Constructing a Novel CAR T-Cell with Anti-Angiogenic Properties for Acute Myeloid Leukemia

​

By Gideon Jancu

Science Research Honors Program, Archbishop Molloy High School

S.U.N.Y. Albany, Science Research in the High School Program

​

Abstract

​

In immuno-oncology, Chimeric Antigen Receptor (CAR) T-cell therapy has become a mainstream field of research. As a type of immunotherapy, it started as a treatment option of last resort, and now is used as a first-line regimen. In the context of Acute Myeloid Leukemia (AML), there have been special roadblocks because the antigens of the AML cells are very heterogeneous, and engineering a receptor for each mutated antigen in a laboratory using autologous cells would be extremely difficult. In addition, the myeloids (precursors to red and white blood cells, and platelets) are extremely toxic, and bone marrow ironically can suppress the immune system once it becomes part of a tumor microenvironment. This paper proposes that to have a maximum effect, CAR therapy should have a dual approach: antigen receptors that are better able to recognize the cancerous antigens, plus the ability to attack the blood vessels that feed the tumor microenvironment (TME) of AML, i.e., have an anti-angiogenic effect. This paper introduces an experimental method for engineering a new CAR T-cell that has the dual function of recognizing AML-specific antigens while delivering an anti-angiogenic parallel attack.

One difficulty to overcome is the preservation of normal hematopoietic function, in other words, the normal process of blood vessel creation when anti-angiogenics are introduced specifically to attack blood vessels. This article proposes a multi-domain CAR design that has an anti-VEGFR2 single-chain variable fragment (scFv), plus a regular AML-specific CAR form.  It will express a soluble endostatin to modulate autocrines. Ultimately, this article proposes pathways for a hybrid approach.

​

1. Introduction

Acute Myeloid Leukemia (AML) is essentially the clonal expansion of the precursors of myeloids.  AML can result in the failure of bone marrow, which can have cascading effects in the body.1  Some treatments, like chemotherapy and the transplant of hematopoietic stem cells, do not prevent significant relapse or refractory rates.2 While immunotherapy has shown promise for many cancers, AML presents particular challenges.  One of these is shared antigen expression between cancer cells and regular myeloid cells.  Another is a TME that is very angiogenic, given that this is bone marrow which is heavily involved in the creation of vasculature.3

It is well-established that angiogenesis is indeed a major reason that AML is difficult to defeat. Vascular endothelial growth factor (VEGF) and VEGF receptor (VEGFR), which spur the creation of blood vessels, are heavily present in the AML TME.4  This is why a hybrid approach of CAR and anti-angiogenics with synergistic effects could be successful.  We will use the term Anti-Angiogenic CAR T (AA-CAR-T) for a method that has both the regular effect of CAR, which is to recognize cancer antigens and introduce cytotoxicity, with anti-angiogenics as a parallel weapon.

​

2. The Angiogenic Microenvironment in AML

In terms of angiogenesis in AML, there is an excess VEGF-A, vascular endothelial growth factor A, which is a protein that plays a crucial role in forming and maintaining blood vessels. There is also an excess of basic fibroblast growth factor (bFGF), and angiopoietin-2, which are angiogenics.  They are secreted by blasts of leukemia and stromal cells.5 (A blast of leukemia is a young, irregular white blood cell which proliferates in the bone marrow; stromal cells are a tissue cells that help structure organs, promote tissue homeostasis, and keep immunity strong.)  bFGF and angiopoietin-2 together can contribute to abnormal blood vessels in marrow.  That supports leukemia, and also makes it resist the usual treatments like chemotherapy.6

Endothelial cells form the inner lining of all blood vessels.  They are a threshold for the exchange of fluids, gases, and immune cells for blood, on the one side, and surrounding tissues, on the other.  They also are heavily responsible for clotting. With respect to AML, they help cancer cells evade the immune system by expressing molecules like PD-L1 (a protein that acts as a checkpoint against immunity, which in normal circumstances is helpful to prevent the immune system from attacking the body), and secreting TGF-β (which is a group of proteins that regulate immunity).7  Essentially, the cancer cells cloak themselves in proteins that work to defeat immunity. When anti-angiogenesis is combined with CAR, cancerous endothelial cells which cloak themselves in those proteins could be defeated at the source.8

​

3. CAR T-Cell Therapy in AML: Challenges and Opportunities

In B-cell malignancies, there is one targetable antigen that can distinguish between malignant and normal hematopoietic cells. B-cells are white blood cells that latch onto and neutralize antigens on targets like bacteria and viruses by becoming plasma and releasing antibodies, the proteins that attack invaders.  AML, on the other hand, does not have a universally targetable antigen – AML is too heterogeneous, as mentioned -- and therefore cannot be targeted in the same way.

