Chimeric antigen receptor (CAR) T-cell therapies have proven to be effective in treating hematologic malignancies but demonstrate only marginal efficacy in eradicating solid tumors. Although several mechanisms can account for these differences, a major cause is thought to derive from CAR T-cell exhaustion, where chronic exposure to tumor antigen can activate feedback pathways that suppress CAR T-cell cytotoxicity. We describe here a strategy to reverse this CAR T-cell exhaustion using a universal anti-fluorescein CAR that concurrently serves as (i) a cancer recognition receptor that enables engagement of multiple cancer cell clones upon addition of a cocktail of bispecific fluorescein-linked tumor-targeting ligands, and (ii) a drug-internalizing receptor that mediates uptake of a CAR T-cell activator comprised of fluorescein linked to an immune stimulant. By attaching a Toll-like receptor 7 agonist (TLR7–1A) to fluorescein, we enable the anti-fluorescein CAR to bind and internalize TLR7–1A, leading to both downregulation of exhaustion markers (i.e., PD-1, TIM3, LAG3) and reactivation of exhausted CAR-T cells without causing the toxicities commonly associated with systemic administration of TLR7 agonists. The resulting rejuvenated CAR-T cells are observed to regress otherwise refractory solid tumors. Moreover, because no other immune cells are altered by this treatment, the data demonstrate that the exhaustion state of the CAR-T cells constitutes a major property that determines the efficacies of CAR T-cell therapies in solid tumors.

Implications:

A novel strategy for rejuvenating exhausted CAR-T cells is described previously that promotes downregulation of exhaustion markers and renewed eradication of cancer cells in a tumor mass.

This article is featured in Highlights of This Issue, p. 671

Chimeric antigen receptor (CAR) T-cell therapies constitute a novel and exciting approach to the treatment of cancers (1–3). Anti-CD19 CAR T-cell therapies, which are currently the most successful CAR T-cell products in the clinic, have achieved 81% and 51% complete remissions in the treatment of acute lymphoblastic leukemias and large B-cell lymphomas, respectively (4–7). A related anti–B-cell maturation antigen CAR T-cell therapy has reported similar response rates of approximately 80% in phase I/II clinical trials of multiple myeloma (8). However, despite analogous efforts to use CAR-T cells to treat solid tumors (9–12), clinical outcomes have been less encouraging, with the best response rates reaching only approximately 30% in phase II clinical trials of melanomas (13). Clearly, obstacles to CAR T-cell efficacies in solid tumors exist that are not encountered in treatment of liquid tumors.

Although characterization of the causes of CAR T-cell failures in solid tumor is still in its infancy, several hypotheses have been offered, including (i) natural selection for antigen-deficient cancer cells during therapy (14, 15); (ii) inefficient infiltration and expansion of CAR-T cells in solid tumors (16–20); (iii) suppression of CAR T-cell cytotoxicity by other immune cells [e.g., regulatory T cells, myeloid-derived suppressor cells, and tumor-associated macrophages (TAM)] and cancer cells in the tumor microenvironment (21–26), and (iv) activation of natural feedback signals that suppress a T-cell's response to chronically presented antigens (13, 27–29). Although the first of these mechanisms can conceivably be prevented by expansion of the repertoire of tumor antigens recognized by the CAR-T cells (30), solutions to the latter three will likely require strategic modulation of CAR T-cell properties in vivo.

In an effort to identify a methodology for continuous and selective manipulation of CAR T-cell properties in vivo, we have explored the possibility of using an engineered CAR to deliver drugs solely into CAR-expressing T cells. For this purpose, we have exploited a universal CAR T-cell design to create a single CAR that can recognize any tumor antigen for which a ligand can be synthesized (31–33). In this approach, a CAR containing the usual cytoplasmic activation domains (e.g., CD3ζ, 4–1BB and CD28) is fused to an extracellular single-chain fragment variable designed to recognize solely fluorescein (33–36). Addition of a bispecific adapter comprised of fluorescein linked to a tumor-specific ligand then creates a molecular bridge between the CAR T-cell and tumor cell, leading to formation of an immunologic synapse that induces both destruction of the cancer cell and proliferation of the CAR T-cell (Fig. 1A). With the increasing availability of bispecific adapters that can form such bridges between anti-fluorescein CAR-T cells and different cancer-specific antigens (33), this universal CAR T-cell strategy has been shown to facilitate engagement and destruction of multiple orthogonal cancer cell clones in antigenically heterogeneous tumors, leading to eradication of the tumors without selection for antigen-deficient clones (33, 34, 37). In view of the financial limitations to creating a series of different CAR-T cells that recognize the diversity of tumor antigens present in the heterogeneous cancer cell clones characteristic of almost all solid tumors (29, 31, 38), the alternative approach of using a single anti-fluorescein CAR in combination with multiple low molecular weight bispecific adapters has been attracting increased attention (31–33, 39).

Figure 1.

Use of an anti-fluorescein CAR and a fluorescein–TLR7a conjugate to selectively activate CAR-T cells in vivo. A, Upon administration of a fluorescein–folate bispecific adaptor to mice bearing a folate receptor-expressing tumor, the bispecific adaptor forms a bridge between the anti-fluorescein CAR and the folate receptor-expressing cancer cell, promoting killing of the cancer cell and proliferation of the CAR T-cell. B, Infiltration of a fluorescein-linked TLR7 agonist into the same tumor can enable binding of the conjugate to the same anti-fluorescein CAR on the T-cell. Subsequent internalization of the CAR by receptor-mediated endocytosis then allows the drug to modulate CAR T-cell properties without affecting other (CAR-negative) cells in the body.

Figure 1.

Use of an anti-fluorescein CAR and a fluorescein–TLR7a conjugate to selectively activate CAR-T cells in vivo. A, Upon administration of a fluorescein–folate bispecific adaptor to mice bearing a folate receptor-expressing tumor, the bispecific adaptor forms a bridge between the anti-fluorescein CAR and the folate receptor-expressing cancer cell, promoting killing of the cancer cell and proliferation of the CAR T-cell. B, Infiltration of a fluorescein-linked TLR7 agonist into the same tumor can enable binding of the conjugate to the same anti-fluorescein CAR on the T-cell. Subsequent internalization of the CAR by receptor-mediated endocytosis then allows the drug to modulate CAR T-cell properties without affecting other (CAR-negative) cells in the body.

Close modal

In our work, we test the ability of an anti-fluorescein CAR to mediate binding, endocytosis, and intracellular delivery of a variety of fluorescein-linked drugs (Fig. 1B). We first demonstrate that addition of a fluorescein-linked fluorescent dye (fluorescein–Alexa Fluor 647) results in rapid internalization of the conjugate, demonstrating that the anti-fluorescein CAR can be exploited to deliver an attached molecule into CAR-T cells. We then show that similar incubation of a fluorescein-conjugated Toll-like receptor 7 agonist (fluorescein–TLR7–1A) leads to endocytosis of the agonist followed by potent activation of the exhausted CAR-T cells in vitro. Finally, we document that systemic administration of the same fluorescein–TLR7–1A conjugate to tumor-bearing mice reduces many immunosuppressive markers in solid tumors and converts an otherwise refractory solid tumor to a CAR T-cell–responsive tumor.

