Engineering Conferences International

Please click Additional Files below to see the full abstract

Sin, Wei-Xiang

Suhas Jagannathan, Narendra

Kairi, Faris

Sandikin, Dedy

Cheung, Ka-Wai

Lee, Yie Hou

Ram, Rajeev J.

Tucker-Kellogg, Lisa

Birnbaum, Michael E.

HIGH-DENSITY MICROBIOREACTOR PROCESS DESIGNED FOR AUTOMATED POINT-OF-CARE 
MANUFACTURING OF CAR T CELLS 
 
Wei-Xiang Sin, Critical Analytics for Manufacturing Personalized-Medicine (CAMP), Singapore-MIT Alliance for 
Research and Technology Centre (SMART), Singapore 
weixiang.sin@smart.mit.edu 
Narendra Suhas Jagannathan, Cancer and Stem Cell Biology, Duke-NUS Medical School, Singapore 
Denise Bei Lin Teo, CAMP, SMART, Singapore 
Faris Kairi, CAMP, SMART, Singapore 
Dedy Sandikin, CAMP, SMART, Singapore 
Ka-Wai Cheung, CAMP, SMART, Singapore 
Yie Hou Lee, CAMP, SMART, Singapore 
Rajeev J. Ram, CAMP, SMART, Singapore; Department of Electrical Engineering and Computer Science, 
Massachusetts Institute of Technology, USA 
Lisa Tucker-Kellogg, CAMP, SMART, Singapore; Cancer and Stem Cell Biology, Duke-NUS Medical School, 
Singapore 
Michael E. Birnbaum, CAMP, SMART, Singapore; Department of Biological Engineering, Massachusetts 
Institute of Technology, USA 
 
 
Key Words: High-density; CAR T cells; Perfusion; Microfluidic; Bioreactor. 
 
Adoptive cell therapies have revolutionized the landscape of cancer immunotherapy, especially in the treatment 
of hematological malignancies. Despite the clinical success of these immunotherapies, the manufacturing of 
autologous cell therapies remains a complex, multi-step process. On the one hand, research- and preclinical-
scale CAR T cell production still largely rely on fed-batch, manual, open processes that lack environmental 
monitoring and control, which can lead to fluctuations in nutrient availability and waste elimination, as well as to 
process and operator variabilities that may confound process development and prove difficult to translate to 
larger culture systems. On the other hand, clinical- and commercial-scale CAR T cell therapies largely rely on a 
centralized manufacturing model, but low manufacturing throughputs and long turnaround times have hindered 
patient accessibility and treatment efficacy. There are increasing efforts to move towards decentralized and 
automated point-of-care manufacturing of autologous cell therapies in closed systems. However, these 
bioreactors typically have large footprints, operate at large working volumes and require high seeding cell 
numbers. These factors, coupled with limited or costly starting materials and GMP-grade reagents, such as 
PBMCs, viral vectors, medium, serum, and cytokines, make these platforms not well suited for high-throughput 
process optimization and not amenable to efficient scale-out of autologous cell therapy production. There is thus 
a need to intensify the scaling-out of autologous CAR T cell production, in order to reduce the wait times and 
cost of goods of these living drugs. 
 
Here, we present a means of conducting anti-CD19 CAR T cell culture-on-a-chip, using a 2 mL perfusion-
capable microfluidic bioreactor, which incorporates integrated mixers, injectors, and sensors for real-time 
monitoring and closed-loop control of flow rate, optical density, temperature, CO2, dissolved O2, and pH levels. 
We established and assessed a CAR T cell production process on this microbioreactor, with T cell activation, 
lentiviral transduction, and CAR T cell expansion steps all happening within the microfluidic chip, achieving cell 
densities exceeding 150 million cells/mL. This translated to more than 200 million viable CAR+ T cells per chip, 
which met the minimum cell dose of tisa-cel (Kymriah) and exceeded the maximum cell dose of axi-cel 
(Yescarta), hence enabling the production of clinical-scale CAR T cell doses in a reaction vessel the size of a 
deck of cards. The anti-CD19 CAR T cells produced at high densities in the microbioreactor were highly 
functional as measured via cytokine secretion and cytotoxic activity when co-cultured with CD19+ tumor cells in 
vitro. Phenotypically, despite somewhat higher proportions of T cells in the microbioreactor expressing 
differentiation and senescence markers, exhaustion markers were comparable to T cells cultured in a gas-
permeable well. Additionally, continuous sensing of the microbioreactor environment provided information 
complementary to the measurement of metabolite levels in spent media. From this data, computational 
modeling estimated donor-specific rates of expansion and metabolism, thus providing a non-destructive readout 
of cell health and state while ensuring sterility. Thus, the process intensification and miniaturization 
demonstrated by the scaled-down perfusion-capable microbioreactor, with its small working volume and 
footprint, can facilitate high-throughput process optimization studies, as well as enable efficient scale-out of cell 
therapy manufacturing. 


