A Novel Method for Highly Sensitive Detection and Isolation of Chimeric Antigen Receptor and Cancer Associated Antigen-Expressing Cells

Chimeric antigen receptor (CAR)-based cellular therapy is a revolutionary approach to treat cancer as witnessed by recent success in clinical trials for various hematological malignancies. Currently, flow cytometry based detection of fluorochrome-tagged antibodies or proteinL that binds to the Extra-Cellular Domain (ECD) of CAR molecule are the widely used methods for the detection of CARexpression. Here, we have developed a novel luciferase based assay for detecting the expression of CAR. Our assay is accurate, highly sensitive (10^-5), and has a broad linearity by taking advantage of the extreme brightness of recently discovered marine luciferases (Gluc/Nluc/Tluc16/Mluc/Loluc/Paluc/Htluc). The assay is based on recombinant fusion protein technology by fusing the ECD of a CAR target in frame with one of the marine luciferases (for detection) along with several small peptide tags -Flag/ Strep -tag II/AcV5/His (for isolation). Initially, a fusion construct was made by cloning the ECD of CD19 fused in frame with Nluc. The fusion protein was produced using 293FT cells, and tested by a simple binding assay that involved 45 minutes incubation at 4^oC followed by washing and detection of luminescence. More than 10³ fold increases in luminescence was observed between FMC63-CARtransduced-T/NK cells and uninfected cells or a non-specific CAR transduced cells. Essentially, identical results were obtained by replacing Nluc with other marine luciferases or by using cells transduced with five distinct CARs targeting CD19. Similar strategy was successfully applied for the specific detection of CARs targeting CD20, CD30, CD33, CD123, CD138, BCMA, and SLAM7. We also show that the small peptide tags in the fusion protein can be used for the specific isolation of CAR^+ve cells using anti-tag antibodies by FACS. Additionally, we purified ECD-fusion proteins for CD19 and CD33 using Strep -Tactin protein purification columns. Purified fusion proteins were fully active, as observed by successful and specific binding to respective CAR-T/NK cells. Furthermore, direct conjugation of purified ECD-fusion proteins with fluorochromes resulted in a single-step detection and/or isolation of CAR^+vecells.

In another embodiment, an antigen positive cell can be detected by fusing the single-chain fragment variable (scFv) of its specific monoclonal antibody with one of the marine luciferases. A fusion construct was made by cloning scFv of a CD19 antibody clone (FMC63) fused in frame with Nluc, followed by the fusion protein production and binding assay. Again, a 10³ fold increase in luminescence was observed between CD19^+ve cell lines (Raji, Nalm6, and BV173) and CD19^-ve cell lines (HL60 or Raji-CD19^-/- by Crispr-Cas9). Similar results were obtained using other marine luciferases in place of Nluc or by replacing FMC63 with five distinct scFv targeting CD19. Next, we made fusion proteins using scFv of antibody clones targeting CD20, CD30, CD33, CD123, CD138, BCMA, and SLAM7 and tested by binding assay using a panel of cell lines comprising 8 hematological malignancies. Strikingly, all the scFv fusion proteins bound strictly to their respective antigen^+ve cell lines.

Until now, only a limited number of targets have been pursued using CAR technology, which can be attributed to the lack of a High Throughput Screening (HTS)-based approach to find novel targets. To check the compatibility of our scFv-luciferase fusion protein technology as a HTS approach to discover novel CAR targets, a library of 78 distinct scFv-Nluc constructs were made, followed by the production of fusion proteins. Binding assays were performed in 96-well plates using Raji, BV173, MM1S, and U266 cell lines. Fusion proteins which are supposed to bind their respective targets were readily identified by the assay - Raji (CD19 and CD20 targeting scFv); BV173 (CD19 and CD33 targeting scFv); MM1S and U266 (CD138, BCMA, and SLAM7 targeting scFv). Most importantly, few novel CAR targets were identified thereby opening up the possibility of applying this screening approach to patient tumor cells, that may result in precision medicine model in cellular-therapy space by tailoring CAR therapy for individual patients. Finally, we demonstrate the utility of scFv-luciferase fusion protein technology in accurately detecting the serial engraftment of B-ALL and AML-patient derived xenografts in NSG mice using scFv of antibodies targeting CD19 and CD33, respectively.