Synthesis fluorescein–drug conjugates

The synthesis of fluorescein–PEG3–Alexa Fluor 647 was performed as shown in Supplementary Scheme S1. FITC (Sigma-Aldrich) was added dropwise to a solution of amino-PEG3-amine (BroadPharm; 3 equiv) and DIPEA (Sigma-Aldrich; 5 equiv) in DMSO (Sigma-Aldrich). The solution was stirred at room temperature (RT) for 1 hour to give FITC–PEG3–amine. The resulting product was purified by preparative reverse-phase high-performance liquid chromatography (yield of 91.9%) with a mobile phase gradient consisting of 20 mmol/L ammonium acetate buffer and 0% to 100% acetonitrile over 30 minutes (xTerra C18; Waters; 10 μmol/L; 19 × 250 mm). For the synthesis of FITC–PEG3–Alexa Fluor 647, Alexa Fluor 647 NHS ester (Thermo Fisher Scientific), FITC–PEG3–amine (∼2 equiv) and DIPEA (∼5 equiv) were dissolved in DMSO, and the solution was stirred at RT for 1 hour. The product was then purified using the conditions described above (yield of 90.2%).

The synthesis of FITC–PEG3–NIR dye was performed as shown in Supplementary Scheme S2. Near-infrared fluorescent (NIR) dye (40), HATU (Sigma-Aldrich; ∼1 equiv), and DIPEA (∼5 equiv) were dissolved in DMSO and stirred for 25 minutes, followed by addition of FITC–PEG3–amine (∼1 equiv). The reaction was stirred at RT overnight and the product was purified as described above (yield of 71%).

The synthesis of FITC–TLR7–1A was performed as shown in Supplementary Scheme S3.

Cancer cell lines

MDA-MB-231 (NCI-DTP Cat# MDAMB-231, RRID: CVCL_0062) and KB (ATCC Cat# CCL-17, RRID: CVCL_0372) cells were obtained from the ATCC. Authentication of MDA-MB-231 was carried out by short-tandem repeat analysis based on the ATCC. Folic acid–free RPMI-1640 (Gibco) containing 10% heat-inactivated FBS and 1% penicillin–streptomycin was used to culture both cell lines. To obtain stable mCherry-expressing MDA-MB-231 cells, MDA-MB-231 cells were firstly transduced with a lentiviral vector, pLv-NLS-mCherry-puro (Vector Builder) and then selected for positive clones in puromycin-containing media. All cells were maintained in 5% CO2 at 37°C and were tested regularly for contamination of Mycoplasma.

Generation of human anti-fluorescein CAR-T cells

Human peripheral blood mononuclear cells were isolated from fresh human peripheral blood samples obtained from a diversity of healthy donors by Ficoll (GE Healthcare Lifesciences) density gradient centrifugation. CD3+ T cells were collected and enriched using an EasySep Human T-Cell Isolation Kit (STEM CELL Technologies) and then cultured in TexMACS medium (Miltenyi Biotech) containing 1% penicillin and streptomycin sulfate plus 2% human serum (Valley Biomedical) in the presence of human IL-2 (100 IU/mL, Miltenyi Biotech). Human T cells were counted every 2–3 days and maintained at 0.5 × 106 cells/mL. Human anti-fluorescein CAR-T cells were generated using a lentiviral vector protocol as previously described (33, 34). The studies were conducted in accordance with Belmont Report and approved by Purdue Institutional Review Board. Written informed consent was obtained from each donor.

Analysis of expression of TLR7 in human T cells

To determine whether TLR7 is expressed in primary human T cells, freshly isolated CD3+ T cells were fixed and permeabilized according to the manufacturer's instructions (Intracellular Flow Cytometry Staining Protocol, BioLegend) and stained with Alexa Flour 488–anti-TLR7 (R&D Systems). Cells were then washed 2 times with Intracellular Staining Permeabilization Wash Buffer (BioLegend) and resuspended in 1 × PBS before analysis by flow cytometry.

Analysis of fluorescein–NIR dye binding to anti-fluorescein CAR-T cells

Anti-fluorescein CAR-T cells (or MDA-MB-231 and KB cells as negative controls) were incubated with fluorescein–NIR dye (10 nmol/L) for 1 hour at RT prior in the presence or absence of 1,000-fold excess sodium fluorescein (10 μmol/L). Cells were then washed 3 times with PBS and NIR dye fluorescence was analyzed by flow cytometry.

Stimulation of CAR T-cell exhaustion and reversal of exhaustion in vitro

To induce CAR T-cell exhaustion, anti-fluorescein CAR-T cells (104/well) were added to folate receptor–expressing mCherry+ MDA-MB-231 cells (104/well) in folic acid–free RPMI-1640 medium, after which CAR T-cell–mediated killing of the mCherry+ MDA-MB-231 cells was initiated by addition of 10 nmol/L fluorescein–folate. Then, every 12 hours thereafter the anti-fluorescein CAR-T cells were transferred into fresh plate of mCherry+ MDA-MB-231 cells (104/well) to assure continuous exposure to tumor antigen. A fraction of the anti-fluorescein CAR-T cells were harvested after 12 hours (1st Round), 24 hours (2nd Round), and 36 hours (3rd Round) to analyze expression of exhaustion markers, PD-1 (BioLegend Cat# 329906, RRID:AB_940483), TIM3 (BioLegend Cat# 345014, RRID:AB_2561720), and LAG3 (BioLegend Cat# 369212, RRID:AB_2728373) by flow cytometry. The number of live mCherry+ (MDA-MB-231) cells were counted by Incucyte S3 every 4 hours and used to calculate the cancer cell killing efficiency. CAR-T cells were considered exhausted when they simultaneously expressed PD-1, TIM3, and LAG3.

For assessment of exhausted CAR T rejuvenation, at the beginning of the third round of exhaustion conditioning (see Fig. 4A), the combined CAR T-cell plus MDA-MB-231 cell culture was incubated overnight with the desired rejuvenating compounds at concentrations ranging from 0.01 to 100 nmol/L. CAR-T cells were then evaluated for rejuvenation by quantitating their cancer cell killing efficiency and analyzing their expression of PD-1, TIM3, and LAG3.

Comparison of CAR T-cell eradication of solid KB cell tumors in the presence and absence of fluorescein–TLR7–1A conjugate

Eight to 10-weeks-old NSG mice (The Jackson Laboratory strain No. 005557) were transferred upon arrival to folic acid-deficient diet (Envigo) to lower their serum folic acid concentrations to levels similar to those in humans and mice in the wild (41). One week later, the mice were implanted subcutaneously with approximately 106 KB cells and tumors were allowed to grow until they reached 30–50 mm3. Mice were then injected intravenously with 8 × 106 anti-fluorescein CAR-T cells, followed by injections with 500 nmol/kg fluorescein–folate both 4 and 24 hours after CAR T-cell infusion. Four days after CAR T-cell infusion, the CAR T-cell–treated mice were divided into a rejuvenation group (fluorescein–TLR7–1A treated) and a control group (saline treated). The fluorescein–TLR7–1A–treated cohort received 500 nmol/kg of fluorescein–TLR7–1A 4 times/week, whereas the control group received an equal volume of saline on the same schedule. Tumor volume was concurrently measured with calipers using the formula (l × w2)/2 (“l” being the largest length across the tumor and “w” being the dimension perpendicular to the longest transept).

