Expanding CAR T cells in human platelet lysate renders T cells with in vivo longevity
Pre-clinical and clinical studies have shown that the infusion of CAR T cells with a naive-like (TN) and central memory (TCM) phenotype is associated with prolonged in vivo T cell persistence and superior anti-tumor effects. To optimize the maintenance of such populations during the in vitro preparation process, we explored the impact of T cell exposure to both traditional [fetal bovine serum (FBS), human AB serum (ABS)] and non-traditional [human platelet lysate (HPL) - a xeno-free protein supplement primarily used for the production of clinical grade mesenchymal stromal / stem cells (MSCs)] serum supplements.
Second generation chimeric antigen receptor with CD28 and CD3ζ endodomain targeting prostate stem cell antigen (PSCA) (P28z) or CD19 (1928z) were constructed and used for this study. After retroviral transduction, CAR T cells were divided into 3 conditions containing either FBS, ABS or HPL and expanded for 7 days. To evaluate the effect of different sera on CAR T cell function, we performed a series of in vitro and in vivo experiments.
HPL-exposed CAR T cells exhibited the less differentiated T cell phenotype and gene signature, which displayed inferior short-term killing abilities (compared to their FBS- or ABS-cultured counterparts) but superior proliferative and anti-tumor effects in long-term in vitro coculture experiments. Importantly, in mouse xenograft model, HPL-exposed CAR T cells outperformed their ABS or FBS counterparts against both subcutaneous tumor (P28z T cells against Capan-1PSCA) and systemic tumor (1928z T cells against NALM6). We further observed maintenance of less differentiated T cell phenotype in HPL-exposed 1928z T cells generated from patient’s PBMCs with superior anti-tumor effect in long-term in vitro coculture experiments.
Our study highlights the importance of serum choice in the generation of CAR T cells for clinical use.
KeywordsCAR T cells Persistence Memory phenotype Manufacture Human platelet lysate
Second generation CAR targeting CD19 with CD28 costimulatory domain
Human AB serum
B lymphoblastic leukemia/lymphoma
Chimeric antigen receptor
CRISPR associated protein 9
Clustered regularly interspaced short palindromic repeats
Cytotoxic T lymphocyte
CellTrace Violet dye
Dominant TGFβ receptor II
Fetal bovine serum
Green fluorescence protein
Good manufacturing practices
Human platelet lysate
Internal ribosome entry site
Mesenchymal stromal / stem cells
Non-transduced T cells
Second generation CAR targeting PSCA with CD28 costimulatory domain
Peripheral blood mononuclear cells
Prostate stem cell antigen
Single chain variable fragment
Transcription activator-like effector nuclease
Central memory T cells
Effector T cells
Transforming growth factor beta 1
Type 2 helper
Naïve T cells
The clinical success of adoptively transferred CD19 targeted chimeric antigen receptor (CAR) modified T cells for the treatment of B cell lymphoma / leukemia has precipitated the extension of this approach to a spectrum of both hematologic malignancies and solid tumors [1, 2]. In parallel, given that in vivo persistence has been shown to correlate with superior outcomes [3, 4], various groups have also explored strategies to enhance T cell longevity ranging from the incorporation of transgenes to support cell expansion (e.g. stimulatory cytokines [5, 6, 7, 8] such as IL12 and IL15 or modified cytokine / inhibitory receptor [9, 10, 11, 12] to protect cells from the suppressive tumor microenvironment) to manufacturing modifications designed to retain less differentiated T cells (e.g. naïve and central memory T cells) in the infused product. The latter strategy includes the isolation of less differentiated T cell subsets prior to ex vivo activation , the incorporation of homeostatic cytokines (e.g. IL7 and IL15 ) known to preserve central memory T cells [15, 16, 17] for ex vivo expansion, or chemical manipulation of signaling pathways known to be involved in T cell differentiation [18, 19, 20], including the activation of Wnt / β-catenin pathway using the GSK3β inhibitor TWS119 [18, 21] or the inhibition of the PI3K/AKT and mTOR pathways using small-molecule inhibitors [22, 23, 24]. Though all have proven to effectively enrich for the target T cell populations of interest, the additional complexity (e.g. use of magnetic beads for isolation) and associated costs serve as an impediment to broad clinical implementation.
