunity might prove to be a critically important determinant of therapeutic success.
We hypothesized that the complex interplay between vaccinia virus, ADSCs and immune cells could have significant impact on the ability of stem cells to function as a Trojan horse that can amplify and deliver oncolytic vaccinia virus. Here we show that interferons protect stem cells against vaccinia virus infection but compromise their function as a Trojan horse. The IFNγ responses to the virus and allogeneic stem cells alone or in combination, however, appear to be highly variable and patient-specific. Our studies indicate that these differences can be associated with subtle allogeneic NK- and T cell-mediated cytotoxic and cytokine responses that can result in the inactivation or complete rejection of the Trojan horse. These findings have significant implications for the development of cell-based delivery platforms for oncolytic viruses and suggest the need for proper screening and patient-specific matching to enable the successful use of off-the-shelf allogeneic cell carriers, as opposed to the more expensive personalized autologous stem cell approach.
Materials and methods
Cell lines, cytokines, and viruses
B16 F10 melanoma, A549 lung carcinoma, and CV-1 monkey kidney cells were obtained from Dr. Boris Minev, and K562 cells were a kind gift from Albert Perez-Ladaga, PhD. Cells were propagated in DMEM (B16, A549) from Gibco (Cat#: 11960069) or RPMI 1640 (K562) from Gibco (Cat: 21870092) supplemented with 10% Fetal Bovine Serum (Omega Scientific, FB-02, USDA certified, heat inactivated), 2 mM l-Glutamine (ThermoFisher Scientific, 25030081, 100×) and Penicillin/Streptomycin (Life Technologies, 15140122, 100×). Human IFNγ (Peprotech, cat# AF3000220UG, 20 mg lyophilized, diluted to 20 μg/ml approx. 1000× stock in 1×PBS supplemented with 0.1% FBS, stored at − 80 °C) and IFNβ (Peprotech, cat# AF30002B5UG, 5 μg lyophilized, diluted to 5 μg/ml approx. 1000× stock in 1×PBS supplemented with 0.1% FBS, stored at − 80 °C) were added to ADSCs for 24 h 1 to 3 days prior to virus infection. Sucrose gradient purified WT1/ACAM2000 and L14 (TK-inserted Turbo-FP635 engineered LIVP strain) vaccinia viruses were obtained from StemVac GmbH, Bernried, Germany. In some experiments the virus and ADSCs were pre-infected for 1 h with constant agitation on an orbital shaker at 37 °C (incubator) before adding them to PBMCs or tumor cell co-cultures.
Adipose-derived stem cells isolation and culture
Non-cancer donor SVF and PBMC were obtained as part of an IRB-approved protocol after informed written consent (International Cell Surgical Society; IRB# ICSS-2016-024). Fresh SVF fractions were plated to attach overnight and next day were washed to remove unattached cells and debris. Media was changed every 3–4 days until the mesenchymal stem cells start to grow and reach 80% confluency. Cell were expanded to 80% confluency and passaged every 3–4 days using TrypLE™ Express (Life Technologies, (1×), no phenol red, Cat# 12604021, 3 min 37C incubator) for up to 10 passages. Note that P12 has normal mesenchymal look and morphology but manifested evidence of gradual loss of immunosuppressive ability. ADSCs were expanded and maintained in 5% Human Platelet Extract (Cook Regentec, Stemmulate, PL-SP-100) in DMEM supplemented with l-Glutamine and Pen/Strep.
Generation of adipose-derived MSC constitutively expressing eGFP
RM20 adipose-derived stem cells at passage 0 were engineered to express eGFP under the control of the CMV promoter. A Lentiviral vector (VectorBuilder) containing eGFP was used to introduce eGFP for constitutive expression. 10,000 eGFP-positive cells were sorted at passage 1 and subsequently at passage 2 using the BioRAD S3 Cell Sorter. eGFP expression was confirmed by flow cytometry and fluorescence microscopy using the Keyence All-in-one Fluorescence Microscope BZ-X700 Series.
PBMC were isolated through standard Ficoll protocol (Ficoll-Paque Plus, GE Healthcare, cat# 95021-205) and co-cultured (100 μl) with ADSC (50 μl) plus minus vaccinia virus (50 μl) for 48 h on 96-well flat-bottom plates and in a total of 200 μl R10 medium (RPMI 1640 supplemented with 10% FBS, l-Glutamine and Pen/Strep). In some experiments the virus (50 μl) and stem cells (50 μl) were premixed and agitated on an orbital shaker at 37C (incubator) for 1 h, and then (100 μl of the mix) was added to the PBMCs without additional washing of any unbound virus. At the end of the 48 h incubation period the cells were recovered for staining and flow analysis directly or after an additional 4–5 h stimulation with K562 cells or PMA/Ionomycin (50 μl) with Monensin/Brefeldin A (50 μl) as needed.