Mice were sacrificed, and derived tumor fragments were dissociated using a human tumor dissociation kit (Miltenyi Biotec) when the tumors in the fluorescein–TLR7–1A–treated group decreased to around 50 mm3. The resulting single-cell suspensions were stained for human CD3 (BioLegend Cat# 317342, RRID:AB_2563410), PD-1 (BioLegend Cat# 329906, RRID:AB_940483), TIM3 (BioLegend Cat# 345014, RRID:AB_2561720), mouse CD11b (BioLegend Cat# 101208, RRID:AB_312791), F4/80 (BioLegend Cat# 123118, RRID:AB_893477), M2 macrophage marker CD206 (BioLegend Cat# 141708, RRID:AB_10900231), and M1 macrophage marker CD86 (BioLegend Cat# 105013, RRID:AB_439782). All the samples were then analyzed by flow cytometry. All animal care and usages were followed by NIH guidelines, and all experimental protocols were approved by the Purdue Animal Care and Use Committee.

Statistical analysis

All experiments were independently repeated at least three times. Data are expressed as mean ± SD or mean ± SEM (for in vivo studies). Differences of the data were analyzed by two-way ANOVA as indicated in the figure legends. GraphPad Prism software (GraphPad Prism, RRID:SCR_002798) were used to carry out all statistical analyses and data graphing, respectively.

Data availability statement

The data generated in this study are available within the article and its Supplementary Data Files.

Analysis of fluorescein conjugates binding and uptake by anti-fluorescein CAR-T cells

The objective of this work was to determine whether a CAR could be exploited to target an immunomodulatory drug selectively to CAR-T cells following intravenous injection in vivo. As an initial study, we synthesized fluorescein conjugates of two structurally unrelated NIR dyes (and Alexa Fluor 647), and after incubating the conjugates with anti-fluorescein CAR-T cells explored whether the conjugates were able to bind and enter the engineered T cells. As shown in Fig. 2; Supplementary Fig. S2, the fluorescein–NIR dye conjugate did indeed bind to anti-fluorescein CAR-T cells (Fig. 2A, green line), but not to cells lacking the anti-fluorescein CAR (e.g., MDA-MB-231 or KB cancer cells; Fig. 2B). That this binding was CAR-mediated could be demonstrated by quantitative blockade of its uptake when the conjugate was administered in the presence of 1,000-fold excess-free fluorescein (Fig. 2A, magenta). To assure that the above binding indeed resulted in internalization of the fluorescein-linked dye, we examined the cells by confocal microscopy and observed that the cells displayed punctate Alexa Fluor 647 fluorescence throughout the CAR T-cell's cytoplasm. We further noted that incubation of the same cells at 0°C (i.e., a nonpermissive temperature for receptor-mediated endocytosis) yielded only plasma membrane–localized fluorescence, with no fluorescence inside the CAR-T cells, even though subsequent raising of the temperature to 37°C allowed internalization of the same fluorescent conjugate. Because similar data were obtained with the fluorescein–NIR dye conjugate, we conclude that the anti-fluorescein CAR can mediate internalization of fluorescein-linked molecules by CAR-mediated endocytosis.

Figure 2.

Fluorescein-conjugated fluorescent dyes bind only to anti-fluorescein CAR-expressing T cells. A, Demonstration that a fluorescein–NIR dye conjugate (either fluorescein–NIR dye or fluorescein–Alexa Fluor 647) binds to anti-fluorescein CAR-T cells in a manner that can be quantitatively blocked by addition of 1,000-fold excess fluorescein. Flow cytometry was performed on anti-fluorescein CAR-T cells in the absence (blue histogram) or presence (green histogram) of the fluorescein–NIR dye conjugate (10 nmol/L), or in the presence of both fluorescein–dye conjugate and 1,000-fold excess fluorescein (magenta histogram). B, Demonstration that CAR-negative cell lines, MDA-MB-231 and KB cells, express no binding sites for fluorescein–dye conjugates (10 nmol/L). C and D, Confocal microscopy evaluation of internalization of fluorescein–Alexa Fluor 647 conjugate by CAR-T cells incubated first for 1 hour at 4°C (C) and then again after transferring the cells to 37°C for 4 hours (D); bar, 10 μm.

Figure 2.

Fluorescein-conjugated fluorescent dyes bind only to anti-fluorescein CAR-expressing T cells. A, Demonstration that a fluorescein–NIR dye conjugate (either fluorescein–NIR dye or fluorescein–Alexa Fluor 647) binds to anti-fluorescein CAR-T cells in a manner that can be quantitatively blocked by addition of 1,000-fold excess fluorescein. Flow cytometry was performed on anti-fluorescein CAR-T cells in the absence (blue histogram) or presence (green histogram) of the fluorescein–NIR dye conjugate (10 nmol/L), or in the presence of both fluorescein–dye conjugate and 1,000-fold excess fluorescein (magenta histogram). B, Demonstration that CAR-negative cell lines, MDA-MB-231 and KB cells, express no binding sites for fluorescein–dye conjugates (10 nmol/L). C and D, Confocal microscopy evaluation of internalization of fluorescein–Alexa Fluor 647 conjugate by CAR-T cells incubated first for 1 hour at 4°C (C) and then again after transferring the cells to 37°C for 4 hours (D); bar, 10 μm.

Close modal

Identification of immune modulators for activation of human T cells

Because a major roadblock in the use of CAR T-cell therapies for treatment of solid tumors arises when the CAR-T cells become exhausted, we next screened multiple T-cell modulating drugs for their abilities to activate CD3+ human T cells, including PI3K inhibitors, Toll-like receptor agonists, and thalidomide analogues, etc. (42). These screens revealed that TLR7 agonists induce the most pronounced

T-cell activation, as assessed by expression of activation markers CD69, CD25, IFNγ, and TNFα, whereas the thalidomide analogue, CC-220, induces the least. The screens further established that a recently disclosed TLR7 agonist, termed TLR7–54, was the one most potent TLR7 agonist examined. However, because TLR7–54 contained no chemical moiety that would enable its conjugation to fluorescein, we elected to derivatize TLR7–54 with a reactive functional group, as shown in Supplementary Scheme S3. We then re-examined the ability of the hydroxymethyl-modified TLR7–54 (now termed TLR7–1A) to activate human T cells. As shown in Fig. 3AD; Supplementary Fig. S3A–S3D, incubation of human CD3+ T cells with either TLR7–54 or its hydroxymethyl derivative (TLR7–1A) promoted a gradual increase in expression of activation markers (CD69, CD25, IFNγ, and TNFα) followed by their decrease at higher agonist concentrations, that is, consistent with the known bell-shaped response curves of T cells to TLR7–1A. Surprisingly, instead of compromising the potency of TLR7–54, derivatization of TLR7–54 with a hydroymethyl moiety enhanced its ability to activate T cells. Taken together, these data suggest that TLR7–1A constitutes an excellent drug for use in enhancing activation of human T cells.

Figure 3.

Activation of human CD3+ T cells upon administration of different concentrations either TLR7–54 or TLR7–1A agonist. Isolated human peripheral blood CD3+ T cells were stimulated with 3 μg/mL anti-CD3+ mAb (Thermo Fisher Scientific Cat# MA1–10175, RRID:AB_11157849) in the absence or presence of increasing concentrations of TLR7–54 or TLR7–1A agonist, after which the percentage of increase in activation markers, CD69 (A) and CD25 (B), was analyzed by flow cytometry. IFN-γ (C) and TNF-α (D) were also quantified in the 24-hour cell-free supernatants of the above stimulated T cells by ELISA. Data shown report the increase in these parameters over baseline levels [vehicle (DMSO) treated]. All data were plotted as mean ± SD, n = 3. Where error bars appear to be missing, they are present but smaller than the symbols.

Figure 3.