Experiments have looked at the suitability of different targets for CAR therapy.  For instance, CD33 has been a common candidate.  CD33 is a protein on the surface of cells like myeloids, including AML myeloids. It is considered a strong biomarker for AML, and has been targeted often in CAR T-cell therapy. 

Another candidate has been FLT3 (also called FMS-like tyrosine kinase 3).  It is a gene that sends instructions to make a protein for forming blood stem cells.  FLT3 mutations are actually the most common mutations in AML, spurring growth of white blood cells that cannot be contained. 

Yet another candidate has been CD123. This is part of the interleukin-3 receptor, which is a protein on the surface of healthy cells, but becomes overexpressed on blood cancers like AML. Interleukin-3 is a cytokine, or messenger protein, that stimulates the bone marrow to create blood cells in the bone marrow.  It has been used as an immunotherapeutic target since it is on cancer cells and cancer stem cells, even though it is also on normal cells.

A fourth candidate has been CLL-1.  This is a protein that is highly expressed on AML, but like the other candidates, is also expressed on blood cell precursors, which makes it risky as a target because of the potential for myelosuppression.9  CLL-1 nevertheless serves as an important biomarker in AML.

Besides the problem of these target antigens also being present in normal cells, and therefore carrying the risk of degrading whatever healthy cells exist around the TME, AML cells in general often occur in a part of the marrow that is particularly resistant to T-cells, where their effects are not especially strong to begin with.10

To overcome some of the obstacles, CAR T-cell strategies in AML have sometimes tried to target more than one antigen at once, suicide-switch integration (a kill switch to permanently disable a system), and “armored” CARs secreting cytokines such as IL-12 or IL-18.11

The genetic engineering of an “armored” CAR is an incredible leap forward in CAR design. The first generation of CAR worked by engineering new receptors on chimeric (lab made, artificial) T-cells to bind to the antigens of target cancer antigens.  “Armored” CARs have three advancements.  First, they secrete cytokines like interleukin, which interfere with immunosuppressive signals in the TME.  Second, they secrete ligands (messengers between cells that bind to specific proteins) like CD40L, and CD40L activates immune cells in the TME itself.  Third, they secrete proteins that act like antibodies to block immune checkpoints. (Immune checkpoints normally tamp down the immune response so as to prevent auto-immune diseases, but in malignant situations they must be de-activated somewhat for immunotherapies to have greater effect.)  

My proposal, which I am calling AA-CAR T-cells, builds on the idea of “armored” CARs by adding angiogenesis inhibitors to the mix: targeting the angiogenic receptors, so that “armored” CARs with triple-functionality, can now exhibit a fourth functionality, anti-angiogenics, to attack the leukemic vasculature that feeds the TME.

​

4. Conceptual Design of an Anti-Angiogenic CAR T-Cell

 

4.1 Dual Antigen Targeting

The AA-CAR T-cell incorporates two distinct antigen-recognition domains:

1. AML-specific CAR domain: directed against leukemia antigens like CLL-1 or CD123 (discussed above).

2. Anti-angiogenic domain: an antibody fragment (also called scFv) that recognizes VEGFR2 (discussed above) which for these purposes is a receptor that gets overexpressed on cancerous endothelial cells of the AML marrow TME.12

A CAR “domain” is a synthetic, lab-made receptor on a CAR, engineered so that T-cells recognize and attack cancer cells. Each CAR has a number of different domains: (1) an extracellular domain, that binds to specific cancer antigens, (2) a transmembrane domain, that anchors it to the cell, and (3) an intracellular signaling domain, that activates the immune cell to destroy the cancer cell. 

For the anti-angiogenic domain, an scFV (a single chain variable fragment), is an antibody fragment that is only about one-sixth the size of a regular antibody, but is preferable here because they are not just smaller but also more flexible and easier to engineer than full antibodies. Another advantage is that by being smaller, they can combine better with other molecules, and they can also penetrate into a tumor more easily for targeted delivery.

