香京julia种子在线播放

Female C57BL/6 mice and apoE ^-/- mice (B6.129P2-Apoe^tm1Unc/J) aged 6 weeks were purchased from Jackson Laboratories (Bar Harbor, Maine, United States). Lrp8 ^-/- C57BL/6 breeder mice were generously provided by Dr. Sohail Tavazoie’s laboratory from the Rockefeller University. Mice were housed five per cages and kept in a temperature-controlled environment (20 ± 2°C, 50 ± 5% relative humidity) with a 12-hour light/dark cycle in an air-conditioned room with free access to food and water. The animals were acclimated for 4–5 days prior to tumor challenge. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Children’s National Hospital, Washington, DC.

The murine melanoma B16-F10 cell line (purchased from ATCC^® CRL-6475, VA), were cultured in DMEM (Life Technologies, CA) supplemented with 10% heat-inactivated FBS (Sigma-Aldrich) and 100 IU/ml Penicillin, 100 μg/ml Streptomycin (Life Technologies).

Six guide RNAs (gRNAs) were designed using IDT’s gRNA design tool, targeting the mouse apoE gene. gRNAs were synthesized as 2-part crRNA: tracrRNA gRNAs with chemical modifications (Integrated DNA Technologies, Inc., IA) and were functionally screened by next generation sequencing (NGS) for high INDEL (insertion/deletion) frequency and low in-frame INDEL rates. gRNAs were prepared by complexing a 1:1 molar ratio of the crRNA: tracrRNA at a final concentration of 100 μM, heating to 95°C and slowly cooling to room temperature. Ribonucleoprotein complexes (RNPs) were formed by the addition of purified Alt-R HiFi Cas9 protein (IDT) to each gRNA at a 1.2:1 molar ratio in 1× PBS to a concentration of 5.6 μM. RNP complexes were allowed to form for 10 min at room temperature before electroporation. RNP complexes (5 μL), Alt-R Cas9 Electroporation Enhancer (3 μL), and 350,000 B16-F10 cells (20 μL) resuspended in Buffer SF were mixed and electroporated using the Lonza 96-well Shuttle System (Lonza, Basel, Switzerland) with electroporation protocol 96-DS-150. Final concentrations for RNP and Alt-R Cas9 Electroporation Enhancer were 1 μM and 4 μM, respectively. Genomic DNA was extracted 48 hours post-transfection using QuickExtract DNA extraction solution (Epicentre Biotechnologies, CA) according to the manufacturer’s specifications. The targeting sequence for each ApoE crRNAs are listed in Supplementary Table 1 .

The lead gRNA (Mm. Cas9.APOE.1-E) resulted in a 99% INDEL frequency with no in-frame INDELs and was electroporated into B16-F10 cells using the Lonza 96-well Shuttle System as previously described. The electroporated cells were plated in 1 well of a 6-well plate and allowed to grow until confluent. The cells were then dissociated by trypsinization, resuspended in media, and counted. The suspension was diluted to 20,000 cells/mL; 4000 cells were added to 1 well of a 96-well plate and then diluted by array dilution. After 5 days of growth, each well was visually screened for single colonies. Wells with only 1 colony were allowed to grow to confluency. Each well was progressively passaged to a larger well until confluent in a 100 mm dish, about 8.8 x 10^6 cells for genomic DNA extraction and further cell passaging. The genomic DNA from each well was subject to quantification of total editing and analysis of INDEL profile by NGS to confirm a monoclonal isolate.

Genomic DNA libraries for sequencing were prepared using IDT rhAmpSeq targeted amplification. In short, the first round of PCR was performed using target-specific primers with universal 5’ tails ( Supplementary Table 2 ); a second round of PCR incorporated P5 and P7 Illumina adapter sequences to the amplicon ends. Libraries were purified using Agencourt^® AMPure^® XP system (Beckman Coulter, Brea, CA, USA) with a 1:1 ratio of beads to reaction by volume and quantified with quantitative real-time PCR (qPCR) before loading onto the Illumina^® MiSeq platform (Illumina, San Diego, CA, USA). Paired end 150 base pair reads were sequenced using V2 chemistry. A sequencing depth of at least 1000 reads was obtained for each sample. Total editing efficiency was calculated and INDEL profiles were evaluated using an IDT custom-built pipeline, CRISPAltRations.