Activation of human CD3+ T cells upon administration of different concentrations either TLR7–54 or TLR7–1A agonist. Isolated human peripheral blood CD3+ T cells were stimulated with 3 μg/mL anti-CD3+ mAb (Thermo Fisher Scientific Cat# MA1–10175, RRID:AB_11157849) in the absence or presence of increasing concentrations of TLR7–54 or TLR7–1A agonist, after which the percentage of increase in activation markers, CD69 (A) and CD25 (B), was analyzed by flow cytometry. IFN-γ (C) and TNF-α (D) were also quantified in the 24-hour cell-free supernatants of the above stimulated T cells by ELISA. Data shown report the increase in these parameters over baseline levels [vehicle (DMSO) treated]. All data were plotted as mean ± SD, n = 3. Where error bars appear to be missing, they are present but smaller than the symbols.

Close modal

Evaluation of the ability of fluorescein–TLR7–1A to rejuvenate exhausted CAR-T cells in vitro

Next, to determine whether an anti-fluorescein CAR might be exploited to deliver a TLR7 agonist into exhausted CAR-T cells in vitro, we required a method to induce CAR T-cell exhaustion. As shown Fig. 4A, we found that exhaustion of anti-fluorescein CAR-T cells could be induced in vitro by serially transferring the CAR-T cells into fresh MDA-MB-231 cells every 12 hours. Then, after each around of cancer cell exposure (i.e., every 12 hours), we used Incucyte S3 to quantitate the CAR-T cells' killing efficiency by counting the number of mCherry+ (MDA-MB-231) cells remaining on the plate. As shown Fig. 4B, the killing efficiency of CAR-T cells decreased from approximately 70% to approximately 30% after the second round of cancer cell exposure and then further declined to approximately 15% after the 3rd round of exposure. Moreover, in concordance with these results, concurrent expression of the three major exhaustion markers (PD-1+TIM3+LAG3+) also increased with successive rounds of cancer cell exposure, reaching approximately 20% after the 3rd round (Fig. 4C).

Figure 4.

Effect of CAR T-cell targeted and nontargeted TLR7–1A agonists on CAR T-cell exhaustion in vitro. A, Protocol for induction of CAR T-cell exhaustion involving serial transfer of CAR-T cells every 12 hours to fresh MDA-MB-231 human breast cancer cells in culture. B and C, Evidence for CAR T-cell exhaustion is established by monitoring (B) a decrease in the ability of the anti-fluorescein CAR-T cells to kill MDA-MB-231 cells in vitro, and (C) an increase in expression of CAR T-cell exhaustion markers (PD-1+TIM3+LAG3+). A representative example of our gating strategy for PD-1+TIM3+LAG3+ CAR-T cells is shown in Supplementary Fig. S5. Rejuvenation of the above exhausted anti-fluorescein CAR-T cells by incubation with either targeted or nontargeted TLR7–1A is demonstrated by (D), a return of their ability to kill MDA-MB-231 cells in culture, and (E) a decrease in expression of their cell surface exhaustion markers. As shown in B, the exhausted CAR-T cells used in D were capable of killing only approximately 10% of the cancer cells before the indicated stimulation. More detailed killing efficiency data are shown in Supplementary Fig. S4. Data shown represent the change in these markers over baseline levels [vehicle (DMSO) treated]. All data were plotted as mean ± SD, n = 3. Where error bars appear to be missing, they are present but smaller than the symbols.

Figure 4.

Effect of CAR T-cell targeted and nontargeted TLR7–1A agonists on CAR T-cell exhaustion in vitro. A, Protocol for induction of CAR T-cell exhaustion involving serial transfer of CAR-T cells every 12 hours to fresh MDA-MB-231 human breast cancer cells in culture. B and C, Evidence for CAR T-cell exhaustion is established by monitoring (B) a decrease in the ability of the anti-fluorescein CAR-T cells to kill MDA-MB-231 cells in vitro, and (C) an increase in expression of CAR T-cell exhaustion markers (PD-1+TIM3+LAG3+). A representative example of our gating strategy for PD-1+TIM3+LAG3+ CAR-T cells is shown in Supplementary Fig. S5. Rejuvenation of the above exhausted anti-fluorescein CAR-T cells by incubation with either targeted or nontargeted TLR7–1A is demonstrated by (D), a return of their ability to kill MDA-MB-231 cells in culture, and (E) a decrease in expression of their cell surface exhaustion markers. As shown in B, the exhausted CAR-T cells used in D were capable of killing only approximately 10% of the cancer cells before the indicated stimulation. More detailed killing efficiency data are shown in Supplementary Fig. S4. Data shown represent the change in these markers over baseline levels [vehicle (DMSO) treated]. All data were plotted as mean ± SD, n = 3. Where error bars appear to be missing, they are present but smaller than the symbols.

Close modal

With this exhaustion model established, we next examined whether the ability of exhausted CAR-T cells to kill cancer cells might be restored by stimulation of the CAR-T cells with either free or fluorescein-conjugated TLR7–1A. As shown in Fig. 4D; Supplementary Fig. S4A, both the nontargeted and CAR-targeted TLR7 agonists were able to increase killing of exhausted CAR-T cells at low TLR7 agonist concentrations, but consistent with the bell-shaped concentration dependence of TLR7 stimulation (43), this enhancement of killing efficiency gradually converted to a suppression at higher TLR7–1A concentrations. As anticipated, the least active immune stimulant tested (CC-220) also exerted the smallest effect on CAR T-cell reactivation. Not surprisingly, the change in exhaustion markers also exhibited a bell-shaped concentration dependence (Fig. 4E; Supplementary Fig. S4B) with an initial inhibition followed by an eventual augmentation. An example of flow cytometry dot pot showing the gating of the exhaustion markers is shown in Supplementary Fig. S5. In conclusion, both nontargeted and CAR T-cell–targeted TLR7 agonists can reverse CAR T-cell exhaustion and enhance CAR T-cell killing when administered at optimal TLR7–1A concentrations.

Evaluation of the ability of fluorescein–TLR7–1A to rejuvenate exhausted CAR-T cells in vivo

To evaluate the ability of fluorescein–TLR7–1A conjugate to reverse CAR T-cell exhaustion in vivo, we required a murine model of T-cell exhaustion where both induction of exhaustion markers and loss of CAR T-cell cytotoxicity would be prominent. Evaluation of several human solid tumor xenografts in NSG mice revealed that human KB cell tumors, which express high levels of folate receptor α, are far less responsive to an anti-fluorescein CAR T-cell therapy than other folate receptor–positive tumors examined. Thus, as shown in Fig. 5A, MDA-MB-231 tumors gradually decrease in size upon treatment with fluorescein–folate plus anti-fluorescein CAR-T cells, whereas KB cell tumors continuously expand. That these very different responses to CAR T-cell therapy do not derive from intrinsically different growth rates in NSG mice can be demonstrated by their nearly identical growth curves in untreated NSG mice. To begin to assess the cellular mechanism underpinning this difference in resistance to CAR T-cell therapy, we performed a cursory comparison of the cellular compositions of the KB and MDA-MB-231 tumor with CAR T-cell therapy. As seen in Fig. 5B, significantly fewer CD3+ T cells were found to infiltrate KB tumors than MB-MDA-231 tumors, and those CD3+ T cells that did infiltrate the KB tumors expressed dramatically more exhaustion markers than those that accumulated in the MB-MDA-231 tumors (Fig. 5C). Although the markers on the TAMs were somewhat more inflammatory in KB tumors than MB-MDA-231 tumors (Fig. 5D), the T cells were much more evenly distributed in the MB-MDA-231 cells than in the KB cells (Fig. 5E). Because sequestering of T cells into pools constitutes a hallmark of immunologically cold tumors, these data collectively demonstrate that the immunologic environment is more suppressed in KB than MDA-MB-231 tumors.