English

High-density microbioreactor process designed for automated point-of- care manufacturing of CAR T cells

Srinivasan, Prasanna

Canova, Christopher

Sangerman, José

Barone, Paul W.

Wolfrum, Jacqueline

Neffling, Milla

Anderson, Ji Young

Miller, Scott

Piras, Graziella

Sinskey, Anthony J.

Braatz, Richard D.

Springs, Stacy L.

Monday, February 5, 2024                                                                                                                        Session 1 
TOWARDS A SCALABLE AAV VECTOR PRODUCTION AT HIGH VOLUMETRIC EFFICIENCY 
 
Prasanna Srinivasan; Massachusetts Institute of Technology, Center for Biomedical Innovation 
prasri@mit.edu  
Christopher Canova; Massachusetts Institute of Technology, Dept. of Chemical Engineering 
José Sangerman; Massachusetts Institute of Technology, Center for Biomedical Innovation 
Paul W. Barone; Massachusetts Institute of Technology, Center for Biomedical Innovation 
Jacqueline Wolfrum; Massachusetts Institute of Technology, Center for Biomedical Innovation 
Milla Neffling; 908 Devices, Boston 
Ji Young Anderson; 908 Devices, Boston 
Scott Miller; 908 Devices, Boston 
Graziella Piras; 908 Devices, Boston 
Anthony J. Sinskey; Massachusetts Institute of Technology, Dept. of Biology 
Richard D.  Braatz; Massachusetts Institute of Technology, Dept. of Chemical Engineering 
Stacy L. Springs; Massachusetts Institute of Technology, Center for Biomedical Innovation 
 
 
Key Words: AAV, Biomanufacturing, Bioprocess, Mechanistic Modeling, Process optimization 
 
Gene therapy has immense potential to revolutionize health care. Adeno-associated virus (AAV) is a commonly 
preferred gene of interest (GOI) carrier because of its non-pathogenicity, broad tropism, ability to engineer its 
properties and established manufacturing platforms. Common approaches to vector production include transient 
transfection of HEK293 cells and Baculovirus infection of insect cells. In particular, the gene therapy market 
segment for treating rare diseases is marginalized because individual therapies targeting a small population 
cannot take advantage of the economies of scale. Furthermore, orphan drugs used for treating rare diseases 
(e.g., Lipodystrophy, hereditary angioedema, etc.) are very expensive and most conditions require prolonged 
period of medication. Although the incidence rate is very low (< 1% of population), rare diseases cause 
significant economic burden on both patients and the health care system. Our goal is to address the 
technological gaps in the rare disease space using engineered devices, mechanistic model predictions, 
metabolic screening and molecular biology approaches to enable cost effective scalable vector production at a 
personalized level. 
 
We had drawn valuable insights to understand the intrinsic limitation of per cell vector genome (vg) titer via 
transient co-transfection of suspended HEK293 cells using PEI mediated delivery of triple plasmids – GOI, Rep-
Cap and Helper plasmids. Our current process typically yields between 3 and 10% of total capsids filled with 
viral DNA (vDNA) cargo. Preliminary findings from our fed-batch shake flask experiments and mechanistic 
modeling elucidated several important aspects that influence vector production such as the plasmid amount 
used for transfection, incompatible kinetics between replication of AAV genome and capsid production, and the 
repression of cap genes by rep proteins. We then developed new experimental strategies to unravel various 
molecular brakes that impede filled capsid production. 
 
We used high-throughput metabolic screening to identify candidate amino acids (AA) that affect vector 
production. Preliminary screening revealed temporally separated Asn and Gln fluxes associated with distinct 
metabolic states that align with saturation of preformed capsids and vDNA replication respectively. More 
comprehensive screening is currently underway to better understand the effects of media components on vector 
production. 
 
We then leveraged our learnings from fed-batch experiments, mechanistic models and metabolic screens to 
develop a rAAV5 vector production process using the Mobius Breez® microbioreactor (a product of Millipore 
Sigma Inc.). Mobius Breez® is a 2mL microbioreactor that enables high cell density culture (>100 million 
cells/mL) in a perfusion mode and is a promising platform to produce a single therapeutic dose at high 
volumetric efficiency. However, a major challenge is to optimize the vector production process at high cell 
densities, which is currently lacking. We demonstrate a new microbioreactor process created by developing new 
transfection strategies and optimizing media formulation to engineer metabolic pathways of cell lines used for 
vector production. Our integrated multidisciplinary approach aims to democratize gene therapy technology and 
accelerate bench-to-bedside clinical translation. 
 
