Honing in on the Tumor Microenvironment

Chimeric antigen receptor T cells (CAR-T) are hot in the biotech world, and the science community has been tinkering with the constructs to improve their efficacy in real-world settings. One of these parameters is manufacturability; allogeneic cells are ideal for large-scale manufacturing and to reduce some of the heterogeneity that is associated with autologous T-cell therapies.

BioPharm International recently spoke to Alexandre Juillerat, PhD, innovation senior scientist at Cellectis, to find out how Cellectis’ CAR construct differs from others in development and how the company’s novel approach to T-cell architecture allows it to be eligible for future improvements in safety and/or efficacy. For background information on the new CAR-T, please read “Cellectis Describes Novel CAR-T Design in New Paper.”

BioPharm: The extensive expansion of T-cell therapies carries the risk of cell senescence. How would Cellectis overcome this challenge in the manufacture of its allogeneic cells?
Juillerat: The cells we are using are from healthy donors. They are not exhausted because of chemotherapy, for example.

BioPharm: The UCART123 investigational agent is said to be the first investigational new drug (IND) filing for a gene-edited T-cell product in the US. What about worldwide? Are there others that are first elsewhere
Juillerat: UCART123 is the first IND filing for a gene-edited CART-cell product in the US. Sangamo filed an IND for clinical trials in HIV/AIDS. Their product, a gene-edited T-cell product, is currently being tested in Phase I/II in the US. In addition to the IND filed for UCART123, we filed CTAs [clinical trial authorization] (equivalent of an IND in the UK) for UCART19. To our knowledge, there is no other IND filed for gene-edited T-cell product. Editor’s note: There may be an IND from Sangamo on the horizon. In a 2011 10-K, Sangamo said the company’s focus is the development of zinc finger DNA-binding protein (ZFP) Therapeutics. Their lead ZFP Therapeutic, SB-728-T, is a ZFN-modified autologous T-cell product for the treatment of HIV/AIDS. In a January 11, 2017 press release, Sangamo wrote, “We are also moving quickly to advance additional ZFP Therapeutic programs into the clinic with the target of filing six more IND applications over the next twelve months."

BioPharm: Would the oxygen-specific construct in your recent study be considered a "bispecific" construct? Why or why not?
Juillerat: 'Bispecific' T-cells are mainly described as engineered T-cells that can recognize two different antigens to allow improved discrimination between healthy tissues and tumors. However, such approaches may suffer from the difficulty to select combinations of suitable target antigens (e.g., avoiding the so-called antigen escape). In our recent paper, we replaced the requirement of finding a second antigen by focusing on the tumor microenvironment. By analogy to other recent synthetic biology strategies based on the use of exogenous small molecule drugs to control CAR T-cell, our oxygen-sensitive system would rather be considered a dual-input system (one antigen-based input and one microenvironment-based input). It should also be noted that our approach can be combined to 'bispecific' strategies to create additional layers of safety.

BioPharm: How many subdomains were added to the CAR constructs, and what are they? (i.e., were there two inputs on one domain)?
Juillerat: On the construct (the chimeric antigen receptor), we have developed contains four functional subdomains: an antigen-targeting domain, a T-cell activation domain, a T-cell co-stimulation domain, and an oxygen-sensitive domain. One of the innovations of our CAR resides in the fact that these domains are spread between three chains of a multichain receptor instead of being piled one after the other in more classical CAR constructs. Such a novel approach allows more freedom in the addition of new functionalities. We also combined the two inputs (antigen-recognition and microenvironment-sensing) on the same chain to strengthen the requirement of having the two functional inputs.