"V体育官网入口" Fine-tuning the CAR spacer improves T-cell potency

CAR-PSCA T cells exhibit potent in vitro antitumor effects but fail to exert in vivo activity in a xenograft tumor model

To target the tumor-associated antigen (TAA) PSCA, which is overexpressed in a variety of solid tumors including prostate, pancreas, and colon cancer, we constructed a retroviral vector encoding a humanized, codon-optimized, second generation CAR with an IgG1-derived hinge-CH2CH3 spacer, a CD28 transmembrane and signaling domain, and the CD3ζ chain, which we entitled our prototype CAR [P1. CAR] (Fig.  1A). This transgenic molecule was efficiently and stably expressed on the surface of activated T cells (95. 9 ± 0. 6%, mean ± SE, n = 8; Fig.  1B), conferring cells with the ability to specifically kill PSCA-expressing target cells (K562-PSCA; 73. 1 ± 5. 9% and Capan-1; 72. 0 ± 11. 1% specific lysis, mean ± SE, n = 5, 40:1 E:T ratio) but not PSCA-negative targets such as K562 and 293T cells (19. 0 ± 2. 6% and 8. 4 ± 2. 0%, respectively). Non-transduced (NT) T cells produced only background levels of lysis (K562; 11. 1 ± 4. 1%, K562-PSCA; 27. 9 ± 7. 0%, 293T cells; 6 V体育2025版. 5 ± 2. 1% and Capan-1; 26. 9 ± 8. 9% specific lysis, mean ± SE, n = 5, 40:1 E:T ratio) (Fig.  1C). To evaluate the in vivo antitumor potential of these CAR T cells, we engrafted 6-week-old NSG mice with 5×106 Capan-1 cells subcutaneously (s. c. - right flank) and after 28 days, when the tumor had reached a volume of > 80 mm3, mice were treated with 10×106 P1. CAR T cells labeled with GFP/firefly luciferase (FFluc). However, despite CAR T-cell treatment, the tumor continued to increase in size at a rate similar to that observed in control (PBS) mice (Fig.  1D). .

To assess whether deficient CAR T-cell trafficking was responsible for this phenomenon, we evaluated T-cell migration by performing sequential luminescence imaging of animals treated with either control (GFP/FFluc) or P1. CAR T cells. As shown in Fig.  1E control T cells rapidly (within 24 h) localized to secondary lymphoid tissues such as the spleen and lymph nodes. In contrast, P1. CAR T cells failed to migrate to either the tumor or secondary lymphoid tissue VSports app下载. Instead the T cells were trapped in the lungs, where the luminescence signal progressively increased. To investigate the mechanism behind this “non-specific” expansion, we examined whether interactions between the IgG1-CH2CH3 spacer region of our P1. CAR and Fcγ receptor-expressing cells could be responsible for this phenomenon. 8-11 Thus, we cultured NT and P1. CAR T cells at a 1:1 ratio with human monocytes, macrophages and NK cells, all of which express different types of FcγRs (CD64—FcγRI, CD32—FcγRII, and CD16—FcγRIII) at varying intensities (Fig.  1F). As shown in Fig.  1G co-culture with monocytes and macrophages, which express CD64 and CD32, induced P1. CAR T-cell expansion and resulted in the elimination of monocytes and macrophages. However, this phenomenon was not observed in co-cultures with human NK cells, suggesting that this recognition was mediated through interaction with the FcγRs I and II and not CD16 (Fig.  1G).