Flow cytometry analysis
Co-cultures of PBMC and stem cells were recovered by pipetting and transferred to V-bottom plates, where they were washed with FACS Buffer (1×PBS with 1% FBS) and surface stained for 30 min at 4C in FACS Buffer supplemented with the following antibody cocktail: anti-human CD3-PerCP/Cy5.5 (BioLegend, cat# 300328, at 1:50), anti-human CD335 or NKp46-PE (BioLegend, cat# 331908, at 1:50), anti-human CD69-APC (BioLegend, cat# 310910, at 1:50). The FACS buffer also contained a viability probe (ThermoFisher Scientific, LIVE/DEAD Fixable Violet Dead Cell Stain Kit, for 405 nm excitation, cat# L34964, at 1:1000). After staining the cells were washed twice with FACS Buffer, fixed in 2% PFA in 1×PBS for 15 min at RT, washed again with FACS Buffer to remove PFA and analyzed on BD FACSAria II. To evaluate cytotoxic functions in some experiments anti-human CD107a-AlexaFluor 488 (BioLegend, cat# 328610) was added directly to the co-cultures at 1:20 (10 μl/well) 5 h prior to recovery and surface staining followed by addition of Monensin at 1:1000 an hour later for additional 4 h incubation at 37 °C (BioLegend, cat# 420701-BL, 1000×).
To evaluate IFNγ production in activated NK, NKT and T cells, in some experiments Brefeldin A was added at 1:1000 an hour after stimulation or 4 h prior to recovery and surface staining (BioLegend, cat# 420601-BL, 1000×). Monensin and Brefeldin A were added together when cells were to be evaluated for both IFNγ production and CD107a surface exposure. Following standard surface and viability staining cells were processed using the eBioscience Intracellular Staining Buffer Set (ThermoFisher, cat# 00-5523). Briefly, following surface staining with or without anti-CD107a-AlexaFluor 488, cells were fixed for 30 min with 1 part Fixation/Permeabilization Concentrate (cat# 00-5123) and 3 parts of Fixation/Permeabilization Diluent (cat# 00-5223), washed twice with 200 μl/well Permeabilization Buffer 10× (cat# 00-8333), diluted 1:10 in double distilled water), and stained with anti-human IFNγ-APC antibody (BioLegend, cat# 502512, at 1:50) in Permeabilization Buffer for 1 h at RT. Cells were washed twice in Permeabilization Buffer, fixed in 2% PFA in 1xPBS for 15 min at RT, washed again with FACS Buffer to remove PFA and analyzed on BD FACSAria II.
NK cell immunosuppression
To evaluate the extent of ADSC-induced immunosuppression following the 48 h co-culture of PBMC with autologous or allogeneic ADSCs in the presence or absence of virus, the 250,000 PBMCs co-cultures were subjected to an additional brief 4 h stimulation at 37 °C (incubator) with 250,000 K562 (physiological stimulation of NK cells) or non-physiological stimulation with PMA/Ionomycin to assess the extent of NK cell viability/irreversible suppression. PMA (25 ng/ml final, Sigma, cat# P8139-5MG, diluted in DMSO to 5 mg/ml stock) and Ionomycin (1 μg/ml final, Sigma, cat# I-0634-1MG, diluted in DMSO to 1 mg/ml stock) were added as 4× solutions in medium. To evaluate cytotoxic activity of NK cells, the PBMC co-cultures were also stimulated in the presence of anti-human CD107a antibody in combination with Monensin/Brefeldin A as described above. Note that CD69+ surface expression was evaluated in duplicate wells in the absence of Monensin/Brefeldin A treatment due to severe interference.
Virus containing samples were stored at − 80 °C and subjected to a three-fold freeze (− 80 °C)/thaw (+ 37 C) cycle followed by sonication on ice-cold water for three 1 min intervals one min apart. Sonicated samples were serially diluted in vaccinia virus infection medium (DMEM supplemented with 2% FBS, l-Glutamine, Penicillin/Streptomycin). Plaque assays were performed in 24-well plates in duplicate wells. Briefly 200,000 CV1 cells were plated in 1 ml D10 medium per well overnight. Supernatants were aspirated and tenfold serial dilutions of the virus containing sample were applied to the CV-1 monolayer at 200 μl/well. Plates were incubated for 1 h at 37C (incubator) with manual shaking every 20 min. 1 ml CMC medium was layered gently on top of the cells and plates were incubated for 48 h. Plaques were counted after fixing the cells by toping the wells with Crystal Violet solution (1.3% Crystal violet (Sigma-Aldrich, C6158), 5% Ethanol (Pure Ethanol, Molecular Biology Grade, VWR, 71006-012), 30% Formaldehyde (37% v/v formaldehyde, Fisher, cat # F79-9), and double distilled water) for 3–5 h at room temperature, followed by washing the plates in tap water and drying overnight. CMC overlay medium was prepared by autoclaving 15 g Carboxymethylcellulose sodium salt (Sigma-Aldrich, C4888) and re-suspending with overnight stirring at RT in 1 L DMEM, supplemented with Pen/Strep, l-Glutamine, and 5% FBS, short-term storage at 4 °C.
