Fred Hutch

In September 2025, I joined the Hans-Peter Kiem Lab at the Fred Hutchinson Cancer Center. One of the research aims at the Kiem Lab is to make in vivo gene therapy more effective for certain diseases. Currently, ex vivo gene therapy yields better patient outcome. Although, there are numerous problems that arise with ex vivo when brought to scale up and cGMP manufacturing. These issues significantly increase production costs, making this therapy an unaffordable option for many patients. On the contrary, in vivo gene therapy is cheaper to manufacture but does not provide the same results. If we can find ways to increase the effectiveness of in vivo gene therapy, we can make cell and gene therapy a more accessible option for all patients.

Lab Procedures:

DNA Extraction
Polymerase Chain Reaction (PCR)
PCR Cleanup
Gel Electrophoresis
Bacterial Transformation
Cloning
Flow Cytometry
Cell passaging, harvesting, and counting
dPCR and titer calculation

Research Projects

Polymeric Nanoparticles used for Gene Delivery

Polymer nanoparticles are non-viral delivery systems. Calling them a nanoparticle is a bit deceiving as their appearance is not exactly spherical. They consist of positively charged polymers that encapsulate a negatively charged nucleic acid cargo. These polymers form clumps or clusters around the cargo, making their morphology non-homogenous. The cluster is able to travel within the body, protecting the cargo from enzymatic degradation and renal clearance, while also delivering it to the target organ or tissue.

In this experiment, we tested the transfection ability of PNPs on human and monkey cell lines as well as in a murine model. The nanoparticles carried GFP encoding RNA. Additionally, they were targeted with CD90 antibodies. The PNPs were plated on the cell lines at varying RNA dosages. Both human and monkey cell line transfection data looked promising, with maximum transfection resulting from the medium dose. The nanoparticles were then administered in vivo on live mice. They were engrafted into the bone marrow and allowed to sit for a few days. Flow cytometry data revealed minimal to zero transfection across every PNP used.

The hypothesis was that the PNPs had little resistance to the enzymes found in the serum of mice blood. The serum likely inhibited the PNPs ability to target and transfect CD90 cells in the mice.

Conditioning and Enrichment of CD45 cells via Astatine-211 monoclonal antibodies

Astatine (As) is a highly radioactive element that has the potential to kill blood cells at very low dosages . The purpose of this project was to find the dosage of astatine that yields the highest edited blood cell abundance post radiation, but with the least harmful side effects. The editing was carried out ex vivo. CD45+ blood cells are taken out of mice and CRISPR was used to achieve the desired edits. These edits consisted of a therapeutic edit as well as a CD45 edit. The therapeutic edit altered the sickle cell gene to express fetal hemoglobin (HbF). Fetal hemoglobin is a protein found in fetal red blood cells. They are known to never sickle. The CD45 edit was a CD45 knockout (KO). This essentially kicks off proteins found on this cell's epitope, nullifying the cell surface receptor. Afterwards, astatine monoclonal antibodies (mAb) are administered to the mice. The mAb targets CD45 cells. The point of this was to clear out CD45 cells prior to introducing the edited stem cells. This helps create space for the transfusion and is often referred to as "conditioning". After the edited stem cells are back in the mice, the mAb is introduced again. This will internalize and kill remaining CD45 cells, although, the cells with our therapeutic edit will remain unharmed due to the CD45 KO edit. This amplifies the percent of edited cells in the mice. This process is referred to as "enrichment". After the antibody runs its course, genetic sequencing is done to determine the abundance of edited blood cells remaining in the mice. The therapeutic threshold for sickle cell disease is around 20%. This was our target.

The experiment was run on three different mice. Each received a different dosage of astatine. We had low, medium, and high dosages. The low dosage did not physically harm the mice, although the edited cell yield was very low. The high dosage killed the mice. The medium dosage achieved our therapeutic threshold while leaving minimal harmful side effects to the mice.

Engineered Virus-like Particles (eVLPs)

I like to think of eVLPs as a "build your own" virus. Essentially, specific plasmids are handpicked and used to create these virus-like particles that infect (or transduce) cells just like a normal virus would. In our case, the only difference is that the payload in an eVLP is a physical base editor while a virus would have a genetic payload that produces the base editor in the cell. In other words, the eVLP delivers a base editor while a virus delivers genetic material that expresses for a base editor. Despite this, the end results are often the same; a gene edit is performed either way.

In this project, various eVLPs were made with differing plasmids. This was to test the effect that certain plasmids would have on the eVLPs ability to transduce cells. There were two types of plasmids that were altered among the eVLPs. The first was the plasmid that expressed for the virus' envelope. Both the COCAL and VSVG envelopes were used. Within each, we had a wild-type (WT) and a knockout (KO). The WT is the unaltered version of the envelope, and in theory it would transduce the best. The KO was the altered version that lost its ability to transduce. Secondly, plasmids that expressed for the base editor were changed. Cas9 is an enzyme found in base editors. Its role is to recognize the PAM site, the nucleotides immediately before the DNA strand of interest. The first plasmid used was NGG. This essentially programs the Cas9 to recognize a three-nucleotide PAM site, consisting of any nucleotide followed by two Gs. The second plasmid was NG. This tells the Cas9 to recognize a two nucleotide PAM site, consisting of any nucleotide followed by one G. Additionally, every eVLP received plasmids that expressed for helper proteins which are critical for virus formation and structure. These plasmids stayed consistent across all eVLPs used.

A total of 8 eVLPs were constructed. Each had different combinations of the envelope and base editor plasmids. The data strongly suggested that a combination of COCAL WT with NGG produced eVLPs with the highest cell transduction rate.

Vector Production

I was given the unique opportunity to intern with the Vector Core team for a month. Vector Core is responsible for designing, constructing, and producing personalized viral vectors for clients. They have the capabilties to produce a wide range of viral vectors. These include lentivirus, adeno-associated virus (AAV), gammaretrovirus, and foamyvirus. Their clients are other research labs within Fred Hutch or small biotech startups in the Greater Seattle Area. Often times, these viral vectors get used in animal experiments. Thus, production of viral vectors are non-cGMP grade.

I was able to help with lentivirus production in both suspension and adherent cell cultures. The workflow consisted of growing and seeding cells, transfections, multiple rounds of harvesting, and resuspension to produce a viable final product. Afterwards, transductions and titrations (functional and infectious) are performed to determine a titer for the produced batch of viral vectors.

My main responsibilities included transfecting plasmids into cells, harvesting viruses from cell culture, and QC assays such as dPCR to determine the infectious titer of our lentiviral vector.

Thank you Megha, Zach, and Logan for opening your doors to me and giving such a cool experience in vector production!

Page updated

Google Sites

Report abuse