​

4.2 Autocrine Expression of Soluble Endostatin

As the word “auto” in autocrine suggests, an autocrine is a form of cell communication in which a cell releases a signal to itself to bind to receptors on the same cell. While there are various types of messengers among cells, autocrines give their own cells instruction.  The autocrine of endostatin is necessary for this to work.  Endostatin is a protein fragment that inhibits angiogenesis. The CARs have to be engineered to secrete soluble endostatin.  It will limit the proliferation and migration of cancerous endothelial cells and VEGF signalling.13  The secretion of endostatin could get into the marrow where it would limit the toxicity to blood vessels outside the TME (avoiding off-target problems for healthy cells). It must be admitted that some studies have shown the difficulties in engineering CARs to carry soluble endostatin in this manner.

​

4.3 Signaling Domains

Another way to make this more effective is to put together CD28 and CD137 in the CAR.  CD28 is a protein on the T-cells that is an antigen receptor. They are necessary to activate T-cells, and to make them proliferate.  CD137 is a protein receptor on T-cells that have been activated. It also increases T-cell proliferation. 

Both of those working together give a significant boost a T-cell response called CD3ζ.14 T-cell activation requires CD3ζ because it is the main means of signaling and translating antigen recognition into intracellular signals.  So, without CD3ζ, the assembly of the T-cell receptor (TCR) could not happen.  The laboratory is where TCRs are made, and these receptors, which are proteins on the surface of T-cells, are engineered to recognize and bind to cancer cell antigens.

​

5. Mechanistic Rationale and Expected Benefits

Once those components are in place, the proposed AA-CAR T-cell would work through a  number of different mechanisms.

1. Direct cytotoxicity.  Like in regular CAR T-cell function, once the CAR T-cells recognize the target antigens, those blasts (the immature, white blood cells in the marrow) would be flooded with cytotoxicity. 

2. Targeted endothelial disruption. The VEGF (discussed above: vascular endothelia growth factor) would be reduced so as to hamper microvascular proliferation.

3. Microenvironment remodeling. The angiogenic cytokines (protein messengers to create blood cells) are downregulated (the stimulus response is reduced by decreasing receptors for signals to create marrow vasculature).

4. Enhanced T-cell infiltration. Once there is less angiogenic signaling, there is also less resistance to the presence of immunotherapeutic cells like the scFV antibody fragments.

Ideally, these combined mechanisms for malignant cell destruction and the suppression of the vasculature in the TME, would be twice as powerful as either strategy on its own.

​

6. Preclinical and Translational Considerations

​

6.1 Safety and On-Target, Off-Tumor Effects

As mentioned before, many of the targets are also expressed on healthy cells, so there is the possibility of off-tumor effects, similar to how an overactive immune system can result in auto-immune disease.  For example, there is a great risk in targeting VEGFR2, which normally works to manufacture blood vessels through the endothelial cells, because the off-tumor effects could be quite serious. There are several ways to potentially limit or prevent off-target vasculature toxicity.

• Suicide genes.  These are genes that make cells vulnerable to substances that usually are not toxic, like certain drugs which could be used to activate cell death, or, in other cases, that themselves contain a toxic substance it releases for programmed cell death (apoptosis, which occurs in all cells throughout the body anyway on an ongoing basis).  These methods could eliminate CAR T-cells that become dangerous.15

• Tumor-restricted promoters.  These control gene expression in cells that are normally cancer-free. A therapeutic gene under the control of a tumor-restricted promoter will only become active in the cells where the promoter is present.  In other words, the T-cells and scFV’s would not even be activated except in response to a cytokine environment with AML.16

• SynNotch circuits.  A SynNotch circuit is a synthetic biological system that uses a Notch receptor (a protein that emits signals for cell development, differentiation, proliferation, and survival) to make T-cells perform specific functions based on their environment.  The receptor detects an antigen on a target cell, which triggers a chain reaction that releases a therapeutic molecule or creates a different receptor.  In practice, SynNotches activate T-cells only in the presence of two specific tumor antigens.  That limits the danger of T-cells affecting anything other than their targets. These could be engineered to have anti-angiogenic effects only in the presence of AML antigens.17

 

6.2 Efficacy Testing

In terms of testing, there would have to be murine (mouse) testing first.  The mice would first be genetically engineered to have a more human-like response, and then AML would be xenografted into their marrow to test any dual-target effects of AA-CAR T-cells.18  Besides imaging tests on vascular density, flow cytometry would be used to identify and count different cell types and keep track of T-cell persistence.  Human clinical trials would take place later.