Modification of the CH2CH3 spacer improves tumor localization

To abrogate FcγR recognition, we made two new CARs - M1. CAR and M2. CAR. In M1. CAR, we mutated the amino acids ELLG (aa233–236) and N (aa297) in the IgG1 CH2 region to PVA and Q, respectively. 9,12 To generate the M2. CAR, we substituted the hinge-CH2CH3 IgG1 framework for that of IgG2 (reported to have the lowest potential for interaction with both human13,14 and murine11 FcγR-expressing cells) and we additionally mutated aa297 (N) to Q (Fig.  2A). Subsequently, we investigated whether these modifications were sufficient to restore the migratory capacity of our CAR T cells. As shown in Fig.  2B, both the M1 and M2. CARs could be expressed at high levels on CD3/28-activated T cells (95. 3 ± 0. 8% and 91. 3 ± 1. 3%, respectively, mean ± SE, n = 8), enabling cells to specifically kill PSCA+ targets (72. 8 ± 12. 9% and 61 V体育官网. 5 ± 5. 4% specific lysis for M1. CAR and 75. 8 ± 5. 5% and 63. 2 ± 6. 1% for M2. CAR against K562-PSCA and Capan-1, respectively, mean ± SE, n = 5, 40:1 E:T ratio), with only background levels of killing against the control PSCA− targets (K562 and 293T) (Fig.  2C). To investigate whether our modifications mitigated FcγR recognition we co-cultured NT, P1, M1, and M2. CAR T cells with monocytes or macrophages (Fig.  2D) and after 3 d quantified residual cells by flow cytometry. As before, co-culture with P1. CAR T cells resulted in the elimination of macrophages/monocytes, while our M1 or M2. CAR and NT co-cultures all showed limited T-cell expansion and retention of macrophages/monocytes, suggesting that our modifications had successfully minimized FcγR recognition (Fig.  2D). However, while M1 and M2. CARs behaved similarly in vitro, our in vivo studies revealed that M2. CAR T cells were more efficiently able to mobilize from the lungs and localize at the tumor (Fig.  2E). Nevertheless, their antitumor activity was underwhelming (Fig.  2F)—for reasons unrelated to target antigen expression (Fig.  S5B), highlighting the need for further CAR optimization. .

CAR T-cell senescence

To determine whether CAR expression adversely impacted T-cell persistence, we initiated a series of in vitro studies to assess long-term T-cell expansion. Thus, we cultured P1.CAR T cells in media supplemented with IL2 (50 IU/mL) (without antigen or FcγR stimulation) and media and cytokines were replenished every 2–3 d for a total of 30 d. The gene expression profile, phenotype, and function of cells were assessed on days 10, 20, and 30 post-transduction. As shown in Fig. 3A, prolonged in vitro expansion did not impact cytolytic capacity, as measured in a short-term (4 h) chromium release assay with DU145 cells (57.0 ± 3.8%; day 10, 56.0 ± 7.0%; day 20, 54.7 ± 7.9%; day 30, n = 3, 40:1). However, when we performed a long-term (6 d) co-culture with DU145 cells (1:2—E:T) we observed an inverse correlation between antitumor activity and cell age (tumor cell fold expansion: 8.8 ± 1.5; day 10, 19.8 ± 1.5; day 20, 32.0 ± 7.2; day 30—mean ± SE, n = 4), which could be ascribed, at least in part, to decreased T-cell proliferation (T-cell fold expansion: 18.1 ± 2.4; day 10, 10.0 ± 1.1; day 20, 4.8 ± 0.6; day 30—mean ± SE, n = 4) (Fig. 3B). To explore this phenomenon further we compared the gene expression profile of P1.CAR T cells cultured for 20 or 30 d with that of cells maintained in culture for 10 d and found that genes associated with maintaining a naïve/central memory phenotype (e.g., CCR7, SELL, CD27, and CD28) were progressively downregulated, whereas those associated with a differentiated T-cell profile [e.g., EOMES, FAS ligand (FASLG), and Granzyme B (GZMB)] were upregulated (Fig. 3C). Indeed, Fig. 3D shows all the genes that were significantly upregulated or downregulated, clearly illustrating that prolonged culture induced cellular differentiation.