MTT viability assay
MTT assays were performed as previously described. Briefly, MTT (ThermoFisher, cat# M-6494, 5 mg/mL stock in 1×PBS, kept at− 20 °C) was added to cells (10 μl to 100 μl cells/well) on 96-well flat-bottom plates at a final concentration of 5 μg/mL and incubated for 1–2 h at 37 °C (incubator). Following incubation, cells were lysed by adding 100 μl of Isopropanol: 1 M HCl (24:1, supplemented with 10% Triton ×100, Sigma-Aldrich, ×100-100ML) and vigorous pipetting to dissolve the Formazan. Plates were read on Tecan InfiniteR 200 Pro and the MTT signal was measured within 1 h by subtracting OD at 650 nm from OD at 570 nm. Cells without MTT or Blank/Medium Alone wells were included as controls to eliminate background signals.
Time course microscopic observations of virus infection were done on a Keyence All-in-one Fluorescence Microscope BZ-X700 Series. ADSCs were engineered to express eGFP and were followed on the GFP channel (1 s exposure) while virus infection with the TurboFP635-engineered L14 virus was monitored on the TRITC channel (3 s exposure). Images at 4× or 10× magnification were collected and overlaid with bright field (phase contrast, 1/50 s exposure).
HLA and KIR analysis
HLA and KIR/MIC typing analysis was done through NGS by ProImmune (Oxford, UK) and Scisco Genetics (Seattle, USA), respectively. The presence/absence of the known KIR ligands A3/A11 (HLA-A), Bw4 (HLA-B) and C1/C2 (HLA-C) epitopes in the HLA alleles of our PBMC and ADSC donors was taken from http://www.dorak.info/mhc/nkcell.html. The − 21 M/T (Methionine/Threonine) dimorphism at the anchor amino acid from the leader sequence that predicts strong/weak binding and presentation of HLA-B-derived leader peptides by HLA-E, which provides inhibitory signaling though the NKG2A/CD94 receptors on NK cells, was taken from the Immuno Polymorphism Database (IPD) at http://www.ebi.ac.uk/ipd/.
Data analysis and statistics
Data was plotted and analyzed for statistical significance using licensed Graph Prism software. Statistical significance was evaluated with the two-tailed Student’s T-test, p < 0.05, and statistically significant correlation was shown with the Pearson coefficient and corresponding p values.
Adipose-derived stem cells provide potent amplification of vaccinia virus that can be restricted by the induction of IFN-mediated anti-viral state
ADSCs promote the oncolysis of resistant tumor cell lines through a combination of virus amplification, tumor cell recruitment and secretion of factors sensitizing the resistant tumor cells to virus infection
ADSCs can be immunosuppressive towards NK cells in both autologous and allogeneic settings
ADSCs can overcome allogeneic immune barriers and amplify VV even in the presence of some allogeneic PBMCs
We next evaluated the potential of ADSCs not only to immunosuppress but also to amplify vaccinia virus when co-cultured with PBMC from other individuals and against possible allogeneic immune barriers. We hypothesized that an unfavorable allogeneic PBMC environment might result in elimination or inactivation of the stem cells before they have a chance to amplify vaccinia virus. Surprisingly, the presence of allogeneic PBMCs didn’t interfere with virus amplification, suggesting that at least in some “permissive” cases the allogeneic stem cells remain “under the radar” and avoid immune recognition even under conditions of highly inflammatory virus infection (Fig. 3c). Moreover, in the presence of the allogeneic PBMCs the overall virus output was increased relative to the combined output of stem cells and PBMCs when infected in separation, suggesting that the stem cells can sensitize some subpopulations of normal PBMC to vaccinia virus infection in a way equivalent to their effect on tumor cells (Fig. 2b, c). These observations raised the relevant question as to whether successful and unrestrained virus amplification by the Trojan horse could be attributed to potent stem cell-mediated inhibition of anti-viral responses irrespective of any allogeneic differences or was it patient/MSC recipient-specific and limited to cases of a particularly close MHC match, hence not necessarily applicable to a larger cohort of MHC-diverse recipients.
The potential of allogeneic stem cells to overcome immune barriers correlates with their ability to suppress virus-induced T, NK and NKT cell responses
The potential of allogeneic stem cells to function as a Trojan horse is restricted by patient-specific differences suggesting that proper matching will be required
Patient-specific immunological barriers regulate the survival of the Trojan horse and can limit the therapeutic potential of both genetically attenuated and wild type vaccinia virus strains
Patients’ resistance to the Trojan horse is associated with possible HLA mismatches and the rapid induction of anti-stem cell cytotoxic and interferon responses
The PBMCs from the highly resistant SIBD01 donor also manifested a much higher frequency of CD3+ NKp46 + NKT-like cells relative to other blood donors tested, which could suggest an alternative explanation for the broader resistance to various allogeneic stem cell lines (Fig. 7d, Additional file 7: Figure S7F). The latter, however, were not found to be associated with significant IFNγ production or cytotoxic activity as measured by CD107a surface exposure at the 48 h coculture timepoint (Additional file 7: Figure S7B–D), but the potential involvement of NKT cells in the anti-viral response or the crosstalk between NK and T cells at earlier timepoints cannot be excluded. We specifically looked at anti-virus and anti-stem cell induced immune responses at earlier timepoints (Additional file 7: Figure S7G). Significant background immune cell activation was evident at 6 h that subsided by the 24 h timepoint. At 6 h only T and NKT cells from the resistant donor demonstrated stem cell-induced CD69 upregulation but not IFNγ responses. At 24 h all three subsets responded to the allogeneic stem cells by upregulation of the CD69 activation marker regardless of the virus. While the NK cell response was the most vigorous, only the NKT subset manifested statistically significant stem cell-induced IFNγ responses, suggesting that NKT cells might be the first immune cell subtype mounting effector cytokine responses against the allogeneic stem cells (Additional file 7: Figure S7G). The NK, NKT and T cells from the permissive donor didn’t upregulate CD69 or produced IFNγ in response to the stem cells at any timepoint. Of note, at the 24 h timepoint the allogeneic stem cells were even able to suppress IFNγ production by the permissive donor NK cells below background (Additional file 7: Figure S7G).
