February 13, 2018

Techniques for Improving Process Development for Exosome-Derived Therapeutics

An interview with Aaron Noyes, Head of Downstream Processing at Codiak BioSciences

Aaron Noyes, Zodiak BioSciencesSome therapeutic targets have long frustrated drug developers hoping to crack the code of mitigating untreatable diseases like pancreatic cancer. One promising biotherapeutics category that may hold the key for achieving this goal is exosomes. These nano-scale membrane sacs are mediators of intercellular communication and can be exploited as carriers of therapeutic molecules (like proteins and RNA) that can alter disease states. 

One developer currently working in this area is Codiak BioSciences of Cambridge, MA. Founded in 2015, this company’s primary focus is to develop therapeutics with exosomal cargo and engineered surface ligands that can affect diseases, especially those with “undruggable” targets. Specifically, Codiak is investigating the delivery of nucleic acids and engineering exosomes that, when combined with a platform production process, can become exosome drug therapeutics that are scalable, consistent, and have a competitive cost of goods to be a viable contender in medicine.

Here, the Process Development Forum speaks with Aaron Noyes, Head of Downstream Processing, at Codiak BioSciences about exosomes, their potential clinical applications, and process innovations that would benefit developers of these types of therapeutics.

Would you please start by giving us a short overview of what are exosomes and the benefits they may have in clinical applications?
Exosomes are cell-derived vesicles produced by living cells and found throughout the body. These native information carriers between cells and tissues can be used by drug developers to deliver payloads to recipient cells and alter the biology of recipient cells as part of the communication. Because these structures are native to living organisms, they are well tolerated by the immune system, which is very important from a therapeutic standpoint. This tolerance can lead to greater potency and enhance unique tropisms, such as migration to tissues including the lymph nodes and pancreas that previously have been difficult to reach with traditional treatments. 

From a biochemical standpoint, exosomes are composed of a phospholipid bilayer associated with transmembrane proteins, peripheral proteins, DNA, RNA, polysaccharides, and cholesterol. The lumen of the vesicle can contain RNA and other species. It's a very complex entity. The potential therapeutic value of exosomes is that developers can alter the properties of the surface and the luminal cargo to make a meaningful difference in disease states. 

For instance, exosomes are useful for delivering nucleic acids to recipient cells and knocking down mRNA expression. Because of the complex yet modifiable surface topology of exosomes, developers can tie delivery to combination therapy, creating a more sophisticated drug product to target very difficult diseases.

What sort of disease states might exosomes help address?
Cancer is certainly one hot area. With pancreatic cancer, for instance, the standard of care does not offer much advantage to the patient. Breakthroughs in this area are desperately needed.

One cancer treatment with enormous potential is for exosomes to deliver siRNA and knock down the expression of oncogenic mRNA. The laboratory of Dr. Raghu Kalluri, one of our founders at The University of Texas, MD Anderson Cancer Center focused on the knockdown of KRAS, a gene encoding a GTPase implicated in several types of cancer, with exosomes containing siRNA. He was able to demonstrate effective knockdown and survival of small animals with pancreatic cancer. This work was published in Nature in 2017.

What are the main process development and manufacturing challenges for this class of compounds to be successful?
The first challenge relates to the heterogeneity of exosomes. Because exosomes are so complex, there is a lot of variety within a population of exosomes. Such variety includes both the surface chemistry and the size. Although similar in size to enveloped viruses, exosomes do not have a monodisperse distribution. They can range in size from 50 nanometers to 150 nanometers in diameter. This size range coupled with the diversity of surface chemistries creates a lot of heterogeneity to handle from a production standpoint. It can be difficult to isolate a large swath of the exosome population while achieving sufficient purity.

In the literature, the most commonly used approach for isolating exosomes is ultra-centrifugation. Many academic groups have found success with this technique, which often relies on density gradients of sucrose or iodixanol. Ultracentrifugation is, however, quite difficult to operate and scale-up. Other isolation approaches such as chromatography haven’t been as successful historically because of the surface heterogeneity of exosomes.
Related to the heterogeneity issue is, how do you characterize the heterogeneity? You can characterize the interior or exterior of exosomes. You can isolate different populations and sub-populations. But, you must have the right analytical tools to understand what you're making and whether that’s meaningful from a clinical standpoint or not.

How could one overcome these challenges?
The first key is to develop a robust approach for isolating exosomes. Then, a single batch of exosomes could be loaded with different payloads to generate multiple distinct drug products. Luminal payloads can be introduced through endogenous loading by recombinant cell lines or through ex vivo loading. One can move to the next level of sophistication by changing the surface of the exosomes, in which case one might convert that first process into a platform that can be tweaked to effect adequate purification for different types, different flavors, and even different sub-populations of exosomes.

Codiak BioSciences has invested a lot of effort to develop advanced technology in a production process that is scalable and that can be used by our partners to produce exosomes. We have done this for the first project in our pipeline, scaled it up, and have seen good scalability of performance.

It is important to develop the initial process with a future platform in mind. All templated processes start with a single purification process, but there are things you can do to increase the likelihood that it will translate well into a platform such as designing around the modality structure, using well understood technology, and establishing robustness with many development batches. We’re at the stage now where we have taken the initial process with native exosomes and successfully employed it to produce engineered exosomes with custom surface properties.

With any approach, you must have the right analytics to tell you that you’re producing a drug substance with the appropriate characteristics. So, we invest quite a bit of time in understanding the biochemical properties with a focus on what matters from a functional standpoint. A myriad of attributes can be studied for a given population of exosomes. We want to focus on the ones that relate to the functional characteristics of exosomes and how they might behave from a safety and efficacy standpoint in the clinic. For exosomes, what you look at is as important as what you don’t look at.

What process innovations could most benefit exosomes?
Having technologies to interrogate individual exosomes would be really useful. This would be analogous to a flow cytometer being used to determine how many cells or what percentage of cells in a CAR T-cell population have a certain characteristic. In lieu of this option, you can rely on population-based statistics similar to what is employed with viral vectors but with this approach, you lose resolution and information. If we could look at each individual exosome and understand more about its characteristics, then we could better understand how exosome structure relates to the overall therapeutic behavior of the pool.

This key characterization technology is lacking. I'm optimistic it will be developed in the next few years because there's such a pressing need and flow cytometry is getting closer from an optical standpoint. Now, we just need to increase the resolution of the available technology with an eye toward how you can make it work broadly for viral vectors and exosomes.

Read More Q&As With Industry Leaders Here 

Tags: novel biotherapeutics, exosomes, pancreatic cancer, cell-derived vesicles, process development, manufacturing