To confirm our microarray analyses and determine if our M1 and M2.CAR T cells exhibited an aging profile similar to that of our P1.CAR T cells, we performed phenotypic analyses to examine their memory profile (Fig. 3E, left panel and Fig. S1A). Although NT cells retained a naive profile over time, there was a progressive increase in effector memory T cells in all three CAR populations (P1, M1, and M2). Furthermore, although NT cells retained expression of both CD27 and CD28 over time, there was a progressive decline in expression of these molecules in all three CAR T-cell products (Fig. 3E, right panel and Fig. S1B], suggesting that CAR expression resulted in accelerated T-cell aging/differentiation.

V体育安卓版 - Tonic signaling is responsible for accelerated CAR T-cell aging

To investigate whether these phenotypic and functional changes could be related to spontaneous CAR signaling, we generated a truncated version of the P1.CAR lacking the intracellular signaling domains (ΔCAR; Fig. 4A), which we expressed on T cells (mean 89.2 ± 1.6% transduction, n = 7) (Fig. 4B). Next, we measured CD3ζ phosphorylation in the absence of cognate antigen stimulation and, as shown in Fig. 4C, all except the ΔCAR T cells exhibited evidence of CD3 signaling, which resulted in a chronic activation state characterized by persistently elevated CD25 levels both on CD8⁺ (Fig. 4D) and CD4⁺ (Fig. S2) T cell. Fig. 4D left panel shows CD25 expression on CD8⁺ T cells from a representative donor and right panel shows summary data (mean ± SE, n = 6). As expected, this tonic signal promoted cell cycle progression, characterized by transition from resting (G₀) state to G₁, S, and G₂/M (Fig. 4E). As a consequence, P1-, M1-, and M2-modified cells exponentially expanded in the absence of the antigenic stimulation (Fig. 4F) and non-specifically produced effector cytokines (Fig. 4G), implicating tonic signaling as the mechanism underlying the accelerated aging we had detected.

The CH2CH3 spacer is responsible for tonic T-cell signaling

To determine if removing the CH2CH3 region would abrogate tonic signaling, we deleted this region entirely (X2.CAR, Fig. 5A), modified activated T cells (87.1 ± 2.0% transduction, mean ± SE, n = 8) (Fig. 5B), and confirmed that these cells were unable to interact with monocytes or macrophages (Fig. S3A). We next monitored the activation status of ΔCAR (negative control), M2.CAR (positive control), and X2.CAR T cells using CD25 expression as a readout. As shown in Figs. 5C and S3B, X2.CAR T cells exhibited a profile similar to that of ΔCAR T cells with minimal CD25 expression over 30 d of culture. This quiescent state was confirmed by cell cycle analysis showing that the majority of X2.CAR T cells were in G₀ (mean 69.2 ± 8.4%, n = 3 - Fig. 5D). Consequently, non-specific expansion and cytokine production was low (Fig. 5E and G) and cells maintained an undifferentiated phenotype (Figs. 5F and S3C). Finally, we assessed the ability of X2.CAR T cells to kill PSCA⁺ targets. In a 4-h ⁵¹Cr release assay (Fig. 5H) transgenic cells exhibited potent antitumor activity against targets expressing high levels of PSCA (K562-PSCA and Capan-1; Fig. S3D). However, when cultured with target cells that expressed low PSCA levels (Fig. S3D), X2.CAR T cells demonstrated little/no tumor cell recognition (DU145—mean 20.7 ± 5.8% vs 57.5 ± 4.3%; X2 vs M2, CFPAC-1—mean 9.9 ± 1.5% vs 28.3 ± 4.0%; X2 vs M2, n = 5, 40:1) (Fig. 5H). Taken together this data suggests that although tonic signaling can be mitigated by removing the CH2CH3 region, deletion of this region can adversely affect antigen recognition and subsequent target lysis.