​

6.3 Safety Testing

Because AA-CAR T-cells have never been manufactured before, there would have to be significant clinical evaluation of safety first. The vascular system, and various biomarkers of vascular health (like VEGF levels throughout other parts of the body), would have to be closely monitored. Also, transposon effects would have to be monitored.  The transponson process is the migration of some parts of DNA to another gene (they are often called “jumping genes”).  If this effect is detected, it means that there is a serious disruption in the genome.19

​

7. Conclusion

Combining anti-angiogenics into CAR technology in the specific way I am proposing could open  new avenues of research. Just as important to the proposal for AA-CAR T-cells, programmed to   simultaneously attack AML and disable leukemic vasculature, is the potential for limiting off-tumor toxicities which has been one of the main drawbacks of immunotherapy. Because the proposal is a departure from current approaches, there would need to be a high degree of preclinical testing. However, the idea of a combined therapy with AA-CAR T-cells, whether through the mechanisms I have proposed or using other pathways, could itself be a starting point for a completely new approach to hematologic immunotherapy.

​

References

Endnotes

​

1. Chen Y, Hoffmeister LM, Zaun Y, Arnold L, Schmid KW, Giebel B, Klein-Hitpass L, Hanenberg H, Squire A, Reinhardt HC, Dührsen U, Bertram S, Hanoun M. Acute myeloid leukemia-induced remodeling of the human bone marrow niche predicts clinical outcome. Blood Adv. 2020 Oct 27;4(20):5257-5268. https://doi: 10.1182/bloodadvances.2020001808. PMID: 33108453; PMCID: PMC7594397.

2. Thol F, Ganser A. Treatment of Relapsed Acute Myeloid Leukemia. Curr Treat Options Oncol. 2020 Jun 29;21(8):66. https://doi: 10.1007/s11864-020-00765-5. PMID: 32601974; PMCID: PMC7324428.

3. Litwin C, Leong KG, Zapf R, Sutherland H, Naiman SC, Karsan A. Role of the microenvironment in promoting angiogenesis in acute myeloid leukemia. Am J Hematol. 2002 May;70(1):22-30. https://doi: 10.1002/ajh.10092. PMID: 11994978.

4. de Bont ES, Fidler V, Meeuwsen T, Scherpen F, Hählen K, Kamps WA. Vascular endothelial growth factor secretion is an independent prognostic factor for relapse-free survival in pediatric acute myeloid leukemia patients. Clin Cancer Res. 2002 Sep;8(9):2856-61. PMID: 12231527.

5. Trujillo A, McGee C, Cogle CR. Angiogenesis in acute myeloid leukemia and opportunities for novel therapies. J Oncol. 2012;2012:128608. https://doi: 10.1155/2012/128608. Epub 2011 Sep 5. PMID: 21904549; PMCID: PMC3167188.

6. Karajannis MA, Vincent L, Direnzo R, Shmelkov SV, Zhang F, Feldman EJ, Bohlen P, Zhu Z, Sun H, Kussie P, Rafii S. Activation of FGFR1beta signaling pathway promotes survival, migration and resistance to chemotherapy in acute myeloid leukemia cells. Leukemia. 2006 Jun;20(6):979-86. https://doi: 10.1038/sj.leu.2404203. PMID: 16598308.

7. Gulley JL, Schlom J, Barcellos-Hoff MH, Wang XJ, Seoane J, Audhuy F, Lan Y, Dussault I, Moustakas A. Dual inhibition of TGF-β and PD-L1: a novel approach to cancer treatment. Mol Oncol. 2022 Jun;16(11):2117-2134. https://doi: 10.1002/1878-0261.13146. Epub 2022 Jan 4. PMID: 34854206; PMCID: PMC9168966.