To study the potential detrimental effects of IFNγ secretion on the interactions between the immune cells and the allogeneic stem cells that are likely to occur in vivo, we performed experiments where the stem cells were pretreated with IFNγ prior to coculture with the PBMCs in the presence or absence of vaccinia virus (Additional file 7: Figure S7H–J). Contrary to our expectations that IFNγ-pretreatment would suppress immune responses due to inhibited virus amplification and improved immunosuppressive potential, as previously reported in the literature , the immune responses against the infected stem cells were further increased, specifically in the permissive donors suggesting that virus-associated infection and IFNγ-production might have the unfortunate effect of sensitizing immune cells to otherwise small and “under the radar” allogeneic differences resulting in compromised rather than improved immunosuppression of T and NK cells. Consistent with previous experiments (Figs. 4b, 5b), we also observed that the NK-specific stimulator cell line K562 induced indirect activation of NKT and T cells (Additional file 7: Figure S7J), providing further support to the notion of intensive NK-NKT-T cell crosstalk and interdependence. Such a crosstalk might also be critically important for the magnitude and kinetics of anti-viral or anti-allogeneic responses, thus playing a key role in the patient-specific restrictions to the Trojan horse.
Thus, while HLA or KIR matching might be informative and indeed play a significant role in determining the permissiveness versus resistance, the ultimate therapeutic efficacy of an allogeneic Trojan horse can be influenced by multiple other patient-specific differences including innate and adaptive immune cell composition as well as activation status or sensitivity to vaccinia virus/allogeneic mismatch. Such complex patient-specific differences would be much more properly evaluated by the development and application of robust companion diagnostic assays.
The goal of the work presented in the current manuscript was to evaluate the potential of using off-the-shelf allogeneic adipose-derived stem cells as a delivery vehicle to potentiate oncolytic virotherapy, leveraging the unique abilities of mesenchymal stem cells to amplify vaccinia virus and to transiently suppress anti-viral immunity. The immune system is known to play a dual role in oncolytic virotherapy of cancer. While the induction or potentiation of tumor-specific immunity is believed to play an important role in the long-term therapeutic efficacy of oncolytic viruses , the presence of innate and adaptive anti-viral immune barriers can significantly restrict the extent of direct vaccinia virus-mediated oncolysis. Importantly, recent data indicate that tumor burden might directly correlate with inadequate responsiveness to immunotherapy . Limited oncolysis and the lack of immediate reduction in tumor burden can therefore compromise the immuno-stimulatory effects of oncolytic viruses and greatly reduce their therapeutic efficacy.
Improving direct oncolysis is contingent on designing effective strategies to overcome the existent immunological barriers to oncolytic viruses that include complement/antibody neutralization, the intrinsic tumor cell anti-viral interferon responses, as well as the elimination of infected cells by innate and adaptive immune cell populations such as NK cells and T cells. The NK cell responses appear to be of particular interest due to their innate characteristics and immediate nature, in sharp contrast to the time needed for expansion of virus-specific T cells [71, 72]. Of note, the latter branches of adaptive immunity appear to be critical for the ultimate clearance of the virus from the patient and warrant safety as well as therapeutic efficacy associated with the induction of tumor-specific immunity.
The therapeutic efficacy of oncolytic viruses is often restricted by the limited permissiveness of the tumor to virus infection and amplification , which reflects patient specific differences, including intrinsic tumor cell resistance , functional interferon anti-viral responses , or the presence of non-permissive tumor-associated stroma [75, 76, 77]. In tumors with relatively modest permissiveness, vaccinia virus might have insufficient time to efficiently colonize the tumor before it gets eliminated by the innate and adaptive anti-viral immune responses. Augmenting the oncolytic potential of vaccinia virus in such circumstances might therefore require a strategy that combines efficient delivery with a boosted local virus amplification that together could provide a favorable environment where virus spread and colonization occurs faster than the induction of protective anti-viral state.