In the current study we sought to address the issues of complexity and cost by exploring whether T cell differentiation status could be influenced by choice of serum supplementation. Whereas traditional CAR T cell cultures are maintained in fetal bovine (FBS) or human AB serum (ABS)-supplemented medium we investigated the impact of exposure to human platelet lysate (HPL) as an alternative xeno-free protein supplement being used for the expansion of mesenchymal stromal / stem cells (MSCs) in clinic. We now demonstrate in a series of in vitro and in vivo experiments, performed in both hematologic and solid tumor models, the profound qualitative impact of serum supplementation on CAR T cell performance.
Donors and cell lines
Peripheral blood mononuclear cells (PBMCs) were obtained from healthy volunteers and B cell lymphoma and B-ALL patients after informed consent on protocols approved by the Baylor College of Medicine Institutional Review Board (H-15152, H-27471, H-19384 and H-31970). Capan-1 (pancreatic cancer cell line) and DU145 (prostate cancer cell line) were obtained from the American Type Culture Collection (Rockville, MD). NALM6 (pre-B-ALL cell line) and Raji (Burkitt lymphoma cell line) were gifted by Dr. Maksim Mamonkin. Cells were maintained in a humidified atmosphere containing 5% carbon dioxide (CO2) at 37 °C. Capan-1 and DU145 cells were maintained in Iscove’s Modified Dulbecco’s Medium (IMDM; Gibco BRL Life Technologies, Inc., Gaithersburg, MD) while NALM6 and Raji cells were maintained in RPMI-1640 media (GE Healthcare Life Sciences, Pittsburgh, PA). 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.) while other cell lines were grown in their specific media containing 10% FBS with 2 mM L-GlutaMAX.
Generation of retroviral constructs and retrovirus production
A second generation CAR construct targeting PSCA was previously generated in our lab . Briefly, our CAR construct is comprised of scFv domain followed by IgG2 derived-Hinge CH3 spacer with CD28 transmembrane / costimulatory and CD3z signaling domains (P28z). Second generation CAR targeting CD19 was generated based on the P28z construct by replacing the anti-PSCA (clone 2B3) scFv domain with an anti-CD19 scFv (clone FMC63) using restriction enzymes XhoI and BamHI (1928z). The γ-retroviral vectors encoding PSCA-IRES-GFP, GFP/Firefly luciferase fusion protein (GFP/FL) and dominant TGFβ receptor II (DNRII), and the retroviral supernatant were generated as previously described [25, 26, 27].
Generation of CAR-modified T cells and gene-modified cell lines
To generate CAR T cells, 1 × 106 PBMCs were plated in each well of a non-tissue culture-treated 24-well plate pre-coated with OKT3 (1 mg/mL; Ortho Biotech, Inc., Bridgewater, NJ) and anti-CD28 (1 mg/mL; Becton Dickinson & Co., Mountain View, CA) antibodies. Cells were cultured in 10% FBS CTL media [50% RPMI-1640, 50% Clicks medium (Irvine Scientific, Inc., Santa Ana, CA) and 2 mM L-GlutaMAX)], which was supplemented with recombinant human IL2 (50 U/mL, NIH, Bethesda, MD) on day 0. On day 3, retroviral supernatant was plated in a non-tissue culture-treated 24-well plate pre-coated with recombinant fibronectin fragment (FN CH-296; Retronectin; TAKARA BIO INC, Otsu, Japan), and centrifuged at 2000 g for 90 min. After removal of the supernatant, OKT3/CD28-activated PBMCs (0.1 × 106/mL) were resuspended in complete medium supplemented with IL2 (100 U/mL) and 2 mL was added to each virus-loaded well, which was subsequently spun at 400 g for 5 min, and then transferred to a 37 °C, 5% CO2 incubator. On day 3 post transduction, T cells were harvested, washed, and cultured in CTL medium containing different serum supplements - FBS, human ABS (Valley Biomedical, Winchester, Virginia), or pathogen-reduced human platelet lysate (HPL; nLiven PR, Cook Regentec, Indianapolis, IN). In this study, a single lot of HPL was randomly selected as previous work has demonstrated lot-to-lot consistency . Cultures were supplemented with fresh medium and IL2 (50 U/mL) every 2–3 days. To co-express CAR and GFP/FL for in vivo bioluminescence imaging, activated T cells were first modified to express the CAR on day 2 and transduced with GFP/FL on day 3 using the same protocol. Transduction efficiency was measured 3 days post transduction by flow cytometry. To track T cell numbers over time, viable cells were manually counted using trypan blue. To generate tumor cell lines overexpressing PSCA/GFP or GFP/FL, we used the same protocol as previously described and isolated the GFP positive fraction using a cell sorter (SH800S, Sony Biotechnology, San Jose, CA). While T cells were generated in CTL medium containing different serum supplements, all in vitro functional assays were performed in CTL medium supplemented with 10% FBS.