 


Towards a scalable AAV vector production at high volumetric efficiency

Lu, Jiuyi

Monday, February 5, 2024                                                                                                                           Keynote 
TIME IS MUSCLE: PLATFORM ADVANCEMENT FOR SWIFT TIME-TO-MARKET IN AAV 
MANUFACTURING 
 
Jiuyi Lu, Sarepta Therapeutics, USA 
JiLu@sarepta.com 
 


Time is muscle: Platform advancement for swift time-to-market in AAV manufacturing

Masri, Fernanda

Yeago, Carolyn

Maheshwari, Gargi

Moscariello, John

Wang, Bryan

24 conference program - Advancing Manufacture of Cell and Gene Therapies VIII

Cerezo-Magaña, Myriam

CHARACTERIZATION OF TRITERPENE-MEDIATED ENDOSOMAL ESCAPE OF CHOLESTEROL siRNA 
 
Myriam Cerezo-Magaña, Department of Clinical Sciences, Oncology, Faculty of Medicine, Lund University, 
Lund, Sweden.  
myriam.cerezo-magana@med.lu.se 
Hampus du Rietz, Department of Clinical Sciences, Oncology, Faculty of Medicine, Lund University, Lund, 
Sweden. Skane University Hospital, Lund, Sweden. 
Lasse Guericke, Department of Clinical Sciences, Oncology, Faculty of Medicine, Lund University, Lund, 
Sweden. 
Sten Wilhelmson, Department of Clinical Sciences, Oncology, Faculty of Medicine, Lund University, Lund, 
Sweden. Skane University Hospital, Lund, Sweden. 
Anders Wittrup, Department of Clinical Sciences, Oncology, Faculty of Medicine, Lund University, Lund, 
Sweden. Skane University Hospital, Lund, Sweden. 
 
 
Key Words: Endosomal escape; RNA therapeutics; siRNA; Triterpenoids, Ursolic acid. 
 
Endosomal entrapment constitutes a major limitation for the productive biological effect of therapeutic RNAs, 
with most endocytosed molecules estimated to be directed for lysosomal degradation. Therefore, characterizing 
the mechanism of endosomal escape and finding novel compounds that boost RNA cytosolic release is of 
outmost relevance for improved oligonucleotide therapeutics. Here, we evaluated ten triterpenoids and saponins 
i.e. pentacyclic sterol derivates without or with carbohydrate moieties, respectively, for their ability to induce 
endosomal membrane damages. We used high resolution live confocal microscopy and galectin 9 (Gal9) as a 
sensor for endomembrane damages. Initial screening shows varying magnitude of endosomal membrane 
damage induction, with Ursolic acid (UrA) standing out with most Gal9+ membrane damage response. 
Endosomal damages by UrA are rapid, with a robust response at 2 h post-dosing, Interestingly, response to UrA 
also involves recruitment of other damage sensors of the ESCRT pathway, which show even more rapid kinetics 
than Gal 9, appearing after 15-30 min. Functionally, UrA cotreatment with cholesterol siRNA (chol-siRNA) 
targeting eGFP leads to a >10-fold lower IC50 compared to chol-siRNA alone in 2D cancer cell cultures. 
Importantly, knockdown is also improved by 30-70% in 3D spheroid cultures indicating substantially lower IC50. 
Using high speed super resolution live cell imaging, our compartment analysis of newly appearing endosomal 
damages indicate that UrA can target Rab5 positive vesicles, i.e. early compartments where siRNA escape is 
preferred for biological effect, but also later structures e.g. LAMP1-positive lysosomes. Our results contribute to 
better understand UrA response and motivate further characterization of UrA mechanism of action, ultimately 
contributing to the development of more efficient oligonucleotide therapies. 
 


Characterization of triterpene-mediated endosomal escape of cholesterol siRNA

Bae, Jeehyeon

INNOVATIVE DEVELOPMENT OF COVID-19 MRNA VACCINE USING THE GOLD NANOPARTICLE 
DELIVERY PLATFORM 
 
Jeehyeon Bae, Chung-Ang University, South Korea 
jeehyeon@cau.ac.kr 
 
 
At present, nearly all of mRNA vaccines approved or under clinical trials utilized lipid nanoparticles (LNPs) for 
mRNA packing and delivery, in which the entire full-length mRNA that encodes spike (S) protein of SARS-CoV-2 
is used. Here, we report an innovative SARS-CoV-2 mRNA vaccine candidate, AuX-RBD mRNA, that uses DNA 
oligonucleotides conjugated gold nanoparticles (AuX) as a mRNA delivery platform loaded with much shorter 
mRNA encoding only the receptor binding domain (RBD) of S protein as an antigen. Intramuscular vaccination 
of AuX-RBD-mRNA induced robust humoral immune responses and generated functional antibodies that 
neutralized and protected from SARS-CoV-2 infection in mice models. In addition, a non-clinical good laboratory 
practice (GLP) study performed in rats at a repeated high dose assured the safety of AuX-RBD-mRNA vaccine 
candidate. A large scale GMP production of AuX-RBD-mRNA vaccine for the clinical study has also successfully 
achieved, and its IND filing has been completed. Thus, our AuX-RBD-mRNA is a good candidate for COVID-19 
vaccine, and we propose that our AuX system can be extensively used as a platform for the development of 
RNA- and DNA-based therapeutics. 
 