Figure 5.

Effect of anti-fluorescein CAR T-cell therapy on the growth and immunologic properties of MDA-MB-231 and KB tumors. NSG mice were implanted with 4 million MDA-MB-231 cells and 1 million KB cells on separate flanks of the same mice, and infused 2 weeks later with either saline or 8 ×106 anti-fluorescein CAR-T cells (i.e., when MDA-MB-231 and KB tumors reached approximately 160 or 80 mm3, respectively; that is, to accommodate their different growth rates). Four and 24 hours after CAR T-cell infusion, all mice were intravenously injected with 500 nmol/kg fluorescein–folate and this injection was repeated once/week thereafter. KB tumor cells express approximately 2 million folate receptors/cell whereas MDA-MB-231 cells express approximately 1.6 million folate receptors/cell. A, Tumor volumes were analyzed in cohorts either left untreated (dashed lines) or treated with both anti-fluorescein CAR-T cells plus fluorescein–folate bispecific adapter (solid lines). On day 18, tumors were resected and both cancer and stromal cells were released using a tumor dissociation kit prior xenografts after treatment to analysis by flow cytometry for (B) CD3+ T cells (percentage of total cells in tumors); C, Exhaustion markers, PD-1+TIM3+ (percentage of all human CD3+ T cells that are positive for both markers); and D, The ratio of CD86+ F4/80+ CD11b+ to CD206+ F4/80+ CD11b+ myeloid cells. All data were analyzed by two-way ANOVA and plotted as mean ±SEM (*, P < 0.05; **, P < 0.01; ***, P < 0.001); n  =  5 mice/group. Data shown are representative of two independent experiments. E, Representative images showing human CD3+ T-cell infiltrations into MDA-MB-231 and KB cell solid tumors; bar, 200 μm. Where error bars appear to be missing, they are present but smaller than the symbols.

Figure 5.

Effect of anti-fluorescein CAR T-cell therapy on the growth and immunologic properties of MDA-MB-231 and KB tumors. NSG mice were implanted with 4 million MDA-MB-231 cells and 1 million KB cells on separate flanks of the same mice, and infused 2 weeks later with either saline or 8 ×106 anti-fluorescein CAR-T cells (i.e., when MDA-MB-231 and KB tumors reached approximately 160 or 80 mm3, respectively; that is, to accommodate their different growth rates). Four and 24 hours after CAR T-cell infusion, all mice were intravenously injected with 500 nmol/kg fluorescein–folate and this injection was repeated once/week thereafter. KB tumor cells express approximately 2 million folate receptors/cell whereas MDA-MB-231 cells express approximately 1.6 million folate receptors/cell. A, Tumor volumes were analyzed in cohorts either left untreated (dashed lines) or treated with both anti-fluorescein CAR-T cells plus fluorescein–folate bispecific adapter (solid lines). On day 18, tumors were resected and both cancer and stromal cells were released using a tumor dissociation kit prior xenografts after treatment to analysis by flow cytometry for (B) CD3+ T cells (percentage of total cells in tumors); C, Exhaustion markers, PD-1+TIM3+ (percentage of all human CD3+ T cells that are positive for both markers); and D, The ratio of CD86+ F4/80+ CD11b+ to CD206+ F4/80+ CD11b+ myeloid cells. All data were analyzed by two-way ANOVA and plotted as mean ±SEM (*, P < 0.05; **, P < 0.01; ***, P < 0.001); n  =  5 mice/group. Data shown are representative of two independent experiments. E, Representative images showing human CD3+ T-cell infiltrations into MDA-MB-231 and KB cell solid tumors; bar, 200 μm. Where error bars appear to be missing, they are present but smaller than the symbols.

Close modal

Next, to assure that a fluorescein-conjugated drug can selectively bind and enter anti-fluorescein CAR-T cells in vivo, we grew KB tumors on the shoulders of NSG mice and then treated the mice via tail vein injection with the fluorescein–NIR dye conjugate shown in Supplementary Scheme S2. After resecting the tumors and separating the cells with a human tumor dissociation kit, we analyzed the resulting cell suspension for NIR dye fluorescence by flow cytometry. As shown in Fig. 6, only CD3-positive cells were found to retain fluorescein-conjugated dye with no fluorescent dye retained in any other nonmalignant or malignant cell types, demonstrating that the fluorescein receptor resides solely on the CD3-positive cell population in the tumor mass. Moreover, that only CAR-transduced cells express a fluorescein receptor could be further established by showing that solely GFP-positive T cells retain the NIR fluorescence. These data argue that a systemically administered fluorescein-linked TLR7a will accumulate almost exclusively in anti-fluorescein CAR-T cells; that is, greatly reducing the possibility of systemic activation of the immune system and the associated global toxicities. When considered together with the fact that TLR7 receptors are located in endosomal compartments that are inaccessible to a fluorescein–TLR7–1A conjugate in any cell lacking a fluorescein receptor, the likelihood of off-target toxicities seems further remote.

Figure 6.

Demonstration that fluorescein-conjugated fluorescent dye (fluorescein–NIR dye) is specifically targeted to anti-fluorescein CAR-T cells in vivo. NSG mice were implanted with 1 million KB cells in one flank and infused with anti-fluorescein CAR-T cells (8 × 106 cells that contained approximately 50% anti-fluorescein CAR-T cells) when KB tumor volumes reached approximately 50 mm3. 500 nmol/kg fluorescein–folate was injected 4 and 24 hours later and again once/week thereafter. Fifteen days after CAR T-cell infusion, mice were tail vein injected with 500 nmol/kg fluorescein–NIR dye, and 4 hours later tumors were dissociated and analyzed by flow cytometry for uptake of fluorescein–NIR dye. CD3+ T cells were detected with anti-human CD3 on the APC-Cy7 channel, and GFP-transfected anti-fluorescein CAR-T cells were detected using the GFP channel.

Figure 6.

Demonstration that fluorescein-conjugated fluorescent dye (fluorescein–NIR dye) is specifically targeted to anti-fluorescein CAR-T cells in vivo. NSG mice were implanted with 1 million KB cells in one flank and infused with anti-fluorescein CAR-T cells (8 × 106 cells that contained approximately 50% anti-fluorescein CAR-T cells) when KB tumor volumes reached approximately 50 mm3. 500 nmol/kg fluorescein–folate was injected 4 and 24 hours later and again once/week thereafter. Fifteen days after CAR T-cell infusion, mice were tail vein injected with 500 nmol/kg fluorescein–NIR dye, and 4 hours later tumors were dissociated and analyzed by flow cytometry for uptake of fluorescein–NIR dye. CD3+ T cells were detected with anti-human CD3 on the APC-Cy7 channel, and GFP-transfected anti-fluorescein CAR-T cells were detected using the GFP channel.