Genome editing of the CCR7 gene in T cells
Guide RNA for the CCR7 gene (gRNA sequence: GGGCAGGTAGGTATCGGTCA) was designed using CRISPRscan  and incorporated into an oligonucleotide primer and used to amplify the gRNA scaffold from PX458 plasmid (gift from Dr. Tim Sauer). gRNA was generated through in vitro transcription with HiScribe™ T7 High Yield RNA Synthesis Kit (New England Biolabs, Beverly, MA) from the DNA template and purified using the RNA Clean & Concentrator kit (Zymo Research, Irvine, CA). Electroporation of 0.25 × 106 T cells was performed in 10 μL of buffer T with 1 μg of gRNA and 1 μg of Cas9 protein (PNA Bio, Newbury Park, CA) by three consecutive 1600 V 10-ms pulses using the Neon Transfection System (Thermo Fisher Scientific, Waltham, MA).
Cells were stained with antibodies for 20 min at 4 °C. All samples were acquired on a Gallios Flow Cytometer (Beckman Coulter Life Sciences, Indianapolis, IN), and data was analyzed using Kaluza 2.1 Flow Analysis Software (Beckman Coulter Life Sciences). Antibodies used in this study are listed in Additional file 1: Table S1.
Total RNA was extracted from T cells maintained in different serum containing CTL medium using the RNeasy plus Mini kit (QIAGEN, Valencia, CA) and quantified using the NanoDrop 2000 (Thermo Fisher Scientific). RNA sequencing and analysis were performed by Novogene Corporation (Sacramento, CA). Heat map was generated using Heatmapper  .
The cytotoxicity and specificity of engineered T cells were evaluated in a standard 5 h 51Cr-release assay, as described previously .
P28z T cells (0.2 × 106 cells) were cocultured with DU145PSCA cells (0.01 × 106 cells, E:T = 20:1) in 200 μL in the presence of Monensin (BD GolgiStop, BD Biosciences, San Jose, CA) and CD107a-APC antibody (H4A3 / 641,581) for 4 h. Similarly, 1928z T cells (0.2 × 106 cells) were cocultured with NALM6 cells (0.2 × 106 cells, E:T = 1:1). Cells were stained for T cell surface markers and the expression of CD107a was analyzed by flow cytometry.
To measure cytokine production, 0.2 × 106 T cells were cocultured with 0.2 × 106 target cells in 200 μL of medium in a single well of a U-bottom 96-well plate for 24 h. Supernatants were collected and stored at − 80 °C. Cytokine levels were analyzed using MILLIPLEX MAP Human CD8+ T Cell Magnetic Bead Panel Premixed 17 Plex (Merck Millipore, Billerica, MA), according to manufacturer’s instructions.
Cell proliferation assay and detection of apoptotic cells
T cells were stained with CellTrace Violet (Thermo Fisher Scientific, Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol. The stained P28z T cells and 1928z T cells (0.5 × 106 cells) were cocultured with either Capan-1PSCA cells (0.1 × 106 cells) or NALM6 (0.5 × 106 cells), respectively, in a 24-well tissue culture plate for 5 days. Cells were harvested and stained for T cell surface markers, Annexin V-APC (BD Bioscience) and 7-AAD (BD Biosciences) and analyzed by flow cytometer.