 
 


Innovative development of COVID-19 mRNA vaccine using the gold nanoparticle delivery platform

Stanton, Matt

DNA LNP DELIVERY 
 
Matt Stanton, Generation Bio, USA 
 


German

DNA LNP delivery

Sahay, Guarav

LNP DELIVERY 
 
Guarav Sahay, Oregon State University, USA 
 


LNP delivery

Please click Additional Files below to see the full abstrac

Wagner, Ernst

CHEMICAL EVOLUTION OF ARTIFICIAL PEPTIDE CARRIERS FOR NUCLEIC ACID DELIVERY 
 
Ernst Wagner, Pharmaceutical Biotechnology, Ludwig-Maximilians-Universität Munich, Germans 
ernst.wagner@cup.uni-muenchen.de 
 
 
Key Words: Cas9, mRNA, pDNA, PMO, siRNA 
 
It took 50 years from first transfections to approval of gene therapies as medical drugs. 9 gene therapy products, 
2 siRNAs, and 8 oligonucleotides reached the market. Controlled, targeted intracellular delivery remains the key 
bottleneck for the success of such therapeutic macromolecules. Different chemical evolution approaches are 
pursued for further refinements of synthetic nanocarriers. Our strategy focuses on a bioinspired sequence-
defined process including (i) artificial amino acids active in specific delivery steps, (ii) precise assembly into 
sequences by solid phase-assisted synthesis, (iii) screening for a defined delivery task and selection of top 
candidates, followed by random/educated variation for a next selection round. The optimal sequence and 
delivery motifs of nanocarriers depends on the type of cargo [1] and application, as will be outlined for pDNA [2], 
siRNA [3-6], mRNA, PMOs [7], or Cas9/sgRNA [8]. 
 
 
 
 
 
References: 
[1] Peng L, Wagner E. (2019) Polymeric Carriers for Nucleic Acid Delivery: Current Designs and Future 
Directions. Biomacromol. 20, 3613. 
[2] Wang S et al (2017). Antitumoral Cascade-Targeting Ligand for IL-6 Receptor-Mediated Gene Delivery to 
Glioma. Mol. Ther. 25, 1556-1566. 
[3] Lee DJ et al (2016) Dual antitumoral potency of EG5 siRNA nanoplexes armed with cytotoxic bifunctional 
glutamyl-methotrexate targeting ligand. Biomaterials 77, 98-110. 
[4] Klein PM et al (2018) Folate receptor-directed orthogonal click-functionalization of siRNA lipopolyplexes for 
tumor cell killing in vivo. Biomaterials 178, 630-642. 
[5] Reinhard S et al (2018) Precise enzymatic cleavage sites for improved bioactivity of siRNA lipo-polyplexes. 
Bioconjug. Chem. 29, 3649-3657. 
[6] Luo J et al (2019) IL4-receptor-targeted dual antitumoral apoptotic peptide - siRNA conjugate lipoplexes. 
Adv. Funct. Mater. 29, 1900697. 
[7] Kuhn J et al (2019) Supramolecular assembly of aminoethylene‐lipopeptide PMO Conjugates into RNA 
Splice‐Switching Nanomicelles. Adv. Funct. Mater. 29, 1906432. 
[8] Kuhn J et al (2020). Delivery of Cas9/sgRNA ribonucleoprotein complexes via hydroxystearyl oligoamino 
amides. Bioconjug. Chem. 2020 Jan 22 
Figure 1 Tumor-targeted delivery. One of three siRNA formulations 
1214.0.1HA (b), versus 1214.0HA (a) or 1214.0.8HA (c), accumulates at 
tumor vessels at 5 min, extravasates within 45 min (d, e) and triggers 80% 
gene silencing (not shown). (f) Timing of 12140.1HA extravasation. 


Chemical evolution of artificial peptide carriers for nucleic acid delivery

Beskrovnaya, Oxana

TfR-Fab CONJUGATED OLIGOS FOR MUSCLE DELIVERY 
 
Oxana Beskrovnaya, Dyne Therapeutics, USA 
 


TfR-Fab conjugated oligos for muscle delivery

Not a member yet