Viruses have been evolving for millions of years, improving their ability to transfer genetic material to the hosts they infect. When it comes to gene transfer, viruses are efficient and effective. So, it’s no wonder that viruses—or rather, viral vectors derived from viruses—have been harnessed to accomplish gene transfer in the name of science and medicine. Viral vectors now carry genetic material in diverse applications, helping to correct genetic defects, prevent infections, and cure cancer.
Viruses are diverse, and so are viral vectors. They come in different shapes and sizes, and they possess different capabilities and shortcomings.
Some integrate their genetic material into the host genome; others leave their genetic material outside the genome. Some engender strong immune responses; others are relatively innocuous. Some carry RNA; others carry DNA. Some can carry large payloads; others are less capacious. Some require “helpers” (which may be provided by the host or added to the mix); others can go it alone. Some are enveloped; others are nonenveloped. Some are suited to in vivo delivery; others do their best work in a Petri dish.
With such diversity, is there any way to guarantee that viral vectors are safe and effective? We put this question to experts of various kinds—specifically, experts in molecular biology, translational development, and bioprocess engineering. The experts responded that, yes, there are ways to prevent viral vectors from failing the patients they’re intended to help.
Is enough too much?
To date, most of the vectors used for gene therapy have been derived from the adeno-associated virus (AAV). AAV vectors are the go-to vectors because they have a record of preclinical and clinical success.
“There has been an increase in clinical evidence for the curative potential of AAV therapies,” says Christian Thirion, PhD, founder and CEO of Sirion Biotech. Perhaps the most compelling evidence concerns the two AAV-based gene therapies that have been approved by the FDA. Zolgensma has been approved for treatment of spinal muscular atrophy (SMA) type 1, and Luxterna has been approved for the treatment of retinitis pigmentosa linked to a PRE65 deficiency.
The gene therapy field, however, has been shaken by reports of three deaths in the ASPIRO trial. Conducted by Audentes Therapeutics, the trial had been evaluating a gene therapy for X-linked myotubular myopathy (MTM). “It was evident,” Thirion notes, “that the problems occurred in the cohort that received a very high vector dose—in this case, 3 × 1014 genome copies per kilogram.”
According to Thirion, similar toxicity was observed in nonhuman primate trials, as was reported by James M. Wilson, MD, PhD, the director of the Gene Therapy Program at the University of Pennsylvania. “The field is trying to understand which critical parameter [accounts for the] severe adverse events in those high-dose cohorts,” Thirion adds. Most indications point to an out-of-control inflammatory immune response, popularly called a “cytokine storm.”
Regardless of the cause, there seems to be a consensus that there are too many particles being introduced in AAV-based gene therapies. “We dose at 100 times the number of cells in the human body,” says Magnus Gustafsson, PhD, head of global business development at Biovian, a contract development and manufacturing organization (CDMO).
Efforts are being made to improve the impact of each particle so that fewer are required. Sirion, for example, is working on a solution in collaboration with a research group led by Dirk Grimm, PhD, at Heidelberg University Hospital. The collaborators use a library of more than 50 million barcoded vectors to infect nonhuman primates. “Basically, [this approach accelerates] nature’s evolution,” Thirion explains. “We are able to identify specific new AAV capsid variants which are far more efficient in delivering the genes in vivo. The potential is that we can lower the vector dose by an order of magnitude per patient.”
“You can achieve a similar effect by improving the promoters driving the therapeutic transgene,” he continues. This approach improves gene expression. He adds that both approaches—barcoding and promoter optimization—can be combined to advantage.
Make it right
The safety of any biological depends at least in part on the diligence with which the biological is manufactured. Viral vectors are no exception.
“We produce them in clean rooms; we control the environment; we control all the raw materials to produce these types of therapies,” says Mayo Pujols, CEO of the CDMO Andelyn Biosciences. “[Our] ultimate goal [is] to ensure that they’re safe from cross-contamination or any type of foreign or bacterial contamination. That’s first and foremost.
“We run multiple in-process tests and conduct final product testing to ensure that these materials are all at safe levels—not just host cell proteins and DNA, but other production products as well.”
AAV vectors are typically created by triple transfection of plasmids into the host cell because it’s quick, robust, cost effective, and time-line effective. “[But] you can end up with 90% of the AAV being empty—10% being active,” warns Gustafsson. “Most of it is junk that needs to be taken care of by an efficient manufacturing system … but then you need to purify away the empty ones—that’s a large part of the immunogenic response that you get.