In the coculture experiments with P28z T cells, 1.25 × 104 Capan-1PSCA cells were plated in 6-well plate on day − 1, then 5 × 105 T cells were added on day 0. For 1928z T cells, 0.1 × 106 1928z T cells were cocultured with 0.1 × 106 NALM6 cells. Cells were harvested, stained and analyzed by flow cytometer every 3 days. To quantify cells by flow cytometry, 20 μL of CountBright Absolute Counting Beads (Thermo Fisher Scientific, Invitrogen) was added and 7-AAD was added to exclude dead cells. Acquisition was halted at 2000 beads.
In vivo study
For the subcutaneous (s.c.) tumor model, NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ mice (NSG mice, Stock No: 005557, 5–7 weeks old, The Jackson Laboratory) were engrafted s.c. (right flank) with Capan-1PSCA cells (5 × 106 cells / animal) and once the tumors were established (day 21) the animals were treated with 1 or 2 × 106 of P28z T cells engineered to express GFP/FL intravenously. For tumor rechallenge, 5 × 106 Capan-1PSCA cells were injected on left flank on day 42 post T cell administration. Tumor size was measured by calipers and tumor volume was calculated as follows: tumor volume (mm3) = length x width2 / 2. T cell migration and distribution were evaluated by injecting mice intraperitoneally with 100 μL of D-luciferin (15 mg/mL, PerkinElmer Inc., Waltham, MA) followed by bioluminescence imaging using an IVIS Lumina II imaging system (Caliper Life Sciences, Inc., Hopkinton, MA), and analyzed by Living Image software (Caliper Life Sciences, Inc.). To assess PSCA expression on residual tumor, mice were sacrificed, tumors were dissected, and single cell suspensions were prepared, as previously published . Subsequently, cells were stained with either anti-PSCA or isotype control followed by Rat anti-mouse IgG1-APC. To distinguish Capan-1 cells, cells were further stained with anti-EpCAM-PE antibody and 7-AAD to exclude dead cells. For the systemic tumor model, 0.5 × 106 NALM6 cells engineered to express GFP/FL were injected into NSG mice intravenously on day − 3, then 5 or 10 × 106 1928z T cells were injected intravenously. To quantify T cells in the mouse peripheral blood, 50 μL of blood obtained by submandibular facial vein bleeding was stained with CD3, CD4, CD8, and CD45, then treated with RBC Lysis Buffer (BioLegend, San Diego, CA) to lyse red blood cells. CD45+CD3+ cells were counted by flow cytometer using CountBright Absolute Counting Beads. In the experiment to track T cell migration and expansion, mice were injected with 0.5 × 106 NALM6 cells followed by 5 × 106 1928z T cells modified to express GFP/FL. All in vivo experiments were performed according to the Baylor College of Medicine Animal Husbandry guidelines with approval from the Institutional Animal Care and Use Committee (IACUC).
Statistical analysis was performed using Graphpad Prism 7 software (GraphPad Software, Inc., La Jolla, CA). The statistical analysis used in each experiment is described in the figure legend.
