Enabling the Rapid Generation of Allogeneic Artificial Antigen Presenting Cell (aAPC) Red Cell Therapeutics with a Loadable MHC System
Christopher L. Moore, Sneha Pawar, Mellissa Nixon, Timothy J. Lyford, Douglas C. McLaughlin, Shamael R. Dastagir, Christopher L. Carpenter, Thomas J. Wickham, and Tiffany F. Chen
Rubius Therapeutics, Cambridge, MA, USA
Poster #B026
RESULTS AND METHODS
Figure 9. Loadable MHC Platform Can Be Extended to MHC class II
Wild-type domains of MHC class II (α and β β) can be genetically encoded as a single-chain fusion with GPA to achieve cell surface expression (A). Expression of single chain versions of human MHC class II constructs was detected by performing flow cytometry on K562 cells transduced with lentivirus encoding–associated constructs and subsequently stained with fluorescent primary antibody specific to HLA-DR (B) or HLA-DP (C).
α = MHC class II alpha ectodomain; β = MHC class II beta ectodomain; DP4 = allele of human MHC class II DP isotype; DRB1 = allele of human MHC class II DR isotype; FSC = forward scatter; GPA = transmembrane domain of glycophorin A; HLA = human leukocyte antigen; HLA-DP =human leukocyte antigen – DP isotype; HLA-DR = human leukocyte antigen – DR isotype; MHC = major histocompatibility complex; RCT = Red Cell Therapeutic.
• The loadable aAPC platform can be developed to utilize signaling mediated by MHC class II
CONCLUSIONS
• Loadable MHC class I and class II molecules can be robustly expressed on the
RCT cell surface
• In the presence of signals 2 and 3, RCTs with loadable MHC can significantly
expand antigen‑specific T cells in a peptide‑ dependent manner
• Rubius Therapeutics’ loadable aAPC system can be applied to produce aAPC
populations presenting multiple antigenic peptides
• Further development of loadable aAPC system may enable effective personalized
neoantigen therapies
ACKNOWLEDGEMENTSPoster design support was provided by Dennig Marketing Group, sponsored by Rubius Therapeutics. We would like to thank Abigail Bracha and Lori Melançon for their editorial contributions.
DISCLOSURESAll authors: Employment with and equity ownership in Rubius Therapeutics.
Figure 5. A Peptide Can Be Loaded Onto Engineered Empty MHCs
The binding kinetics of a high affinity fluorescent peptide to loadable MHC constructs was measured by varying duration of peptide pulsing prior to washing and detection by flow cytometry (A). Peptides used for pulsing were either internally labeled or externally labeled at the c-terminus (B). The intensity of fluorescent signal on the cells post peptide loading at room temperature and washing was normalized to maximum signal, producing a normalized binding percentage of HPVE7 fluorescently labeled peptide (In and Ex) to MHC constructs (ds HLA-A2 and wt HLA-A2) at 10 ng/ml peptide concentration for different amounts of time (C).
ds = disulfide engineered; Ex = externally labelled; HLA-A2 = human leukocyte antigen A2; HPVE7 = human papillomavirus E7 oncoprotein; In = internally labelled; MHC = major histocompatibility complex; RT = room temperature; wt = wild-type.
• Loadable MHC constructs can bind to exogenous peptide at ambient conditions.
• An internally labeled peptide loads more rapidly than a peptide possessing a C‑terminal extension, suggesting that loadable MHC constructs prefer peptides with an appropriate length
• Peptides load more rapidly on disulfide‑engineered HLA‑A2 compared to wild‑type HLA‑A2
Figure 6. Loadable MHC Constructs Are Stable, Independent of Peptide Pulsing
RCTs expressing unloaded HLA-A2 constructs were pulsed with high affinity peptide from HPVE7 then incubated for increasing amounts of time in media at 37°C prior to measurement by flow cytometry (A). Mean fluorescence intensity (MFI) values for peptide-loaded MHC constructs (ds + peptide and wt + peptide) compared to non-peptide loaded constructs (ds and wt) at different time points over a 10-day period. MFI values were measured via flow cytometry following staining with an antibody specific for HLA-A2 (B).
ds = disulfide engineered; HLA-A2 = human leukocyte antigen A2; HPVE7 = human papillomavirus E7 oncoprotein; MFI = mean fluorescence intensity; MHC = major histocompatibility complex; RCT = Red Cell Therapeutic; RT = room temperature; wt = wild-type.