Inclusion of CH3 as a spacer restores cytolytic abilities without accelerating cell aging

To generate an effective CAR that retained its cytolytic capacity, we generated an additional vector with an intermediate length spacer comprising only the IgG2 hinge-CH3 domain (X₃2.CAR—mean 86.4 ± 2.2% transduction, n = 8; Fig. 6A and B). To first determine whether transgenic expression of this construct enabled T-cell recognition of targets expressing both high (K562-PSCA and Capan-1) and low (DU145 and CFPAC-1) PSCA levels, we performed a chromium release assay. As shown in Fig. 6C, inclusion of the CH3 region facilitated recognition of all PSCA-expressing target cells (K562-PSCA; mean 74.8 ± 2.8%, Capan-1; 68.8 ± 6.6%, DU145; 48.4 ± 5.2%, CFPAC-1; 19.6 ± 3.5%, n = 5, 40:1) with no non-specific target recognition. Furthermore, based on assessment of phenotype (Fig. 6D and E), cell cycle status (Fig. 6F), expansion (Fig. 6G), and cytokine production (Fig. 6H) this enhanced antitumor activity was not gained at the expense of non-specific T-cell activation. Finally, since this construct lacked the CH2 region, as expected there was no evidence of Fc–FcγR interaction (Fig. S4).

In vivo CAR antitumor activity

Based on our previous in vitro data (presented in Figs. 1, 2, 5, and 6), we predicted that in vivo: (i) P1. CAR T cells would be trapped in the lungs and rapidly eliminated, (ii) M1 and M2.CAR T cells would traffic to the tumor (and secondary lymph nodes) but would not control tumors due to cell senescence as a consequence of tonic signaling, (iii) X2.CAR T cells would effectively traffic to the tumor, persist in vivo and kill “high” PSCA-expressing tumor cells but fail to eliminate those expressing low levels of target antigen, whereas (iv) X₃2.CAR T cells would effectively traffic to the tumor, persist in vivo and control tumor growth, resulting in a survival benefit. To assess if this was indeed the case, NSG mice were engrafted s.c. with Capan-1 cells and when the tumor had reached a volume of 80 mm³, animals were administered i.v. with 10×10⁶ FFluc-CAR T cells (P1, M1, M2, X2, or X₃2). In vivo T-cell migration and proliferation was monitored by luminescence imaging, while tumor volume was measured by calipers. As shown in Figs. 7A and S5A, only M1, M2, X2, and X₃2 CAR T cells were able to escape the lungs, and persist at the tumor and secondary lymphoid organs (Fig. 7B and C), which resulted in delayed tumor growth. However, the antitumor activity mediated by X₃2.CAR T cells was significantly stronger than that exhibited by the other three constructs (Fig. 7D), resulting in a survival benefit in this cohort of animals (Fig. 7E). Finally, although all tumors eventually relapsed, in the X₃2.CAR T cell-treated group, the recurrent tumors were PSCA negative (Fig. S5B), illustrating that tumor immune escape was due to antigen modulation rather than CAR deficiency.

Donors and cell lines

Peripheral blood mononuclear cells (PBMCs) were obtained from healthy volunteers after informed consent on protocols approved by the Baylor College of Medicine Institutional Review Board. K562 (chronic erythroid leukemia cell line), 293T (human embryonic kidney cell line), Capan-1 (pancreatic cancer cell line), DU145 (prostate cancer cell line), and CFPAC-1 (pancreatic cancer cell line) were obtained from the American Type Culture Collection (Rockville, MD). Cells were maintained in a humidified atmosphere containing 5% carbon dioxide (CO₂) at 37 °C. K562 cells were maintained in RPMI-1640 media (GE Healthcare Life Sciences, Pittsburgh, PA), whereas 293T cells were maintained in Dulbecco's modified eagle medium (DMEM, GE Healthcare Life Sciences). Capan-1, DU145, and CFPAC-1 cells were maintained in Iscove's Modified Dulbecco's Medium (IMDM; Gibco BRL Life Technologies, Inc., Gaithersburg, MD). Capan-1 cells were grown in IMDM containing 20% heat-inactivated fetal bovine serum (FBS) (Hyclone, Waltham, MA) with 2 mM L-GlutaMAX (Gibco BRL Life Technologies, Inc.), whereas other cell lines were grown in their specific media containing 10% FBS with 2 mM L-GlutaMAX.