Expanding CAR T cells in HPL results in maintenance of a less differentiated T cell phenotype
Effector function of CAR T cells expanded in HPL
HPL cultured CAR T cells showed a higher proliferative capacity leading to potent anti-tumor response in long-term in vitro coculture experiments
HPL-expanded P28z T cells show enhanced in vivo anti-tumor effects
HPL-cultured 1928z T cells exhibit higher proliferative capacity resulting in the elimination of NALM6 tumors
HPL cultures maintain a less differentiated phenotype of 1928z T cells generated from patients with B cell lymphoma and B cell leukemia
TGFβ1 in part plays an important role in maintaining a less differentiated CAR T cell phenotype
In our studies using two different CAR models, HPL-exposed T cells consistently outperformed their ABS or FBS counterparts, leading us to try and identify the component(s) that specifically influenced T cell phenotype. We performed human proteomic analysis of ABS and HPL, and found that HPL contains higher levels of transforming growth factor beta 1 (TGFβ1) compared with ABS (55.4 ± 5.6 vs 2.1 ± 1.1 ng/mL, HPL vs ABS, n = 3, in press). Since previous studies have noted that TGFβ1 prevents T cell differentiation and promotes the survival of activated and memory T cells [34, 35, 36], to explore the specific effects of TGFβ1 on memory T cell phenotype, we supplemented our FBS and ABS cultures with recombinant TGFβ1 (5 ng/mL) to normalize levels to that seen in 10% HPL cultures. As an additional control we used T cells transgenically expressing a dominant negative TGFβ receptor II (DNRII) [26, 37] to neutralize TGFβ1 in HPL. Interestingly, with TGFβ1 supplementation we observed a higher percentage of CCR7+ cells in FBS and ABS cultures, and substitution of DNRII T cells abrogated the effect of HPL on CCR7 expression (Additional file 2: Figure S7a). Taken together, therefore, these results suggest that TGFβ1 has an impact on the maintenance of a less differentiated T cell phenotype. Not surprisingly, though, the killing ability of P28z T cells cultured with recombinant TGFβ1 was impaired in both in vitro and in vivo experiments (Additional file 2: Figure S7b and S7c), suggesting that the combination of various proteins with TGFβ1 contribute to the maintenance of less differentiated T cell phenotype with the retention of effector function.
The correlation between superior clinical outcomes and in vivo T cell persistence has led to the development of various strategies (genetic modification, mechanical isolation, chemical manipulation) designed to preserve/enrich for cells of a less differentiated phenotype within the infusion product. To achieve the same goal with minimal complexity we explored the impact of serum (protein) exposure on CAR T cell phenotype and discovered that simple replacement of traditional FBS or ABS serum with HPL, a GMP-grade xeno-free supplement, arrested CAR T cell differentiation at the TN and TCM phase. HPL-exposed CAR T cells exhibited high proliferative capacity and enhanced long-term in vivo persistence compared to their FBS or ABS counterparts, resulting in superior anti-tumor effects. This data supports the incorporation of HPL in the preparation of clinical grade CAR T cell products for patient administration. Of note, HPL, which is derived from multiple transfusable donors’ platelets, was initially developed to support the ex vivo expansion of MSCs for clinical use in a spectrum of autoimmune diseases including GVHD , Crohn’s disease , amyotrophic lateral sclerosis  and multiple sclerosis . However, to the best of our knowledge, our study is the first to evaluate the effect of HPL-exposure on CAR T cell phenotype or function.
A number of groups have conducted clinical trials using second generation CARs expressing either CD28 or 41BB costimulatory endodomains and correlative studies have demonstrated that the incorporation of CD28 enhances cytotoxicity but is associated with diminished T cell persistence when compared with 41BB [42, 43]. Thus, in the current study we chose to focus on enhancing the longevity of CAR.CD28 T cells using serum supplementation. Using unmodified T cells and T cells modified with two different CAR constructs (P28z and 1928z) we found that cells expanded in HPL contained a significantly higher percentage of less differentiated T cells according to CCR7 expression. Although other makers (e.g. CD62L, CD27 and CD127) frequently used to define memory phenotypes were not different across the serum conditions it should be noted that CCR7 expression (and its associated gene expression profile signature) was not solely responsible for the enhanced anti-tumor effects seen in HPL-exposed cultures given that knocking out this gene did not diminish the effector function of HPL-exposed cells (Additional file 2: Figure S5 - CCR7KO HPL-cultured T cells). Instead, it appears that the less differentiated profile of HPL-exposed P28z T cells, as shown in RNAseq analysis and in vitro proliferation assays, is key in promoting enhanced in vivo anti-tumor effects.
To identify which factor(s) in HPL are responsible for the impact on the T cell differentiation profile we performed proteomic analysis, comparing soluble protein(s) contained in HPL and ABS. However, given the complexity of this serum supplement such assessments have proven challenging. For example, we found that 69 of the 640 proteins assessed were differentially up- or down-regulated by > 10-fold between the two sera . This included TGFβ, which was present in 25-fold greater levels in HPL and importantly has been reported to prevent T cell differentiation [34, 35, 36]. In our study we confirmed this finding using recombinant TGFβ1, as shown in Additional file 2: Figure S7a. However, TGFβ1 is immunosuppressive to T cells [44, 45], as highlighted by the detrimental effect of exogenous TGFβ1 on the cytolytic capacity of our CAR T cells (Additional file 2: Figure S7b and S7c), suggesting that the phenotypic and functional characteristics of HPL-exposed CAR T cells is likely a result of multiple soluble proteins, of which TGFβ1 may be one.