• The addition of a peptide does not appear to alter the stability of surface-exposed wild‑type of disulfide engineered HLA‑A2 constructs
• Loadable MHC construct expression remains stable for up to 10 days
Figure 3. Rubius Therapeutics Is Developing an aAPC Platform to Be Used With Personalized Neoantigens
• An aAPC approach utilizing loadable forms of the 5 most prevalent human leucocyte antigen (HLA) class I alleles in the U.S. population can cover approximately 70% of patients
• Platform is applicable to develop aAPCs for MHC I and MHC II
AG = antigen; RTX-aAPC = artificial antigen-presenting cell.
Figure 4. Empty Loadable MHC Constructs Can Be Stably Expressed at High Levels on the Cell Surface of an RCT
Wild-type, disulfide-engineered, and peptide-fused MHC constructs for HLA-A2 (α1, α2, and α3) were genetically encoded as single-chain fusions with β-2-microglobulin (β2m) and anchored by glycophorin A (GPA) (A). Expression in the RCT platform was detected by performing flow cytometry on cells stained with a fluorescent primary antibody specific to ββ2m (B) and converting geometric mean fluorescence intensity (MFI) values to copy number via Bangs beads (C).
• Robust expression of empty MHC molecules on the surface of RCTs can be achieved
• Disulfide engineering slightly improves expression compared to wild‑type HLA‑A2, but is lower than a genetic fusion with antigenic peptide (sc‑Trimer)
INTRODUCTION
• Current peptide neoantigen vaccine approaches are promising, but do not adequately stimulate and expand patient T cells to the levels required to achieve robust efficacy
• Rubius Therapeutics has developed allogeneic artificial antigen presenting cell (aAPC) Red Cell Therapeutics™, which express the required signals for complete T cell activation: a tumor‑specific antigen on the major histocompatibility complex (MHC), a co-stimulatory ligand and a cytokine
• To use the aAPC approach with personalized neoantigens, Rubius Therapeutics has developed a loadable MHC system that enables the rapid generation of aAPCs
Figure 1. The RED PLATFORM® Is Designed to Generate Allogeneic, Off-the-Shelf Cellular Therapies
OBJECTIVES
• To engineer loadable MHC class I and class II to achieve robust expression on the cell surface
• To determine whether an empty MHC can be loaded with exogenous peptide
• To determine if RCTs expressing loadable MHC can activate TCRs in a peptide- dependent manner
RESULTS AND METHODS
Figure 2. RTX-aAPC Is a Cellular Therapy That Drives Antigen-Specific Activation and Proliferation of T cells
• Artificial antigen‑presenting cells (RTX‑aAPCs) are engineered to simultaneously express on the cell surface a tumor‑specific antigen on MHC I, a co-stimulatory ligand and a cytokine to mimic the human immunobiology of T cell-APC interactions
MHC = major histocompatibility complex; RTX-aAPC = artificial antigen-presenting cell; TCR = T cell receptor.
Signal 3 cytokine
Signal 2 co-stimulatory agonist
Antigen-specific TCR
MHC I
Signal 1 Tumor antigen
RCT
T CELLS
RED PLATFORM®
• The enucleated reticulocytes are RCTs that express hundreds of thousands of biotherapeutic proteins on the cell surface
• Delivered at a dose of <1% of total red blood cell volume in the body
• Universal, scalable and consistent manufacturing process
EXPANSION & �DIFFERENTIATION
PROGENITOR �CELL COLLECTION
ONE �HEALTHY�O- DONOR
ENUCLEATION & MATURATION
GENETIC �ENGINEERING
100-1000’s �OF DOSES
RED CELL THERAPEUTIC
RTX-aAPC (Loadable) Personal Antigens RTX-aAPC (MULTI-AG)
Library of intermediate aAPCs ready
Peptides synthesized for individual patient
Target multiple tumor-specific antigens
Prototype
ds HLA-A2
wt HLA-A2
sc-Trimer HLA-A2
β2m %
68.9
63.5
84.6
Copy #
99,837
62,748
188,779
C
β2m
FS
Cwt HLA-A2 ds HLA-A2 sc-Trimer HLA-A2
Wild-type single-chain
dimer(wt HLA-A2)
Disulfide-engineered single-chain
dimer(ds HLA-A2)
Single-chain trimer
(sc-Trimer HLA-A2)
GPA
α2 α1
β2mα3
GPA GPA
α2
disulfide antigenic peptide
α1
β2mα3
α2 α1
α3 β2m
BA
1 5 10 15 30 45 900
50
100
HPVE7 Fluorescent Peptide Loading at RT
Minutes
Bin
din
g % ds + HPVE7_Ex
ds + HPVE7_In
wt + HPVE7_Ex
wt + HPVE7_In
HPVE7_In
HPVE7_Ex TPQLDLMY E GG K
N C
N C
TAMRA
TPQLDLMY K
CB
A K562 cellsTransduce with
wt or ds HLA-A2Pulsing with fluorescent
peptide for different amounts of time
Detection via flow cytometry
1 3 5 7 100
5000
10000
15000
20000
Stability Testing for Loaded vs Non-loaded HLA-A2
Days Post-Pulsing
HLA
-A2
MFI ds + peptide
wt + peptide
ds
wt
B
RCT cells expressing wt or ds HLA-A2
HPVE7 peptidepulse (90 min)
Incubate in media for different
periods of time
Wash
Detection via flow cytometry
A
Figure 7. Peptide-Loaded MHC Constructs Can Functionally Engage the TCR
Functional stability for peptide-pulsed HLA-A2 was assessed by TCR activity by pulsing, washing and incubating for increasing amounts of time in media at 37°C for 0, 3, and 6 days incubation (A). Relative luminescence data normalized to untreated Jurkat cells. TCR signaling from peptide-pulsed loadable constructs (ds HLA-A2, wt HLA-A2) and the single-chain trimer control construct (sc-Trimer) at different time periods (B).