Anti (α)-mouse CTLA-4, and mouse IgG2b isotype antibodies were purchased from BioXCell (West Lebanon, NH). COG133 and JQ1 were purchased from Tocris (Minneapolis, MN). Dynabeads™ Mouse T-Activator CD3/CD28 for T-Cell Expansion and Activation kit, Vybrant™ DyeCycle™ Violet Stain kit, SYTOX™ red dead cell stain kit, CellTrace™ far red cell proliferation kit, Live/Dead fixable aqua dead cell stain kit, Brilliant stain buffer and mouse IL-2 Carrier-Free recombinant protein were purchased from Thermo Fisher (Waltham, MA).

Cell culture supernatant was collected after centrifuge at 1,200 rpm for 10 min at 4°C. The concentrations of the following immune molecules were determined using the mouse Cytokine & Chemokine 36-plex ProcartaPlex Panel, a magnetic bead-based multiplex immunoassay (Thermo Scientific) following our previous protocol (19). Briefly, cell culture supernatant samples were mixed with antibody-linked polystyrene beads on a 96-well plate and incubated at room temperature (RT) for 2 h on an orbital shaker at 500 rpm. After washing, plates were incubated with biotinylated detection antibody for 30 min at RT. Plates were then washed twice and then labeled beads were re-suspended in streptavidin-PE. Each sample was measured in duplicate along with standards (8-point dilutions) and the buffer control. Plates were read using a Luminex Bio-plex 200 system (Bio-Rad Corp.) for quantitative analysis.

Spleens were collected from mice euthanized by CO₂ narcosis and cervical dislocation. Spleens were pulverized through a 40-μm mesh cell strainer and treated with ACK lysing buffer (Thermo Fisher) for 10 seconds to remove erythrocytes. According to the manufacture’s instruction of the Pan T Cell Isolation Kit (Miltenyi Biotec), 10 µL of Pan T Cell Biotin Antibody Cocktail was added per 10⁷ splenocytes and incubated for 5 min at 4°C. Subsequently, 20 µL of Pan T Cell MicroBead Cocktail was added per 10⁷ cells. Following incubation for 10 min at 4°C, the mixed cell suspension was applied onto the LS column (Miltenyi Biotec). The flow-through containing unlabeled cells, representing the enriched T cells were collected. T cells were cultured in RPMI 1640 media containing 30 U/mL IL-2 (Thermo Fisher) and stimulated with Dynabeads^® Mouse T-Activator CD3/CD28 magnetic beads (Thermo Fisher) at a 1:1 ratio (cell: bead).

WT and apoE^-/- mice or WT and lrp8^-/- mice, were inoculated with either WT B16 or apoE^-/- B16 cells, with anti-CTLA4 antibody on day 0. These vaccinated splenocytes (VS) were harvested on day 7 and co-cultured with either WT B16 or apoE^-/- B16 cells for 48hr, following which IFNγ levels in media were compared with ELISA assay. To set up co-culture, a total of 5 × 10⁵ freshly isolated mouse T cells or splenocytes were plated in a volume of 600 μl per well of 24-well plates, then they were co-cultured with 5 × 10⁴ B16 cells and stimulated with or without CD3/CD28 Dynabeads. Cells were exposed to apoE agonist COG133 at 0, 0.3, 3, 9, 15, 30 µM or human anti-APOE antibody at 1 µg/ml, 10 µg/ml and 20 µg/ml at 37°C under 5% CO2 for 24 hr or 48 hr. Supernatants were collected from triplicate wells, and IFNγ was assayed using the mouse uncoated IFNγ ELISA kit from Invitrogen (Carlsbad, CA). Readings were measured at 450 nm using the EnSpire 2300 Multilabel plate reader (Perkin Elmer, Waltham, Massachusetts, US).