Ex-vivo expanded autologous or allogeneic mesenchymal stem cells represent a unique delivery platform that offers the combined power of boosted local virus amplification and the ability to transiently suppress anti-viral innate and adaptive immunity, which is key to the success of oncolytic virotherapy in vivo. Given that the application of personalized autologous MSCs is an approach that is rather expensive and impractical for clinical development, the feasibility and limitations of using established and easily available allogeneic stem cell lines needs to be evaluated and investigated further. In the current study, we demonstrate that the impressive ability of mesenchymal stem cells to amplify vaccinia virus as well as to recruit and sensitize tumor cells to virus infection might significantly improve the therapeutic potential and broaden efficacy against resistant and low-permissive tumors. In addition, MSCs demonstrate the ability to effectively suppress anti-viral NK cell responses that represent the earliest and most significant innate immune barrier to tumor colonization and oncolysis [71, 72]. Of note, while MSCs appear to be immunosuppressive against NK cells in both autologous and allogeneic settings, as to date it has been shown in the literature, we demonstrate that their immunosuppressive ability and virus amplification potential remain subject to significant allogeneic barriers associated with the production of IFNγ and direct cytotoxicity by both NK and T cells. Such rapid and exaggerated allogeneic responses or the outright rejection of the stem cells, therefore, represent a significant obstacle to the use of off-the-shelf MSC- or alternative cell-based delivery platforms in the clinic.
The hypoimmunogenic and immunosuppressive characteristics of mesenchymal stem cells have justified their extensive use in the treatment of various autoimmune and inflammatory conditions, demonstrating similar persistence and equivalent therapeutic efficacy in both autologous and allogeneic settings. Conflicting reports demonstrate the superiority of using autologous stem cells, which probably reflects different disease background or requirements for brief immunosuppression versus long-term persistence and tissue repair . In fact, numerous studies have already challenged the paradigm that allogeneic stem cells can be used in a one-size-fits-all universal donor setting, due to their hypoimmunogenic features. Instead, it becomes increasingly clear that MSC are not immune privileged and can induce allogeneic responses resulting in their ultimate rejection [79, 80].
Here, we argue that, to successfully utilize the potential of MSCs for the purposes of oncolytic virotherapy, it is necessary to build a much deeper understanding of the underlying mechanisms that control the balance between their immune evasive and immunogenic characteristics, and specifically how these are affected by the inflammatory environment associated with vaccinia virus infection. Importantly, long-term persistence of the allogeneic stem cells is irrelevant for the purposes of oncolytic virotherapy, while their short-term immunosuppressive and virus amplification potentials are essential and indispensable. Our demonstration that the combination of IFNγ production and direct cytotoxicity are associated with inability of the allogeneic Trojan horse to deliver and amplify vaccinia virus is critically important for the advancement of this therapeutic modality, as IFNγ plays a key role in both the control of vaccinia virus infection and the regulation of the immunosuppressive properties of MSCs. Multiple studies have demonstrated that exposure to IFNγ can significantly improve the immunosuppressive functions of MSCs by upregulation of IDO, iNOS/NO, COX2/PGE, and PD-L1 . This would give stem cells the unique ability to support virus colonization by subverting IFNγ-driven immune activation and anti-viral immunity, but previous studies as well as our own findings indicate that IFNγ can also exacerbate small allogeneic differences and increase the immunogenicity of stem cells, potentially through up-regulated expression of MHC Class I/II or other costimulatory molecules. It should be emphasized that, while IFNγ-pre-treatment might be beneficial to enhance MSC-mediated immunosuppression for the treatment of autoimmune and inflammatory conditions, where long-term engraftment is unnecessary or unaffected by minor allogeneic differences, this is certainly not the case with oncolytic virus approaches that are critically dependent on the ability of the Trojan horse to amplify the virus. Accordingly, inappropriate and rapid IFNγ responses against the allogeneic stem cells might be sufficient to inactivate the Trojan horse, even in the absence of outright rejection and compromised persistence in vivo.
Our findings reveal that the responses to vaccinia virus and allogeneic stem cells alone or in combination are highly patient-specific, thus demonstrating the need for further mechanistic studies aimed to validate the relative contribution of IFNγ and direct cytotoxicity for the inactivation of the allogeneic Trojan horse. Understanding the basis for this patient-specific permissiveness versus resistance to the Trojan horse is challenging as it can reflect multiple sources of variability, including MHC I/II mismatches, differences in the MHC-binding Killer Cell Immunoglobulin-like receptors (KIR Haplotype or ligands) or differential frequency/activation status of innate immune cell populations such as NK or NKT cells. The use of allogeneic stem cells for oncolytic virus delivery is greatly facilitated by the lack of requirement for long-term survival and engraftment of the cells, making this approach likely to work across insignificant MHC mismatch barriers and restricting patient-specific resistance to relatively small groups of patients, who can be excluded from clinical trials with the use of simple diagnostic assays, as the ones presented in this study. Alternatively, a more frequent or broader patient-specific resistance to allogeneic Trojan horses would require the establishment of allogeneic MSC banks and matching the patients to the available stem cell lines, based on MHC typing data or in vitro diagnostic assays, which take into account all of the complex immunological characteristics of patients and particularly those conferring rapid and exaggerated anti-viral/-stem cell responses able to significantly limit therapeutic efficacy in highly resistant individuals.