"VSports在线直播" Generation of retroviral constructs and retroviral transfection

Our original second generation CAR-PSCA has an IgG1-derived hinge-CH2CH3 spacer and CD28 and CD3ζ intracellular domains (P1.CAR).¹⁵ To generate our modified CARs, we synthesized the following DNA segments (Invitrogen, Grand Island, NY): (i) IgG1-hinge-CH2CH3 spacer region with mutations of ELLG (aa233–236) and N (aa297) to PVA and Q, respectively—M1.CAR; (ii) IgG2-hinge-CH2-CH3 with N297Q mutation—M2.CAR; (iii) IgG2-hinge—X2.CAR; and IgG2-hinge-CH3—X₃2.CAR. To generate the novel CAR constructs, the spacer sequences in the original P1.CAR construct were removed by enzymatic digestion (BamH1 and PflMI) and replaced with the new spacers. The γ-retroviral vectors encoding the fusion protein (GFP/FFluc) and the retroviral supernatant were generated as previously described.^32,33

Generation of CAR-modified T cells

To generate CAR T cells, 1×10⁶ PBMCs were plated in each well of a non-tissue culture-treated 24-well plate that had been pre-coated with OKT3 (1 mg/mL; Ortho Biotech, Inc., Bridgewater, NJ) and CD28 (1 mg/mL; Becton Dickinson & Co., Mountain View, CA). Cells were cultured in complete media (RPMI-1640 containing 45% Clicks medium (Irvine Scientific, Inc., Santa Ana, CA), 10% FBS, and 2 mM L-GlutaMAX), which was supplemented with recombinant human IL2 (50 U/mL, NIH, Bethesda, VA) on day 1. On day 3, retroviral supernatant was plated in a non-tissue culture-treated 24-well plate (1 mL/well) that had been pre-coated with a recombinant fibronectin fragment (FN CH-296; Retronectin; TAKARA BIO INC, Otsu, Japan), and centrifuged at 2,000g for 90 min. After removal of supernatant, OKT3/CD28-activated PBMCs (0.1×10⁶/mL) were resuspended in complete media supplemented with IL2 (100U/mL) and 2mL was added to each well of a 24-well plate, which was subsequently spun at 400g for 5 min, and then transferred to a 37 °C, 5% CO₂ incubator. Subsequently, cells were split and fed every 2–3 d with fresh media plus IL2 (50 U/mL). To track T-cell numbers overtime, viable cells were counted using trypan blue. To co-express CAR and GFP-firefly luciferase (GFP/FFluc) (for in vivo bioluminescence imaging), T cells were first modified to express the CAR, as previously described, and were transduced after 24 h to co-express GFP/FFluc using the same protocol.

Generation of K562 modified to express PSCA (V体育安卓版)

We synthesized PSCA (Invitrogen, Grand Island, NY),³⁴ which was incorporated into the pVITRO1-blasti-mcs vector (Invivogen, San Diego, CA) by enzymatic digestion (AgeI and NheI) and transfected into K562 using the GeneJuice® Transfection Reagent (EMD Millipore, Darmstadt, Germany). Transfected cells were selected and maintained in the presence of 10 ng/mL of Blasticidin (Invivogen, San Diego, CA).

FcγR-expressing cell preparation

Monocytes were isolated from PBMCs using human CD14 microbeads (MACS system; Miltenyi Biotec Inc., San Diego, CA). Macrophages were generated by culturing monocytes with 100 ng/mL GM-CSF for 7 d. NK cells were expanded by stimulating 5×10⁶ PBMCs with 5×10⁶ irradiated K562-mbIL15-41BBL^35,36 in 500 U/mL IL2 in G-Rex10 devices (Wilson Wolf Manufacturing, Minneapolis, MN)³⁷ followed by depletion of residual CD3⁺ cells using CD3 microbeads (MACS system; Miltenyi Biotec, Inc.).