5x10⁶ wild type (WT) and apoE^-/- B16 cells were irradiated at 60 Gy and then cultured in 20 mL DMEM media supplemented with 10% FBS for 48 hr in T75 flask. The conditioned media (CM) was collected and centrifuged at 1100 rpm at room temperature for 5 min. The supernatant was aliquoted and stored at -80°C.

For testing cell cycle, 1x10⁶ T cells in 1 mL DMEM complete media were incubated with 5 µM Vybrant™ DyeCycle™ Violet Stain (Thermo Fisher) at 37˚C for 30 min. And then 5 µl 7AAD were added and incubated for 10 min prior to analysis. Total 25,000 cells were analyzed per measurement. The same forward and side scatter gates were applied to each sample, and within that gate we measured the intensity of vibrant cell cycle dye. Samples were analyzed on a flow cytometer using 405 nm excitation and 440 nm emission. To compare the growth rates of B16 WT and apoE^-/- cells, the Click-iT Edu Alexa Fluor 488 flow cytometry assay kit was used in conjunction with the FxCycle Violet stain from Invitrogen (Carlsbad, CA). 50,000 cells were plated per well of a 6 well-plate and cell cycle was analyzed at 48 hr as per the manufacturer’s directions.

Mouse serum was collected from naïve WT C57/BL6 mice and also from both naïve and tumor-bearing C57/BL6 apoE^-/- mice. ApoE levels in mouse sera were quantified using the mouse apoE ELISA^PRO kit from Mabtech (Cincinnati, OH) as per the manufacturer’s directions.

RNA was extracted and gene expression was directly measured via counts of corresponding mRNA in each sample using an nCounter murine PanCancer Immune Profiling Panel (NanoString, Seattle, WA, USA). For full details, see our previous publication (19). Briefly, 100 ng of high-quality total RNA were hybridized with reporter probes, and then biotinylated capture probes at 65°C for 16–18 hr before being placed into the nCounter Prep station in which samples were affixed to a cartridge. Cartridges were then read by the nCounter Digital Analyzer optical scanner. Further advanced immune-profiling analysis was performed using nSolver 4.0 analysis software with nCounter advanced analysis package (NanoString Technologies) with identified immune cell types. Genes were grouped into 14 immune cell types and 39 immune functions according to the manufacturer’s designation (19).

Quantitative real-time PCR (qPCR) was performed using TaqMan^® Gene Expression Master Mix (Life Technologies) in a QuantStudio 7 Flex Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA) following the methods that we published previously (19). Each reaction was performed in triplicate, including no template controls and amplification of a housekeeping gene, GAPDH. Gene-specific assays were Mm01307192_m1 for apoE, Mm00464608_m1 for Lrp1, Mm01328171_m1 for Lrp2, Mm00474030_m1 for Lrp8, Mm01177349_m1 for Ldlr, Mm00443298_m1 for Vldlr, Mm99999915_g1 for Gapdh (Life Technologies, Thermo Fisher). Changes in relative gene expression normalized to GAPDH levels were determined using the ΔΔCt method. Results were averaged and statistically analyzed using t-tests.

C57BL/6 wild type (WT) mice, C57/BL6 apoE knockout (apoE^-/-) mice, and C57/BL6 lrp8 knockout (lrp8^-/-) mice were injected subcutaneously in the right flank with 1 × 10⁴ freshly prepared B16 WT or apoE^-/- tumor cells in 100 µl 1xPBS on day 0 and euthanized once the tumor reached 20mm in any dimension. Tumor growth was recorded every day by measuring the diameter in 2 dimensions using a caliper when appropriate as we have previously published (19, 20). Briefly, tumor volume was calculated using the following formula: large diameter² × small diameter × 0.52. A tumor size of 20 mm in diameter in any dimension was designated as the endpoint, and mice were euthanized at that time. Euthanasia was achieved through cervical dislocation after CO2 narcosis. If a tumor impaired the mobility of an animal, became ulcerated, or appeared infected, or a mouse displayed signs of “sick mouse posture”, the mouse was euthanized. All the procedures are approved by the IACUC at CNMC and are in accordance with the humane care of research animals.