Given that amplification of the virus in the stem cells proceeds within hours to several days after infection, while it takes more than a week for the adaptive T cell responses to contain virus spread and oncolysis, it becomes evident that the kinetics of the anti-viral/stem cell responses within the first few days of treatment would be critical for the ability of the Trojan horse to have a significant therapeutic effect in vivo. Consequently, the crosstalk and interplay between different innate and adaptive branches of the immune system need to be further investigated as these might be directly associated with the speed and magnitude of the anti-viral/stem cell immunity. The combination of exaggerated innate immune mechanisms sensitizing adaptive immunity against the virus or the allogeneic stem cells can be particularly detrimental. We were very interested by the finding that a patient can be broadly resistant to several allogeneic stem cell lines and that this broad resistance was also associated with unusually high frequency of NKp46 + CD3+ NKT-like cells, which were also the only population that expanded in numbers in response to both the virus and the allogeneic stem cells. Unfortunately, our attempts to link this NKT-like population of cells with effector functions able to directly interfere with the Trojan horse were unsuccessful, as these cells didn’t manifest any significant contribution to IFNγ production or cytotoxic activity. NKT cells are known to play a role in the control of viral infections, but their involvement in vaccinia virus responses and immunity hasn’t been investigated fully yet. The fact that this population of cells is unlikely to be responsible for the direct rejection or inactivation of the Trojan horse does not eliminate the possibility that NKT or NKT-like cells are critically important for directing and accelerating coordinated NK and T cell responses against the stem cells or the virus. Detailed kinetic and mechanistic studies would be necessary to evaluate the possibility that innate immune cells like NKT cells provide early cytokine help sensitizing NK and T cell to the presence of the virus and/or potential allogeneic differences, as suggested by some of our preliminary data (Additional file 7: Figure S7D). The importance of the crosstalk between the innate (NK/NKT cells) and adaptive (B/T cells) arms of the immune system, as revealed by our data, suggests the existence of correlation between patient-specific NK and T cell responses. Despite the rather limited cohort of PBMC donors tested, it becomes evident that patients mount coordinated NK, NKT and T cell responses against the virus and allogeneic stem cells (Fig. 5d). On the other hand, the comparative responsiveness to the virus and stem cells appears discordant and highly patient-specific (Additional file 5: Figure S5C), which cannot be explained by HLA typing data alone and suggests the involvement of innate immunological differences that might not be the same with respect to the virus or possible MHC mismatches. These innate immunological differences can be associated with differences in the NK or NKT cells and more precisely with the balance of signaling through their activating versus inhibitory receptors such as NKG2A/NKG2C/NKG2D or KIR receptors, which like certain HLA alleles are highly patient-specific and have already been linked to resistance/susceptibility to infectious agents, autoimmune diseases, and cancer [81, 82, 83, 84, 85, 86].
Overall, our data indicate that while autologous mesenchymal stem cells are potentially the optimal vehicles for the delivery and amplification of oncolytic viruses, the use of properly matched allogeneic stem cell lines in combination with robust companion diagnostic assays could provide a more practical and commercially viable alternative that guarantees consistent stem cell quality and validated amplification potential. This approach also provides the unique opportunity to utilize readily available off-the-shelf cell-based delivery platforms in a highly efficient personalized fashion.
We demonstrate that mesenchymal stem cells, in addition to protecting naked viruses, can amplify vaccinia virus, enhance its delivery, and suppress anti-viral innate and adaptive immunity to potentiate oncolytic virotherapy. Our data also reveal that the ability of stems cells to amplify and deliver vaccinia virus is subject to allogeneic barriers that require proper patient-to-stem cell matching. We have also demonstrated that the resistance of patients to allogeneic carrier cells is associated with the induction of anti-stem cell IFNγ and cytotoxic responses, likely reflecting patient-specific HLA or KIR Haplotype mismatches. This study provides fundamental understanding of the molecular principles behind patient-specific resistance and will guide the future clinical development of cell-based delivery platforms for oncolytic virus therapy of cancer.
DDD and AAS conceptualized the project. DDD (designed/performed experiments, analyzed data, and wrote the manuscript); AS (generation of ADSC lines and eGFP-transfection); IM (SVF and PBMC donor coordinator); DN (recombinant vaccinia virus generation and plaque purification); MOK (established ADSC characterization and expansion protocols); IP and AV (preparation and validation of virus stocks, GFP transfection of stem cells, established MTT and plaque assays to evaluate sensitivity of tumor cell lines to VV); EL and MB (provided SVF for ex vivo expansion of ADSCs), BM (MHC Analysis); Aladar A. Szalay (coordinated the project and edited the manuscript); All authors read and approved the final manuscript.
Acknowledgements and Funding
The authors would like to thank for the PBMC and SVF research sample aliquotes donated by patients at the California Stem Cell Treatment Center. This Calidi Bioterapeutics’ funded research was supported by the Foundation of the International Cell Surgical Society at Rancho Mirage, by the cell and tissue imaging facility at the University of Wuerzburg, and by the Stem Cell Facility at StemVac GmbH. We are thankful for the generous research grant support and the donations of graduate student fellowships by the management of Calidi Biotherapeutics. We are thankful to Mr. Stephen Keane and Mrs. Ulrike Szalay for reviewing and editing this manuscript. The authors would also like to express their gratitude to Dennis Young and the Cytometry Core Facility of the Moores Cancer Center, San Diego, CA, for providing technical services in support of this study.