ds = disulfide engineered; HLA-A2 = human leukocyte antigen A2; HPVE7 = human papillomavirus E7 oncoprotein; MHC = major histocompatibility complex; RCT = Red Cell Therapeutic; RLU = relative luminescence units; sc = single chain; TCR = T cell receptor; wt = wild-type.
• Loadable MHC constructs demonstrate functional TCR engagement directly after peptide pulsing
• At day 3, a peptide can still be detected functionally for the wt HLA‑A2, but not for the ds HLA‑A2
• By day 6, only the sc‑Trimer is activating the TCR, suggesting that covalent linkage of the peptide prolongs functional stability of MHC constructs
Figure 8. RCTs With Peptide-Loaded MHC Constructs Achieve Robust Expansion of Primary CMV-Specific T Cells
RCT with peptide-pulsed HLA-A2 (wild-type and disulfide engineered, wt and ds, respectively), was assessed for CMV-specific TCR activity and functional expansion of primary T cells by incubating with PBMC from patients exposed to CMV in the presence of RCT expressing 4-1BBL and IL-12 (A). Following 4 hours of incubation, the extent of TCR signaling in each condition was measured by performing flow cytometry on the cell mixture to identify the frequency of T cells demonstrating elevated expression levels of NFAT (B) and Nur77 (C). After 5 days of incubation, staining with fluorescently labelled CMV-HLA-A2 tetramer and measurement by flow cytometry was performed to quantify CMV-specific T cell expansion.
CMV = cytomegalovirus; ds = disulfide engineered; HLA-A2 = human leukocyte antigen A2; hr = hour; MHC = major histocompatibility complex; wt = wild-type; NFAT = nuclear factor of activated T-cells (T cell activation marker); Nur77 = nuclear receptor intracellular transcription factor 77 (T cell activation marker); RCT = Red Cell Therapeutic; UNT = untransduced.
• Loadable MHC constructs can achieve robust signal 1 (TCR) activation
• When utilized in the context of 4‑1BBL and IL‑12, peptide‑pulsed loadable MHC constructs can achieve robust expansion of primary T cells
0 3 60
2
4
6
8
10
Days Post-Pulsing
Fold
Lu
min
esc
en
ce (R
LU)
ds HLA-A2
wt HLA-A2
sc trimer
HPVE7 peptidepulse (90 min)
Incubate in media (days)
Wash
HPVE7 TCRactivation
RCT cells expressing wt, ds or sc-Trimer HLA-A2
B
A
ds HLA-A2 wt HLA-A2 UNT-pulsed0
500
1000
1500
CM
V T
etr
ame
r+ C
ou
nts
CMV-specific T cell expansion(5 days)
Nur77% Upregulation(4 hr)
10:1 2:1 0.4:10
10
20
30
Cell ratio (RCT:T cell)
Nu
r77
%
ds HLA-A2
wt HLA-A2
UNT-pulsed
UNT-nonpulsed
NFAT % Upregulation(4 hr)
ds HLA-A2
wt HLA-A2
UNT-pulsed
UNT-nonpulsed
10:1 2:1 0.4:10
20
40
60
80
Cell ratio (RCT:T cell)
NFA
T %
CMV peptide pulse(90 min)
Wash Add signal 2 and 3RCT cells expressing
wt or ds HLA-A2CMV+ T cell activationA
B
Single-chain dimer MHC II
HLA-DR
FS
C
HLA-DRB1
HLA-DP
FS
C
HLA-DP4
GPA
α βA B C
AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics / October 26-30, 2019 / Boston, MA