Tumor was fixed in 10% neutral buffered formalin (pH 6.8–7.2; Richard-Allan Scientific, Kalamazoo, Michigan, US) for paraffin embedding and sectioning. Five μm tissue sections were cut with a microtome, and sample processing and IHC staining were performed as previously described (19) using rabbit polyclonal to CD45 and CD3 antibodies (1:200. Abcam, Cambridge, Massachusetts, US). Isotype-matched antibodies were used for negative controls.

Statistical analysis of nanostring gene expression, normalization, clustering, Pathview plots and fold-changes were performed using the Advanced Analysis Module in the nSolver™ Analysis Software version 4.0 from NanoString Technologies (NanoString Technologies, WA, USA) following our published methods (19). Briefly, raw data for each sample were normalized to the geometric mean of housekeeping genes using the geNorm algorithm. Pathway scores were calculated as the first principal component (PC) of the pathway genes’ normalized expression. Each cell type score has been centered to have mean 0 and as abundance estimates (cell type scores) are calculated in log2 scale, an increase of 1 corresponds to a doubling in abundance. All differentially expressed genes were subjected to KEGG term analysis, with significance accepted at p < 0.05. The Benjamini-Yekutieli method was used to control the false discovery rate. All statistical analyses of nanostring data were carried out in R v3.4.3 software.

Statistical significance for each set of experiments was determined by the unpaired 2-tailed Student’s t-test, and the specific tests were indicated in the figure legends. The data are expressed as the mean ( ± SD), with p<0.05 considered statistically significant.

We accessed the raw RNA-seq data of melanoma biopsies through the Gene Expression Omnibus database (accession number: GSE78220) (21) - Raw reads mapping to the reference genome (GRCh38) were performed on quality-checked and trimmed reads using STAR 2.4.1c (22). The reference annotation is Ensembl v86. The overlap of reads with annotation features found in the reference.gtf were calculated using HT-seq v0.6.1 (23). - The output computed for each sample (raw read counts) was then used as input for DESeq2 (24) analysis. Raw counts were normalized using DESeq2 function “rlog,” and normalized counts were used for downstream analysis. Statistical calculations were performed using GraphPad Prism software (GraphPad Software, San Diego, CA, USA) or R Software (Version 4.0).

RNA was extracted and gene expression was directly measured via counts of corresponding mRNA in B16-F10 cells using an nCounter murine PanCancer Immune Profiling Panel (NanoString, Seattle, WA, USA). We evaluated the presence of multiple genes associated with immunosuppression in the B16-F10 melanoma cell line, ApoE was the most highly expressed immune-suppressive transcript in the cell line. ( Figure 1A ) ApoE is also constitutively detected in the serum of wild type (WT) C57/BL6 mice at high levels ( Figure 1B ). To evaluate the systemic levels secreted from the mouse melanoma tumor itself, we measured the level of serum apoE in apoE KO (apoE^−/−) C57/BL6 mice inoculated subcutaneously (s.c.) with 10⁴ WT B16 (F10) cells injected in the right thigh. Blood and tumors were collected at various sizes until the tumors reached a max of 21 mm in any dimension. The serum levels of apoE in these tumor-baring apoE^-/- mice increased progressively and considerably with tumor growth ( Figure 1C ).