Flow cytometry

^{"VSports最新版本" 51}Chromium-release assay

The cytotoxicity and specificity of engineered T cells were evaluated in a standard 4–6-h ⁵¹Cr-release assay, as previously described.¹⁵

Co-culture experiments

For co-culture experiments with FcγR-expressing cells, T cells and target cells were co-cultured at a 1:1 ratio in 2 mL of complete media in a 24-well plate for 3 d. Subsequently cells were harvested and stained with CD3, CD4, and CD8 (T cells) and CD33 (monocyte/macrophage) or CD16 (NK cells). For co-culture experiments with tumor cells, 5×10⁴ T cells were co-cultured with 1×10⁵ GFP/FFluc-DU145 cells in 4 mL of complete media in a 6-well plate for 6 d. To quantify cells by flow, CountBright™ Absolute Counting Beads (C36950; Invitrogen, Eugene, OR) were added (50 μL) and 7-AAD was added to exclude dead cells. Acquisition was halted at 5,000 beads.

Cytokine detection

To measure cytokine production 1×10⁶ T cells, which were maintained in culture for 10 d after transduction, were plated in a single well of a 24-well tissue culture plate with 2 mL of complete media and cultured for 24 h. Subsequently, supernatants were collected and stored at −80 °C. Cytokine levels were analyzed using MILLIPLEX MAP High Sensitivity Human Cytokine 13 Plex (Merck Millipore, Billerica, MA) according to manufacturer's instructions.

Microarray analysis

Total RNA was extracted from T cells cultured for different culture period using the RNeasy Mini kit (QIAGEN, Valencia, CA) and quantified using the NanoDrop 2000 (Thermo Fisher Scientific, Inc., Waltham, MA). RNA expression profiling was performed using the GeneChip PrimeView Human Gene Expression Array (Affymetrix, Inc., Santa Clara, CA) by Genome Exploration, USA (Memphis, TN).

In vivo study

NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ mice (NSG mice, 6–8 weeks old, The Jackson Laboratory) were engrafted s.c. (right flank) with Capan-1 cells (5×10⁶/animal) and once the tumors were > 80 mm³ (∼ day 28) the animals were treated with 10×10⁶ engineered GFP/FFluc-T cells i.v. Tumor size was measured by calipers and tumor volume was calculated formulas follows: tumor volume (mm³) = length × width × width/2. T-cell migration and distribution were evaluated by bioluminescence imaging recorded two times per week using a Lumina IVIS imaging system (Caliper Life Sciences, Inc., Hopkinton, MA), and analyzed by Living image software (Caliper Life Sciences, Inc.). To assess PSCA expression, mice were sacrificed, tumors were dissected, and single cell suspensions were prepared, as previously published.³⁸ Briefly, extracted tumors were minced and dissociated by incubating the cells with 200 U/mL collagenase IV (Gibco BRL Life Technologies, Inc.) at 37 °C in a water bath for 2 h. Every 20 mins, cells were vortexed for 1 min and at the end of the incubation period tissue debris and dead cells were removed by density centrifugation using Lymphoprep (Axis-Shield, Oslo, Norway).

Statistics

Statistical analysis was performed using Graphpad Prism 6 software (GraphPad Software, Inc., La Jolla, CA). Two-way ANOVA was used for ⁵¹Cr-release assay and in vivo tumor growth. One-way ANOVA was used in comparing the cytolytic function of T cells cultured for 10, 20, or 30 d. To analyze microarray results for different T cells of different ages generated from three independent donors, we used FDR-corrected ANOVA. All other experiments were analyzed using an unpaired two-tailed t-test.

Study approval

For human cells, healthy volunteers gave written informed consent according to protocols approved by the Baylor College of Medicine Institutional Review Board and performed in accordance with the guidelines established by the Declaration of Helsinki. All animal studies were performed under a protocol approved by the Animal Research Committee of Baylor College of Medicine.