Close modal

Finally, to evaluate the ability of our fluorescein–TLR7–1A conjugate to reactivate exhausted CAR-T cells in vivo, we exploited the capacity of KB cells to create a strongly immunosuppressive tumor microenvironment that would significantly challenge the ability of the fluorescein–TLR7–1A to reprogram an exhausted CAR T-cell into a tumoricidal CAR T-cell (Fig. 7). As shown in Fig. 7A and B, supplementation of the anti-fluorescein CAR T-cell therapy with systemically administered fluorescein–TLR7–1A significantly retarded KB tumor growth compared with mice treated with CAR-T cells alone. Although no improvement was observed in either total T-cell infiltration (Fig. 7D) or the inflammatory state of TAMs (Fig. 7F), the exhaustion markers on the CAR-T cells decreased dramatically (Fig. 7E). An example of flow cytometry dot pot showing the gating of the exhaustion markers is shown in Supplementary Fig. S6. Together with data showing that co-administration of fluorescein–TLR7–1A causes no additional weight loss (Fig. 7C), these data argue that a CAR T-cell–targeted TLR7 agonist can significantly improve exhausted CAR T-cell performance without causing any enhanced systemic toxicity.

Figure 7.

Rejuvenation of anti-fluorescein CAR-T cells in vivo following intravenous injection of fluorescein–TLR7–1A conjugate. A, Schema for in vivo studies. Mice were injected subcutaneously on day −7 with 106 KB cells and then infused on day 1 with 8 × 106 anti-fluorescein CAR-T cells. At 6 hours, 24 hours, and 9 days later, mice were intravenously injected with fluorescein–folate to induce engagement of CAR-T cells with folate receptor positive KB cancer cells. Then on days 4–7 and 11–14, mice were intravenously injected with fluorescein–TLR7–1A. Tumor volumes (B) and animal body weight changes (C) were measured every 3 days. Tumors were resected and dissociated into component cells on day 16 and human CD3+ T cells were determined as a percentage of all tumor cells in D. PD-1+TIM3+ cells as a percentage of all human CD3+ T cells are shown in E, and the ratio of mouse CD86+ to CD206+ cells also expressing myeloid markers F4/80 and CD11b in the anti-fluorescein CAR T-cell treatment groups are shown in F. D–F, Do not show results from mice that were not treated with fluorescein–folate, because in the absence of fluorescein–folate the CAR-T cells cannot engage folate receptor–expressing cancer cells, and therefore cannot proliferate; that is, leading to their rapid disappearance from circulation. All data are plotted as mean ±SEM, n = 5. Data shown are representative of at least two independent experiments. Data were analyzed by two-way ANOVA (**, P < 0.01; n.s., not significant).

Figure 7.

Rejuvenation of anti-fluorescein CAR-T cells in vivo following intravenous injection of fluorescein–TLR7–1A conjugate. A, Schema for in vivo studies. Mice were injected subcutaneously on day −7 with 106 KB cells and then infused on day 1 with 8 × 106 anti-fluorescein CAR-T cells. At 6 hours, 24 hours, and 9 days later, mice were intravenously injected with fluorescein–folate to induce engagement of CAR-T cells with folate receptor positive KB cancer cells. Then on days 4–7 and 11–14, mice were intravenously injected with fluorescein–TLR7–1A. Tumor volumes (B) and animal body weight changes (C) were measured every 3 days. Tumors were resected and dissociated into component cells on day 16 and human CD3+ T cells were determined as a percentage of all tumor cells in D. PD-1+TIM3+ cells as a percentage of all human CD3+ T cells are shown in E, and the ratio of mouse CD86+ to CD206+ cells also expressing myeloid markers F4/80 and CD11b in the anti-fluorescein CAR T-cell treatment groups are shown in F. D–F, Do not show results from mice that were not treated with fluorescein–folate, because in the absence of fluorescein–folate the CAR-T cells cannot engage folate receptor–expressing cancer cells, and therefore cannot proliferate; that is, leading to their rapid disappearance from circulation. All data are plotted as mean ±SEM, n = 5. Data shown are representative of at least two independent experiments. Data were analyzed by two-way ANOVA (**, P < 0.01; n.s., not significant).

Close modal

We have demonstrated that an endocytosing CAR can be exploited to deliver four different fluorescein-attached molecules selectively into anti-fluorescein CAR-T cells, thereby avoiding their uptake into CAR-negative cells. This specificity for the engineered CAR-T cells required (i) expression of the anti-fluorescein CAR solely on T cells, (ii) internalization of the fluorescein–drug conjugate exclusively by CAR-expressing T cells, and (iii) accessibility of the TLR7 agonist to a Toll-like receptor 7 within the internalizing T-cell endosome. Fortunately, all three conditions were met using the combination of anti-fluorescein CAR T-cell and fluorescein-linked drug described here. Although the nontargeted TLR7–1A was readily membrane permeable, its conjugate to fluorescein-(PEG)3 was impermeable, thereby preventing the conjugate from passively entering any CAR-negative cell. Similar to an unmodified T-cell receptor (44), the anti-fluorescein CAR also readily internalized, thereby mediating delivery of the fluorescein-attached drug into the cell. Although the internalized fluorescein–drug conjugate could not escape its entrapping endosome, because all Toll-like receptors 7 are located within endosomal compartments (45), the fluorescein–TLR7–1A conjugate needed only to dissociate from the internalized CAR and diffuse within the same endosome to an unoccupied TLR7 to induce an immune response. Importantly, this series of requirements for effective signaling was critical to the success of the CAR T-cell rejuvenation strategy, because nontargeted TLR7 agonists have proven to be too toxic for systemic administration due to their systemic stimulation of cytokine release by multiple immune cell types (46–48). Indeed, the only TLR7 agonists approved to date by the FDA (i.e., resiquimod and imiquimod) are indicated solely for localized topical applications on the skin (46, 49).

One of the limitations of using a TLR7 agonist for immune cell stimulation is that its activation of the immune system follows bell-shaped curve (50). Thus, use of a TLR7 agonist to rejuvenate an exhausted CAR T-cell requires determination of an optimal concentration, because injection of either too little or too much will have a reduced impact on T-cell properties. Fortunately, the data collected above suggest that optimal CAR T-cell stimulation can be achieved over a concentration range of at least a two orders of magnitude. Nevertheless, whenever a targeted TLR7 agonist is used for CAR T-cell rejuvenation in humans, it will be important to either define a universal concentration window over which the TLR7 agonist induces CAR T-cell activation or develop a personalized assay to define the ideal TLR7 agonist dose for each patient.

Although only two fluorescent dyes and a TLR7 agonist were delivered into CAR-T cells in this study, the large structural differences among these molecules suggest that almost any small-molecular weight drug might be targeted to CAR-T cells using a similar strategy. Although reversal of CAR T-cell exhaustion was the sole application explored here, the same internalization pathway should conceivably be usable to target drugs that can proliferate, kill, differentiate, or otherwise modulate CAR-T cells in vivo. Thus, if a CAR T-cell clone were to exhibit an unacceptable toxicity, it could feasibly be eliminated with a fluorescein-linked cytotoxic drug. In contrast, if a promising CAR T-cell preparation were to fail to proliferate in vivo, immunomodulators that promote CAR T-cell expansion might be delivered. Given the plethora of challenges that CAR-T cells encounter following infusion into a patient, a mechanism that can allow their post-infusion manipulation could prove invaluable. And while engineering a different delivery strategy for each immunomodulatory drug might accomplish the same objective, use of the pre-existing CAR to perform the same function for almost all drugs should prove much simpler.

Q. Luo reports a patent for Wo2021178887A1 pending. J.V. Napoleon reports a patent for WO2021178887A1 pending. P.S. Low reports grants from Umoja Biopharma during the conduct of the study; and grants from Umoja Biopharma outside the submitted work; as well as reports a patent for WO2021178887A1 pending and licensed to Umoja Biopharma. No disclosures were reported by the other authors.