8. Xing H, Yang X, Xu Y, Tang K, Tian Z, Chen Z, Zhang Y, Xue Z, Rao Q, Wang M, Wang J. Anti-tumor effects of vascular endothelial growth factor/vascular endothelial growth factor receptor binding domain-modified chimeric antigen receptor T cells. Cytotherapy. 2021 Sep;23(9):810-819. https://doi: 10.1016/j.jcyt.2021.05.008. Epub 2021 Jul 7. PMID: 34244079.

9. Watarai M, Miwa H, Shikami M, Sugamura K, Wakabayashi M, Satoh A, Tsuboi K, Imamura A, Mihara H, Katoh Y, Kita K, Nitta M. Expression of endothelial cell-associated molecules in AML cells. Leukemia. 2002 Jan;16(1):112-9. https://doi: 10.1038/sj.leu.2402326. PMID: 11840270.

10. Xu, B., Zhou, Z., Wen, Y. et al. The immunometabolic landscape of the bone marrow microenvironment in acute myeloid leukemia. Exp Hematol Oncol 11, 81 (2022). https://doi.org/10.1186/s40164-022-00332-8.

11. Minagawa K, Al-Obaidi M, Di Stasi A. Generation of Suicide Gene-Modified Chimeric Antigen Receptor-Redirected T-Cells for Cancer Immunotherapy. Methods Mol Biol. 2019;1895:57-73. https://doi: 10.1007/978-1-4939-8922-5_5. PMID: 30539529.

12. Ballmer-Hofer, K., A.C. Hyde, C., Schleier, T., & Avramovic, D. (2018). ScFvs as Allosteric Inhibitors of VEGFR-2: Novel Tools to Harness VEGF Signaling. International Journal of Molecular Sciences, 19(5), 1334. https://doi.org/10.3390/ijms19051334.

13. Shichiri M, Hirata Y. Antiangiogenesis signals by endostatin. FASEB J. 2001 Apr;15(6):1044-53. doi: 10.1096/fj.99-1083com. PMID: 11292666.

14. Degirmencay A, Thomas S, Holler A, Burgess S, Morris EC, Stauss HJ. Exploitation of CD3ζ to enhance TCR expression levels and antigen-specific T cell function. Front Immunol. 2024 May 30;15:1386132. https://doi: 10.3389/fimmu.2024.1386132. PMID: 38873603; PMCID: PMC11169823.

15. Gargett T, Brown MP. The inducible caspase-9 suicide gene system as a "safety switch" to limit on-target, off-tumor toxicities of chimeric antigen receptor T cells. Front Pharmacol. 2014 Oct 28;5:235. https://doi: 10.3389/fphar.2014.00235. PMID: 25389405; PMCID: PMC4211380.

16. Montaño-Samaniego M, Bravo-Estupiñan DM, Méndez-Guerrero O, Alarcón-Hernández E, Ibáñez-Hernández M. Strategies for Targeting Gene Therapy in Cancer Cells With Tumor-Specific Promoters. Front Oncol. 2020 Dec 14;10:605380. https://doi: 10.3389/fonc.2020.605380. PMID: 33381459; PMCID: PMC7768042.

17. Simon S, Bugos G, Salter AI, Riddell SR. Synthetic receptors for logic gated T cell recognition and function. Curr Opin Immunol. 2022 Feb;74:9-17. https://doi: 10.1016/j.coi.2021.09.003. Epub 2021 Sep 24. PMID: 34571290; PMCID: PMC8901444.

18. Antonella Antonelli, Willy A. Noort, Jenny Jaques, Bauke de Boer, Regina de Jong-Korlaar, Annet Z. Brouwers-Vos, Linda Lubbers-Aalders, Jeroen F. van Velzen, Andries C. Bloem, Huipin Yuan, Joost D. de Bruijn, Gert J. Ossenkoppele, Anton C. M. Martens, Edo Vellenga, Richard W. J. Groen, Jan Jacob Schuringa; Establishing human leukemia xenograft mouse models by implanting human bone marrow–like scaffold-based niches. Blood 2016; 128 (25): 2949–2959. https://doi: https://doi.org/10.1182/blood-2016-05-719021.

19. Hackett PB, Largaespada DA, Switzer KC, Cooper LJ. Evaluating risks of insertional mutagenesis by DNA transposons in gene therapy. Transl Res. 2013 Apr;161(4):265-83. https://doi: 10.1016/j.trsl.2012.12.005. Epub 2013 Jan 10. PMID: 23313630; PMCID: PMC3602164.