The following authors DD, AS, IM, DN are receiving salaries and stock options, while BM and AASz are equity owners and receive full or partial salary from Calidi Biotherapeutics. EL and MB are employees and shareholders of California Stem Cell Treatment Center and stock options holders in Calidi Biotherapeutics. IP and AV have received financial support from Calidi Biotherapeutics.
Availability of data and materials
The datasets supporting the conclusions of this article are included within the article and its additional files.
Consent for publication
All authors have consented to the publication of this manuscript.
Ethics approval and consent to participate
All work presented has been done with proper ethics approval and informed written consent to participate under an IRB-approved protocol (International Cell Surgical Society; IRB# ICSS-2016-024).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 1.Leoni V, Gatta V, Palladini A, Nicoletti G, Ranieri D, Dall’Ora M, et al. Systemic delivery of HER2-retargeted oncolytic-HSV by mesenchymal stromal cells protects from lung and brain metastases. Oncotarget. 2015;6:34774–87. https://doi.org/10.18632/oncotarget.5793.CrossRefPubMedPubMedCentralGoogle Scholar
- 3.Alcamí A, Symons JA, Smith GL. The vaccinia virus soluble alpha/beta interferon (IFN) receptor binds to the cell surface and protects cells from the antiviral effects of IFN. J Virol. 2000;74:11230–9. https://doi.org/10.1128/JVI.74.23.11230-11239.2000.CrossRefPubMedPubMedCentralGoogle Scholar
- 4.Kober C, Weibel S, Rohn S, Kirscher L, Szalay AA. Intratumoral INF-γ triggers an antiviral state in GL261 tumor cells: a major hurdle to overcome for oncolytic vaccinia virus therapy of cancer. Mol Ther Oncolytics. 2015;15009:1–14.Google Scholar
- 13.Melen GJ, Franco-Luzón L, Ruano D, González-Murillo Á, Alfranca A, Casco F, et al. Influence of carrier cells on the clinical outcome of children with neuroblastoma treated with high dose of oncolytic adenovirus delivered in mesenchymal stem cells. Cancer Lett. 2016;371:161–70. https://doi.org/10.1016/j.canlet.2015.11.036.CrossRefPubMedGoogle Scholar
- 17.Trachtenberg B, Velazquez DL, Williams AR, McNiece I, Fishman J, Nguyen K, et al. Rationale and design of the transendocardial injection of autologous human cells (bone marrow or mesenchymal) in chronic ischemic left ventricular dysfunction and heart failure secondary to myocardial infarction (TAC-HFT) trial: a randomized, double-blind. Am Heart J. 2011;161:487–93. https://doi.org/10.1016/j.ahj.2010.11.024.CrossRefPubMedGoogle Scholar
- 18.Lechanteur C, Baila S, Janssens ME, Giet A, Briquet A, Baudoux E, et al. Large-scale clinical expansion of mesenchymal stem cells in the gmp-compliant, closed automated quantum® cell expansion system: comparison with expansion in traditional T-flasks. J Stem Cell Res Ther. 2014;4:1–11. https://doi.org/10.4172/2157-7633.1000222.CrossRefGoogle Scholar
- 20.Spaggiari GM, Capobianco A, Abdelrazik H, Becchetti F, Mingari MC, Moretta L. Mesenchymal stem cells inhibit natural killer-cell proliferation, cytotoxicity, and cytokine production: role of indoleamine 2,3-dioxygenase and prostaglandin E2. Blood. 2008;111:1327–33. https://doi.org/10.1182/blood-2007-02-074997.CrossRefPubMedGoogle Scholar
- 22.DelaRosa O, Lombardo E, Beraza A, Mancheño-Corvo P, Ramirez C, Menta R, et al. Requirement of IFN-gamma-mediated indoleamine 2,3-dioxygenase expression in the modulation of lymphocyte proliferation by human adipose-derived stem cells. Tissue Eng Part A. 2009;15:2795–806. https://doi.org/10.1089/ten.TEA.2008.0630.CrossRefPubMedGoogle Scholar
- 23.Najar M, Raicevic G, Boufker HI, Fayyad Kazan H, De Bruyn C, Meuleman N, et al. Mesenchymal stromal cells use PGE2 to modulate activation and proliferation of lymphocyte subsets: combined comparison of adipose tissue, Wharton’s Jelly and bone marrow sources. Cell Immunol. 2010;264:171–9. https://doi.org/10.1016/j.cellimm.2010.06.006.CrossRefPubMedGoogle Scholar
- 24.Spaggiari GM, Capobianco A, Becchetti S, Mingari MC, Moretta L. Mesenchymal stem cell-natural killer cell interactions: evidence that activated NK cells are capable of killing MSCs, whereas MSCs can inhibit IL-2-induced NK-cell proliferation. Blood. 2006;107:1484–90. https://doi.org/10.1182/blood-2005-07-2775.CrossRefPubMedGoogle Scholar