To further investigate the function of apoE, we generated apoE^-/- B16-F10 cell lines with CRISPR-Cas9 gene deletion. To confirm suppression or deletion of apoE protein, we performed western blot analysis on total protein lysates from B16 WT and apoE^-/- single clones. Our data shows absent apoE expression in the apoE^-/- single clones ( Figure 1D ). The apoE protein secreted from WT B16 and apoE knockout clones in culture media was quantified by ELISA assay and levels correlated with the protein expression pattern of western blot analysis confirming KO of apoE in the KO cell lines. ( Figure 1E )

We then evaluated the in vitro proliferation rate ( Figure 1F ) and cell cycling ( Figure 1G ) of the WT and KO cell lines to determine if targeting apoE influenced cell viability and proliferation. There is no statistically significant difference between WT and ApoE^-/- cells in their proliferation rate or cell cycles ( Figures 1F, G ).

To investigate the effect of tumor cell secreted apoE on activated T cells, splenic derived C57/BL6 T cells were cultured in control media and melanoma B16 WT or apoE ^-/- conditioned media (CM) in which the T cells were activated with CD3/CD28 beads for 48 hr. Cytokine secretion, apoptosis and proliferation of the cultured mouse T cells were examined and compared to T cells cultured in the RPMI media that served as controls ( Figure 2 ). IFNγ production ( Figure 2A ) and T cell viability ( Figures 2B, C ) were significantly (P < 0.05) suppressed when T cells were cultured in WT B16 CM at 48 hr of incubation as compared to the RPMI media control, whereas apoE^-/- conditioned media was similar to RPMI control media alone and did not inhibit T cell activity nor viability ( Figures 2A–C ). In addition, cell cycle distribution showed that WT conditioned media arrested activated T cells in the G0/G1 phase while apoE ^-/- conditioned media-maintained T-cells in the S and G2M phase of the cell cycle, similar to control media without prior exposure to tumor cells. ( Figures 2D, E ). These results indicate that conditioned media derived from B16 tumor cells induces arrest of stimulated T cells in the G0/G1 phase of the cell cycle, resulting in apoptosis and suppression of cytokine production. In contrast, the singular absence of apoE in conditioned media rescues the activated T cell phenotype when stimulated with CD3/CD28 beads. To further define the cytokine response, we used a ProcartaPlex multiplex immunoassay, and quantified multiple cytokines/chemokines in the stimulated T-cell media from the same experiment. Ten out of 27 detectable cytokines suppressed by WT conditioned media returned to baseline levels or were upregulated when T cells were cultured in apoE ^-/- CM. These included effectors: IFNγ, IFNα, TNFα; stimulator, IL18; inflammatory factor: IL4, IL13; chemo attractive factor: MCP-1, MCP3, MIP-1α, MIP-1β; and regulatory factor: IL-10 (P < 0.05) cytokines. IL6, CCL-5 (Rantes) and GroαKC were downregulated ( Supplement Figure 1A ).

To further delineate the T-cell suppressive effect directly, COG133, a fragment of apoE peptide, which competes with the apoE holoprotein for binding the low-density lipoprotein (LDL) receptors and acts as an apoE mimetic, was tested to determine its effect on T cell activation. T cells were activated with CD3/CD28 dynabeads and the effects of COG133 (0, 3, 9, 15 and 30µM concentrations) on T cell apoptosis were evaluated. The T cell viability was diminished in a dose dependent manner ( Supplement Figure 1B ). These results confirm the immunomodulatory role of apoE on activation of T cells showing robust and extensive suppression of T cell function.

Activation of innate antigen presenting cells like dendritic cells (DC) are critical for effective induction of immunity. We tested the effect of conditioned media on toll like receptor (TLR7/8) stimulated primary bone marrow derived dendritic cell activation. B16 WT conditioned media (CM) suppressed DC activation as determined by cytokine production, but this effect was absent in in media from apoE^-/- B16-F10 cells ( Supplement Figure 2 ). Multiplex results showed that secretion of the suppressive cytokine IL10 is diminished when DC were cultured in apoE^-/- CM compared to culture in WT CM. In a pro-inflammatory fashion IL1α, IL1β, RANTES, MIP1α, MIP1β, IL28 are all increased with apoE^-/- CM compared to WT CM ( supplement Figure 2 ). To further delineate the DC suppressive effect directly, we tested the effect of the apoE peptide mimetic, COG133 on DC activation. In a dose dependent fashion, COG 133 induced secretion of the anti-inflammatory cytokine, IL-10 from activated DC. Furthermore, COG133 suppressed IL-1α, IL-1β and IL-23 in a dose dependent manner suggesting the immune-modulatory role of ApoE on DC function ( Supplement Figure 3 ).