Q. Luo: Conceptualization, data curation, formal analysis, validation, investigation, visualization, methodology, writing–original draft. J.V. Napoleon: Methodology, writing–original draft, synthesized some of the compounds. X. Liu: Methodology, synthesized some of the compounds. B. Zhang: Methodology. S. Zheng: Data curation, investigation, visualization, writing–original draft, writing–review and editing. P.S. Low: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, methodology, writing–original draft, project administration, writing–review and editing.

This work was funded by a grant from Umoja Biopharma (No. 40003064) and gift from the Hurvis Foundation (No. 14121659). Q. Luo, J.V. Napoleon, S. Zheng, and P.S. Low were supported by the grant from Umoja Biopharma. B. Zhang was supported by the Hurvis Foundation. The authors would like to acknowledge the Purdue Center for Cancer Research, the Purdue Institute for Drug Discovery, the Purdue Flow Cytometry Core, the Purdue Imaging Facility and the Chemical Genomic Facility for their services.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Charrot
S
,
Hallam
S
.
CAR-T cells: future perspectives
.
Hemasphere
2019
;
3
:
e188
.
2.
Holzinger
A
,
Barden
M
,
Abken
H
.
The growing world of CAR T-cell trials: a systematic review
.
Cancer Immunol Immunother
2016
;
65
:
1433
50
.
3.
Almåsbak
H
,
Aarvak
T
,
Vemuri
MC
.
CAR T-cell therapy: a game changer in cancer treatment
.
J Immunol Res
2016
;
2016
:
5474602
.
4.
Cummins
KD
,
Gill
S
.
Chimeric antigen receptor T-cell therapy for acute myeloid leukemia: how close to reality?
Haematologica
2019
;
104
:
1302
8
.
5.
Maude
SL
,
Laetsch
TW
,
Buechner
J
,
Rives
S
,
Boyer
M
,
Bittencourt
H
, et al
.
Tisagenlecleucel in children and young adults with B-cell lymphoblastic leukemia
.
N Engl J Med
2018
;
378
:
439
48
.
6.
Neelapu
SS
,
Locke
FL
,
Bartlett
NL
,
Lekakis
LJ
,
Miklos
DB
,
Jacobson
CA
, et al
.
Axicabtagene ciloleucel car T-cell therapy in refractory large B-cell lymphoma
.
N Engl J Med
2017
;
377
:
2531
44
.
7.
O'Leary
MC
,
Lu
X
,
Huang
Y
,
Lin
X
,
Mahmood
I
,
Przepiorka
D
, et al
.
FDA approval summary: tisagenlecleucel for treatment of patients with relapsed or refractory B-cell precursor acute lymphoblastic leukemia
.
Clin Cancer Res
2019
;
25
:
1142
6
.
8.
Hosen
N
.
Chimeric antigen receptor T-cell therapy for multiple myeloma
.
Cancers
2019
;
11
:
2024
.
9.
Kershaw
MH
,
Westwood
JA
,
Parker
LL
,
Wang
G
,
Eshhar
Z
,
Mavroukakis
SA
, et al
.
A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer
.
Clin Cancer Res
2006
;
12
:
6106
15
.
10.
Lamers
CH
,
Sleijfer
S
,
Vulto
AG
,
Kruit
WH
,
Kliffen
M
,
Debets
R
, et al
.
Treatment of metastatic renal cell carcinoma with autologous T-lymphocytes genetically retargeted against carbonic anhydrase IX: first clinical experience
.
J Clin Oncol
2006
;
24
:
e20
2
.
11.
Maus
MV
,
Haas
AR
,
Beatty
GL
,
Albelda
SM
,
Levine
BL
,
Liu
X
, et al
.
T cells expressing chimeric antigen receptors can cause anaphylaxis in humans
.
Cancer Immunol Res
2013
;
1
:
26
31
.
12.
Morgan
RA
,
Yang
JC
,
Kitano
M
,
Dudley
ME
,
Laurencot
CM
,
Rosenberg
SA
.
Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2
.
Mol Ther
2010
;
18
:
843
51
.
13.
Martinez
M
,
Moon
EK
. CAR-T cells
for solid tumors: new strategies for finding, infiltrating, and surviving in the tumor microenvironment
.
Front Immunol
2019
;
10
:
128
.
14.
Majzner
RG
,
Mackall
CL
.
Tumor antigen escape from CAR T-cell therapy
.
Cancer Discov
2018
;
8
:
1219
26
.
15.
Gardner
R
,
Finney
O
,
Smithers
H
,
Leger
KJ
,
Annesley
CE
,
Summers
C
, et al
.
CD19CAR T cell products of defined CD4:CD8 composition and transgene expression show prolonged persistence and durable MRD-negative remission in pediatric and young adult B-cell all
.
Blood
2016
;
128
:
219
-.
16.
Maude
SL
,
Frey
N
,
Shaw
PA
,
Aplenc
R
,
Barrett
DM
,
Bunin
NJ
, et al
.
Chimeric antigen receptor T cells for sustained remissions in leukemia
.
N Engl J Med
2014
;
371
:
1507
17
.
17.
Kalos
M
,
Levine
BL
,
Porter
DL
,
Katz
S
,
Grupp
SA
,
Bagg
A
, et al
.
T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia
.
Sci Transl Med
2011
;
3
:
95ra73
.
18.
Porter
DL
,
Hwang
WT
,
Frey
NV
,
Lacey
SF
,
Shaw
PA
,
Loren
AW
, et al
.
Chimeric antigen receptor T cells persist and induce sustained remissions in relapsed refractory chronic lymphocytic leukemia
.
Sci Transl Med
2015
;
7
:
303ra139
.
19.
Caruana
I
,
Savoldo
B
,
Hoyos
V
,
Weber
G
,
Liu
H
,
Kim
ES
, et al
.
Heparanase promotes tumor infiltration and antitumor activity of CAR-redirected T lymphocytes
.
Nat Med
2015
;
21
:
524
9
.
20.
Stylianopoulos
T
,
Martin
JD
,
Chauhan
VP
,
Jain
SR
,
Diop-Frimpong
B
,
Bardeesy
N
, et al
.
Causes, consequences, and remedies for growth-induced solid stress in murine and human tumors
.
Proc Natl Acad Sci U S A
2012
;
109
:
15101
8
.
21.
Gajewski
TF
,
Woo
SR
,
Zha
Y
,
Spaapen
R
,
Zheng
Y
,
Corrales
L
, et al
.
Cancer immunotherapy strategies based on overcoming barriers within the tumor microenvironment
.
Curr Opin Immunol
2013
;
25
:
268
76
.
22.
Feig
C
,
Jones
JO
,
Kraman
M
,
Wells
RJ
,
Deonarine
A
,
Chan
DS
, et al
.
Targeting CXCL12 from FAP-expressing carcinoma-associated fibroblasts synergizes with anti–PD-L1 immunotherapy in pancreatic cancer
.
Proc Natl Acad Sci U S A
2013
;
110
:
20212
7
.
23.
Moon
EK
,
Wang
LC
,
Dolfi
DV
,
Wilson
CB
,
Ranganathan
R
,
Sun
J
, et al
.
Multifactorial T-cell hypofunction that is reversible can limit the efficacy of chimeric antigen receptor-transduced human T cells in solid tumors