- 30.Yang Y, Chen Q-H, Liu A-R, Xu X-P, Han J-B, Qiu H-B. Synergism of MSC-secreted HGF and VEGF in stabilising endothelial barrier function upon lipopolysaccharide stimulation via the Rac1 pathway. Stem Cell Res Ther. 2015;6:250. https://doi.org/10.1186/s13287-015-0257-0.CrossRefPubMedPubMedCentralGoogle Scholar
- 32.Cai L, Johnstone BH, Cook TG, Liang Z, Traktuev D, Cornetta K, et al. Suppression of hepatocyte growth factor production impairs the ability of adipose-derived stem cells to promote ischemic tissue revascularization. Stem Cells. 2007;25:3234–43. https://doi.org/10.1634/stemcells.2007-0388.CrossRefPubMedGoogle Scholar
- 39.Selmani Z, Naji A, Zidi I, Favier B, Gaiffe E, Obert L, et al. Human leukocyte antigen-G5 secretion by human mesenchymal stem cells is required to suppress T lymphocyte and natural killer function and to induce CD4+ CD25highFOXP3+ regulatory T cells. Stem Cells. 2008;26:212–22. https://doi.org/10.1634/stemcells.2007-0554.CrossRefPubMedGoogle Scholar
- 41.Valencia J, Blanco B, Yáñez R, Vázquez M, Herrero Sánchez C, Fernández-García M, et al. Comparative analysis of the immunomodulatory capacities of human bone marrow—and adipose tissue-derived mesenchymal stromal cells from the same donor. Cytotherapy. 2016;18:1297–311. https://doi.org/10.1016/j.jcyt.2016.07.006.CrossRefPubMedGoogle Scholar
- 42.Bourin P, Bunnell BA, Casteilla L, Dominici M, Katz AJ, March KL, et al. Stromal cells from the adipose tissue-derived stromal vascular fraction and culture expanded adipose tissue-derived stromal/stem cells: a joint statement of the International Federation for Adipose Therapeutics (IFATS) and Science and the International S. Cytotherapy. 2013;15:641–8.CrossRefGoogle Scholar
- 44.García-Arranz M, Herreros MD, González-Gómez C, de la Quintana P, Guadalajara H, Georgiev-Hristov T, et al. Treatment of Crohn’s-related rectovaginal fistula with allogeneic expanded-adipose derived stem cells: a Phase I-IIa clinical trial. Stem Cells Transl Med. 2016;5:1441–6. https://doi.org/10.5966/sctm.2015-0356.CrossRefPubMedPubMedCentralGoogle Scholar
- 45.Gao W, Zhang L, Zhang Y, Sun C, Chen X, Wang Y. Adipose-derived mesenchymal stem cells promote liver regeneration and suppress rejection in small-for-size liver allograft. Transpl Immunol. 2017; July:0–1. https://doi.org/10.1016/j.trim.2017.07.005.
- 58.Abboud G, Tahiliani V, Desai P, Varkoly K, Driver J, Hutchinson TE, et al. Natural killer cells and innate interferon gamma participate in the host defense against respiratory vaccinia virus infection. 2016;90:129–41.Google Scholar
- 63.Don’skoi BV, Chernyshov VP, Osypchuk DV. Measurement of NK activity in whole blood by the CD69 up-regulation after co-incubation with K562, comparison with NK cytotoxicity assays and CD107a degranulation assay. J Immunol Methods. 2011;372:187–95. https://doi.org/10.1016/j.jim.2011.07.016.CrossRefGoogle Scholar
- 67.Maccario R, Podestà M, Moretta A, Cometa A, Comoli P, Montagna D, et al. Interaction of human mesenchymal stem cells with cells involved in alloantigen-specific immune response favors the differentiation of CD4+ T-cell subsets expressing a regulatory/suppressive phenotype. Haematologica. 2005;90:516–25.PubMedGoogle Scholar
- 75.Arulanandam R, Batenchuk C, Angarita FA, Ottolino-Perry K, Cousineau S, Mottashed A, et al. VEGF-mediated induction of PRD1-BF1/Blimp1 expression sensitizes tumor vasculature to oncolytic virus infection. Cancer Cell. 2015;28:210–24. https://doi.org/10.1016/j.ccell.2015.06.009.CrossRefPubMedGoogle Scholar
- 78.Ben Nasr M, Vergani A, Avruch J, Liu L, Kefaloyianni E, D’Addio F, et al. Co-transplantation of autologous MSCs delays islet allograft rejection and generates a local immunoprivileged site. Acta Diabetol. 2015;52:917–27. https://doi.org/10.1007/s00592-015-0735-y.CrossRefPubMedPubMedCentralGoogle Scholar
- 82.Jafari D, Nafar M, Yekaninejad MS, Abdolvahabi R, Pezeshki ML, Razaghi E, et al. Investigation of killer immunoglobulin-like receptor (KIR) and HLA genotypes to predict the occurrence of acute allograft rejection after kidney transplantation. Iran J Allergy Asthma Immunol. 2017;16:245–55.PubMedGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.