To further investigate the influence of tumor secreted apoE on immune cell function, immunogenic B16 tumor cells (B16 cells treated with Myc inhibitors (0.25µM BET+0.25µM JQ1 for 4 days) were irradiated at 60 Gray and co-cultured with naïve C57/BL6 splenocytes in the presence of either apoE mimetic COG133 at 0.3, 3 and 9 µM or anti-APOE blocking antibody at 1, 10 and 30 µg/ml concentrations. Prior work from our laboratory has shown the immunogenic effect of treating cancer cells with Myc inhibitors and irradiation (19). IFNγ production was quantified by ELISA at 48h. Splenocytes produced high levels of IFNγ when co-cultured with Myc-inhibited immunogenic B16 cells as shown in Figure 3 . Exposure of these cells to apoE mimetic COG133 repressed IFNγ production (6-fold reduction) ( Figure 3A ), while the presence of Anti-APOE antibody enhanced IFNγ production (3-fold increase) ( Figure 3B ). We also tested IFNγ production from vaccinated splenocytes following co-culture with treated and untreated B16 cells in the presence of COG 133. To obtain vaccinated splenocytes, 10^4 WT B16 tumor cells and 100µg/ml anti-CTLA4 antibody were administered to C57BL/6 mice on day 0, and splenocytes were collected at day 7 after tumor cell inoculation. Compared with naïve splenocytes, vaccinated splenocytes produced dramatically greater level of IFNγ, especially when they were co-cultured with Myc inhibited B16 tumor cells. This high level IFNγ was also suppressed by COG 133 at a dose dependent manor ( Figure 3C ). In addition to IFNγ, we also quantified other cytokine/chemokines using ProcartaPlex multiplex immunoassay. Fourteen out of 23 detectable cytokines were significantly upregulated when splenocytes were co-cultured with Myc-inhibited B16 tumor cells, including effectors: IFNγ, TNFα; stimulators, IL18, G-CSF, M-CSF; inflammatory factor: IL6; chemo attractive factors: CCL-5 (Rantes), CXCL-1, CCL-2, CCL-7, CXCL-2, CCL3; and regulatory factors: IL-10, IL6 (P < 0.05). Within these 14 cytokines, four of them including IFNγ, IL6, IL18 and RANTES (CCL5) were suppressed after exposure to apoE peptide mimetic COG133 in a dose dependent manner ( Supplement Figure 4 ).

The pattern of expression of apoE receptors on T cells and dendritic cells is not fully characterized. Here we examined the expression of apoE receptors on T cells, dendritic cells (DC) and macrophages using qPCR. All five of the receptor transcripts were shown to be expressed on T cells, DCs and macrophages. Ldlr and lrp8 were dominantly expressed on T cells ( Figure 4A ) whereas lrp1, lrp8 and ldlr are highly expressed on DC ( Figure 4B ). Lrp1 is highly expressed in macrophages ( Figure 4C ). Vldlr expression was relatively low, and lrp2 was barely detectable on these three cell types. We further examined the effect on expression of these receptors by stimulating T cells with CD3/CD28 beads and stimulating DCs and macrophages with a TLR7/8 agonist. Expression of lrp1, lrp8, ldlr and vldlr were all significantly upregulated on T cells following activation ( Figure 4A ), whereas only lrp2 and lrp8 were increased on DC after TLR stimulation ( Figure 4B ). This data together with previous studies (25, 26) suggest that ldlr and lrp8 (apoER2) are dominantly expressed and may be prominently engaged in apoE-mediated immune cell suppression.