.
Clin Cancer Res
2014
;
20
:
4262
73
.
24.
Zhang
Y
,
Ertl
HC
.
Starved and asphyxiated: how can CD8(+) T cells within a tumor microenvironment prevent tumor progression
.
Front Immunol
2016
;
7
:
32
.
25.
Fischer
K
,
Hoffmann
P
,
Voelkl
S
,
Meidenbauer
N
,
Ammer
J
,
Edinger
M
, et al
.
Inhibitory effect of tumor cell–derived lactic acid on human T cells
.
Blood
2007
;
109
:
3812
9
.
26.
Rodriguez-Garcia
A
,
Palazon
A
,
Noguera-Ortega
E
,
Powell
DJ
. Jr,
Guedan
S
.
CAR-T cells hit the tumor microenvironment: strategies to overcome tumor escape
.
Front Immunol
2020
;
11
:
1109
.
27.
Newick
K
,
O'Brien
S
,
Moon
E
,
Albelda
SM
.
CAR T-cell therapy for solid tumors
.
Annu Rev Med
2017
;
68
:
139
52
.
28.
McLane
LM
,
Abdel-Hakeem
MS
,
Wherry
EJ
.
CD8 T-cell exhaustion during chronic viral infection and cancer
.
Annu Rev Immunol
2019
;
37
:
457
95
.
29.
Mirzaei
HR
,
Rodriguez
A
,
Shepphird
J
,
Brown
CE
,
Badie
B
.
Chimeric antigen receptors T-cell therapy in solid tumor: challenges and clinical applications
.
Front Immunol
2017
;
8
:
1850
.
30.
Yu
S
,
Li
A
,
Liu
Q
,
Li
T
,
Yuan
X
,
Han
X
, et al
.
Chimeric antigen receptor T cells: a novel therapy for solid tumors
.
J Hematol Oncol
2017
;
10
:
78
.
31.
Minutolo
NG
,
Hollander
EE
,
Powell
DJ
.
The emergence of universal immune receptor T-cell therapy for cancer
.
Front Oncol
2019
;9.
32.
Liu
D
,
Zhao
J
,
Song
Y
.
Engineering switchable and programmable universal CARs for CAR T therapy
.
J Hematol Oncol
2019
;
12
:
69
.
33.
Lee
YG
,
Marks
I
,
Srinivasarao
M
,
Kanduluru
AK
,
Mahalingam
SM
,
Liu
X
, et al
.
Use of a single CAR T-cell and several bispecific adapters facilitates eradication of multiple antigenically different solid tumors
.
Cancer Res
2019
;
79
:
387
96
.
34.
Lee
YG
,
Chu
H
,
Lu
Y
,
Leamon
CP
,
Srinivasarao
M
,
Putt
KS
, et al
.
Regulation of CAR T-cell–mediated cytokine release syndrome-like toxicity using low molecular weight adapters
.
Nat Commun
2019
;
10
:
2681
.
35.
Kawalekar
OU
,
O'Connor
RS
,
Fraietta
JA
,
Guo
L
,
McGettigan
SE
,
Posey
AD
Jr
, et al
.
Distinct signaling of coreceptors regulates specific metabolism pathways and impacts memory development in CAR-T cells
.
Immunity
2016
;
44
:
380
90
.
36.
Zhang
B
,
Napoleon
JV
,
Liu
X
,
Luo
Q
,
Srinivasarao
M
,
Low
PS
.
Sensitive manipulation of CAR T-cell activity using a chimeric endocytosing receptor
.
J Immunother Cancer
2020
;
8
:
e000756
.
37.
Lu
YJ
,
Chu
H
,
Wheeler
LW
,
Nelson
M
,
Westrick
E
,
Matthaei
JF
, et al
.
Preclinical evaluation of bispecific adaptor molecule controlled folate receptor CAR T-cell therapy with special focus on pediatric malignancies
.
Front Oncol
2019
;
9
:
151
.
38.
Marofi
F
,
Motavalli
R
,
Safonov
VA
,
Thangavelu
L
,
Yumashev
AV
,
Alexander
M
, et al
.
CAR-T cells in solid tumors: challenges and opportunities
.
Stem Cell Res Ther
2021
;
12
:
81
.
39.
Pellegrino
C
,
Favalli
N
,
Sandholzer
M
,
Volta
L
,
Bassi
G
,
Millul
J
, et al
.
Impact of ligand size and conjugation chemistry on the performance of universal chimeric antigen receptor
T cells
for tumor killing
.
Bioconjug Chem
2020
;
31
:
1775
83
.
40.
Mahalingam
SM
,
Dudkin
VY
,
Goldberg
S
,
Klein
D
,
Yi
F
,
Singhal
S
, et al
.
Evaluation of a centyrin-based near-infrared probe for fluorescence-guided surgery of epidermal growth factor receptor positive tumors
.
Bioconjug Chem
2017
;
28
:
2865
73
.
41.
Leamon
CP
,
Reddy
JA
,
Dorton
R
,
Bloomfield
A
,
Emsweller
K
,
Parker
N
, et al
.
Impact of high and low folate diets on tissue folate receptor levels and antitumor responses toward folate–drug conjugates
.
J Pharmacol Exp Ther
2008
;
327
:
918
25
.
42.
Cresswell
GM
,
Wang
B
,
Kischuk
EM
,
Broman
MM
,
Alfar
RA
,
Vickman
RE
, et al
.
Folate receptor beta designates immunosuppressive tumor-associated myeloid cells that can be reprogrammed with folate-targeted drugs
.
Cancer Res
2021
;
81
:
671
84
.
43.
Edwards
S
,
Jones
C
,
Leishman
AJ
,
Young
BW
,
Matsui
H
,
Tomizawa
H
, et al
.
TLR7 stimulation of APCs results in inhibition of IL-5 through type I IFN and notch signaling pathways in human peripheral blood mononuclear cells
.
J Immunol
2013
;
190
:
2585
92
.
44.
Lou
J
,
Rossy
J
,
Deng
Q
,
Pageon
SV
,
Gaus
K
.
New insights into how trafficking regulates T-cell receptor signaling
.
Front Cell Dev Biol
2016
;
4
:
77
.
45.
Petes
C
,
Odoardi
N
,
Gee
K
.
The Toll for trafficking: toll-like receptor 7 delivery to the endosome
.
Front Immunol
2017
;
8
;
1075
.
46.
Biffen
M
,
Matsui
H
,
Edwards
S
,
Leishman
AJ
,
Eiho
K
,
Holness
E
, et al
.
Biological characterization of a novel class of toll-like receptor 7 agonists designed to have reduced systemic activity
.
Br J Pharmacol
2012
;
166
:
573
86
.
47.
Harrison
LI
,
Skinner
SL
,
Marbury
TC
,
Owens
ML
,
Kurup
S
,
McKane
S
, et al
.
Pharmacokinetics and safety of imiquimod 5% cream in the treatment of actinic keratoses of the face, scalp, or hands and arms
.
Arch Dermatol Res
2004
;
296
:
6
11
.
48.
Savage
P
,
Horton
V
,
Moore
J
,
Owens
M
,
Witt
P
,
Gore
ME
.
A phase I clinical trial of imiquimod, an oral interferon inducer, administered daily
.
Br J Cancer
1996
;
74
:
1482
6
.
49.
Meyer
T
,
Surber
C
,
French
LE
,
Stockfleth
E
.
Resiquimod, a topical drug for viral skin lesions and skin cancer
.
Expert Opin Investig Drugs
2013
;
22
:
149
59
.
50.
Rodell
CB
,
Ahmed
MS
,
Garris
CS
,
Pittet
MJ
,
Weissleder
R
.
Development of adamantane-conjugated TLR7/8 agonists for supramolecular delivery and cancer immunotherapy
.
Theranostics
2019
;
9
:
8426
36
.

Supplementary data