The CAR T cell fields are rapidly growing with the effort to enhance their potency with additional genetic modifications (cytokine, cytokine receptor, switch receptor). The discovery and utilization of gene editing techniques such as ZFN, TALEN and CRISPR/Cas9 should further accelerate the development of new generation of CAR T cells (knock out inhibitory receptor [46, 47, 48], knock in CAR into specific locus [49, 50]). Our study highlights the importance of essential culture supplementation in order to improve CAR T cell manufacturing without additional gene modifications. Optimum serum choice can provide improved cellular phenotype for infusion products that may further be improved with the continued advancements in CAR T cell engineering.
This work was supported by Cook Regentec via a Sponsored Research Agreement with Baylor College of Medicine. We are grateful to Texas Children’s Hospital Small Animal Imaging Facility, Texas Children’s Hospital Flow Cytometry Core Laboratory, and the support of Cell Processing and Vector Production Shared Resource core in the Dan L. Duncan Comprehensive Cancer Center.
NW and JV conceived the study and designed the experiments. ATC, NW, MKM, EC and CTD performed experiments and analyzed data. CAR and PL provided material support. NW and AML wrote the manuscript. MKM, EC and CTD critically reviewed the manuscript. All authors read and approved the final manuscript.
This work was supported by Cook Regentec via a Sponsored Research Agreement with Baylor College of Medicine.
Ethics approval and consent to participate
Collection of human peripheral blood mononuclear cells (PBMCs) were obtained from healthy donors and patients after informed consent on protocols approved by the Institutional Review Board (IRB) at Baylor College of Medicine (H-15152, H-27471, H-19384 and H-31970). Mice were housed and treated in accordance with Baylor College of Medicine Animal Husbandry and Institutional Animal Care and Use Committee guidelines (AN-5639).
Consent for publication
Emanuele Canestrari and Christina T. Dann are employees of Cook Regenetec. Ann M. Leen and Juan F. Vera are consultants and have ownership interests (including stock and patent) in Marker Therapeutics, Inc. and Allovir. The other authors declare that they have no competing interests.
- 24.Klebanoff CA, Crompton JG, Leonardi AJ, Yamamoto TN, Chandran SS, Eil RL, et al. Inhibition of AKT signaling uncouples T cell differentiation from expansion for receptor-engineered adoptive immunotherapy. JCI insight. 2017;2.Google Scholar
- 37.Bollard CM, Tripic T, Cruz CR, Dotti G, Gottschalk S, Torrano V, et al. Tumor-specific T-cells engineered to overcome tumor immune evasion induce clinical responses in patients with relapsed Hodgkin lymphoma. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2018;36:1128–39.CrossRefGoogle Scholar
- 38.Introna M, Lucchini G, Dander E, Galimberti S, Rovelli A, Balduzzi A, et al. Treatment of graft versus host disease with mesenchymal stromal cells: a phase I study on 40 adult and pediatric patients. Biology of blood and marrow transplantation : journal of the American Society for Blood and Marrow Transplantation. 2014;20:375–81.CrossRefGoogle Scholar
- 39.Dhere T, Copland I, Garcia M, Chiang KY, Chinnadurai R, Prasad M, et al. The safety of autologous and metabolically fit bone marrow mesenchymal stromal cells in medically refractory Crohn's disease - a phase 1 trial with three doses. Aliment Pharmacol Ther. 2016;44:471–81.PubMedCrossRefPubMedCentralGoogle Scholar
- 41.Dahbour S, Jamali F, Alhattab D, Al-Radaideh A, Ababneh O, Al-Ryalat N, et al. Mesenchymal stem cells and conditioned media in the treatment of multiple sclerosis patients: clinical, ophthalmological and radiological assessments of safety and efficacy. CNS neuroscience & therapeutics. 2017;23:866–74.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.