To functionally define the role of the lrp8 receptor engagement in apoE suppression, we isolated vaccinated splenocytes from lrp8^-/- mice. These cells were co-cultured with BET/JQ1 treated (Myc-suppressed) immunogenic B16 cells for 48hr with or without exposure to the apoE mimetic COG133. ELISA results show that IFNγ production from splenocytes was inhibited by COG133 in WT mice, however the inhibitory effect was lacking in splenocytes from the lrp8^-/- mice ( Figure 4D ). In addition, using ProcartaPlex multiplex immunoassay, we also quantified other cytokines/chemokines produced in the reaction. Seventeen out of 27 detectable cytokines showed a similar pattern to IFNγ inhibition in WT mice that was reversed in lrp8^-/- splenocytes. These included effector function: IFNγ; stimulator function, IL18, GM-CSF, G-CSF; inflammatory cytokines: IL2, IL3, IL4, IL5, IL9, IL13, IL23, IL12p70; chemo attractant factors: MCP-1, MCP3, MIP-1α, CCL-5 (Rantes); and regulatory factors: IL6, IL-10 (P < 0.05). Representative data are shown in Supplement Figure 5 . These findings suggest that T-cell function is at least partially inhibited by apoE through the lrp8 receptor pathway.

To additionally investigate the role of apoE on melanoma tumor growth in vivo, 1x10⁴ WT or apoE^-/- B16 cells were injected into the right flank of WT, apoE^-/- and lrp8^-/-mice (n=9). The mice from each group were monitored for tumor growth and survival. The results show that targeting apoE in both the tumor cells and the host (apoE^-/- B16 cells injected into apoE^-/- C57/BL6 mice) results in delayed tumor growth ( Figure 5 ), and significant rejection of tumor cell inoculation with improved overall survival ( Figure 6 ). Tumor growth was also impaired when apoE^-/- B16 cells were injected into lrp8^-/- mice, suggesting apoE/LRP8 receptor engagement is important in mediating the apoE protective effect on tumor growth. Complete deletion of apoE in the tumor cells and in the host was most effective for inhibition of tumor growth, while apoE^-/- B16 cells injected into WT mice, or WT B16 cells injected into apoE^-/- mice or lrp8^-/- mice partially suppressed tumor growth compared to control. The suppressive effect observed on tumor growth in vivo with apoE suppression appears indirect as apoE KO cells proliferate normally.

To evaluate if the protective effect against tumor growth with apoE targeting is immune mediated, we harvested the first 3 mice from each group that grew tumors to 15mm for tumor immune profiling. Nanostring analysis of immune cell infiltrates and activation of immune signaling pathways revealed that the apoE ^-/- mice in which apoE ^-/- cells were inoculated, demonstrated the greatest number of immune cell infiltrates ( Figure 7A ) as well as the highest activation of immune signaling pathways ( Figure 7B ) based on the expression of signature marker gene transcripts when compared to the other groups. Wild type tumors in apoE ^-/- mice or WT mice receiving the apoE ^-/- cells demonstrated more activation of immune pathways and cellular infiltrates as determined by RNA transcripts than WT controls, but these effects were only partial and inferior to the immunity induced when apoE was abolished in the system (apoE^-/- cells in apoE^-/- B57/BL6 mice). The expression level of the gene transcripts of multiple activation markers for T cells ( Figure 8A ) and dendritic cells (DCs) ( Figure 8B ) were also compared within 6 tumor groups. Results showed that the markers for activation T cells including interleukin-2 receptor alpha chain (IL2RA, or CD25), CD69, CD8a, CD28, check point inhibitors PD-L1 and CTLA4, and T cell exhaustion marker Tim-3 were all significantly enhanced in the tumor from the apoE ^-/- mice in which apoE ^-/- cells were inoculated (apoE ^-/- /apoE ^-/- ) compared with control group (wt/wt), PD1 and Lag 3 were slightly increased, but they were not statistically significant (data not shown). CD28 was also upregulated in the tumor from the apoE ^-/- mice in which wt tumor cells were inoculated (wt/apoE ^-/- ) ( Figure 8A ). For DCs, the activation genes including CD40, CD70, CD80, CD83, CD86, CD11b and CD11c were all significantly increased in