“It’s important to use someone who knows what they’re doing. It’s not like the monoclonal antibody field where you use Protein A to fish out the monoclonal antibodies and get rid of the rest, and you get a 95% clean product to start with.”
There are, however, differences between empty and full capsids that can be exploited by, for example, affinity chromatography, ultracentrifugation, and anion-exchange. Companies are also working on using smaller plasmids that transfect better, Gustafsson points out. Companies are also exploring alternatives to triple transfection such as baculovirus-based expression and transduction-based methods.
Viral vectors in vaccines
In the gene therapy field, much effort has been made to discover or create vectors for which there is no preexisting immunity, and which do not engender an inflammatory response. Effort has also been made to deal with instances in which preexisting immunity cannot be avoided. For example, there have been studies looking into clearing existing antibodies prior to administering the AAV vectors. “You could even re-dose the vectors,” Thirion suggests.
The situation is nearly reversed with vaccines, where driving an immune response is “what it’s all about,” remarks Chris Mason, MD, PhD, professor of cell and gene therapy at University College London and CSO of Avrobio. For example, vaccine developers may opt to use adenovirus.
Adenovirus has fallen out of favor as a gene therapy vector because of its immunogenicity. “[However, adenovirus] is being used to transport the COVID-19 vaccines,” Mason points out, “and it’s doing that very successfully.”
Of course, it’s important to avoid having a full infectious cycle with a full disease. This can mean using an inactivated virus to generate immunity against the active form of the virus, the prototype being measles.
It could mean using an attenuated virus into which is inserted a gene of interest (GOI) and which will induce an attenuated infection, and in turn multiply and present the encoded protein to the immune system. The Institut Pasteur attempted this for COVID-19, for example, “but it didn’t work so well,” admits Christian Bréchot, MD, PhD, professor of infectious disease at the University of South Florida and president of the Global Virus Network.
Another option is to use a vector which will not be able to replicate fully in the body when injected, but will be able to transfer the GOI to host cells (as was the case with the AstraZeneca and Johnson & Johnson COVID-19 vaccines).
Vaccines can also be directed against cancer or chronic infection. In these cases, a vaccine is considered to be a therapeutic rather than a prophylactic. Bréchot points to three types of safety concerns with therapeutic vaccination.
The first concern, as with any vaccination, is whether the vector itself will generate too much of an inflammatory response.
The second concern is whether the vaccine will change the natural course of the chronic infection, such as hepatitis, in a way that might be deleterious. “That is something you test on animal models and preclinical models before moving to the clinic,” Bréchot explains. “So far, this type of safety concern has really not been a major issue.”
The third concern is whether there will be long-term consequences due to viral vector administration. Such consequences, Bréchot notes, are “due to the vector’s interaction with the host cells, in particular, integrations.”
Retroviral vectors integrate their GOI into the host genome, and thus pass it along whenever the cell divides. “That’s good if you want long-term therapy,” says Mason.
But inserting the wrong sequence into the wrong place can wreak havoc. Several years ago, patients treated for severe combined immunodeficiency developed leukemia as a result of insertional oncogenesis.
Bréchot explains that the field responded, in part, by changing the genetic organization of the vector, the promoter, etc., so that “even if there were to be an insertion in some sensitive site, that could not have the same consequence.”
Lentiviral vectors derived from the HIV retrovirus have become useful in the development of chimeric antigen receptor T-cell therapies and other ex vivo cell-based gene therapies. Here, cells are removed, transduced, and returned to the patient. One advantage, according to Mason, is that “you can target very specific cell populations, because you can do it on a selection basis,” and use many orders of magnitude less vector. He says that another advantage is that “you can freeze the product down, which gives you time to actually do a number of tests to make sure it meets the safety and potency criteria that you require, prior to infusing it back into the patient.”
Researchers are also looking into ways to mediate site-specific integration to further assure against an insertional oncogenesis event.
Non-integrating versions of the lentiviral vector have been developed in a collaboration between the Pasteur Institut and the vaccine company Theravectys. These vectors are designed to target the GOI to dendritic cells in vivo, stimulating “a well-controlled T-cell response” against cancers and chronic viral and bacterial infections, says Bréchot, who is chair of the company’s scientific advisory board. The effective dose of the lentiviral vector-based vaccine—which can be used either therapeutically or prophylactically—is generally 1,000 to 10,000 times lower than that necessary with adenoviral vectors, reducing the potential for an adverse reaction to the vector.