apoE ^-/- /apoE^-/- group compared with controls. CD70 was also increased in wt/apoE ^-/- and apoE ^-/- /wt groups. ( Figure 8B ). These observations show that apoE secreted from the tumor or produced in the host impair immunity and establish the potent role that apoE plays in suppressing tumor immunity in the mouse melanoma model. Surprisingly, there was no upregulation of immune pathway scores nor enhanced immune cells scores in lrp8 ^-/- mice. These observations do not correlate with in vitro findings nor in vivo growth rates, but this may have been specific to the three mice sampled that developed tumors early in this group. To validate nanostring RNA transcript results in the apoE targeted group, we performed immunohistochemistry staining of the immune cell marker CD45 (lymphocyte common antigen) and CD3 (T cell marker) on the same tumor samples that we used for nanostring analysis. Results showed that the apoE ^-/- mice in which apoE ^-/- cells were inoculated, have significantly more CD45 ( Figures 9A, B ) and CD3 ( Figures 9C, D ) positive immune cell infiltrates than WT control. The other groups were not studied as these gross observations do not have the same objectivity or power of analysis as the nanostring assay.

In the mouse model, apoE appears to be an immune modulator critical for enabling tumor cell growth through suppression of immune activation. To determine if knocking out apoE in tumor cells induced immunogenicity, we vaccinated wild type and apoE^-/- mice ( Figure 10A ) or wild type and lrp8^-/- mice ( Figure 10B ) with WT B16 or apoE^-/- B16 tumor cells and CTLA4 Ab and then collected splenocytes from these vaccinated mice 6 days later. We then co-cultured the splenocytes with WT B16 or apoE^-/- B16 cells to evaluate IFNγ secretion as a marker of induced immunity. ApoE^-/- B16 cells induced robust immunity in whatever circumstance they were tested either as the primary immunogen or with re-stimulation of splenocytes ( Figures 10A,B ). These findings re-iterate the potent inhibitory effect of apoE on immune cell activation and present an opportunity to exploit this pathway to enable tumor immunity and cancer immunotherapy.

The potent immune-suppressive effects of ApoE in the melanoma mouse model suggests that APOE may be an important regulator of immunity in human melanoma. To evaluate its association with human melanoma, we analyzed TCGA melanoma datasets based on RNA-seq gene expression values (measured by RSEM algorithm) in 462 patient tumors. We evaluated expression and correlation with other checkpoints ( Figure 11 ), immune cell infiltrates ( Figure 12 ) and patient survival ( Figure 13 ). APOE is abundantly present, however, it did not correlate with PD-L1 or PD1 expression, two checkpoints expressed on tumors, but did positively correlate with APOC1 expression APOC1 was used as a positive control in this analysis ( Figure 11 ). APOE did not correlate with RNA-seq gene expression of T-cell, neutrophil and dendritic cells infiltrates whereas PD-L1 expression correlated with the presence of these three cell phenotypes ( Figure 12 ). Also, APOE did not correlate with survival at a 30% bifurcate gene analysis, whereas the expression of PD-L1 and PD1 both positively correlated with survival of cutaneous melanoma ( Figure 13 ). The high expression of checkpoints like PD-L1 and PD1 associated with cell infiltrates and survival curves, suggest that these genes are expressed in inflammatory tumors that have a better prognosis. APOE expression however appears to be independent of inflammatory phenotype in these tumors and may act as a separate and independent pathway in suppressing anti-tumor T-cell immunity.