Conventional vaccines enhance the body’s immune response against a foreign protein. In an autoimmune disease, where the body’s misdirected immune system attacks its own cells, could an “inverse vaccine” turn down, or better still turn off, a pathological immune response? This question intrigued Lawrence Steinman, MD, PhD, a professor of neurology, pediatrics, and genetics at Stanford University and chairman of Pasithea Therapeutics.

Pasithea, dedicated to developing drugs for neurological disorders, recently announced encouraging preclinical results that support the efficacy of a tolerizing, inverse DNA vaccine for multiple sclerosis (MS). Based on experiments conducted with Hooke Laboratories, Pasithea reported that intramuscular injections of the candidate vaccine (PAS002) delayed the onset of paralysis, and reduced severity of peak disease. Prophylactic administration also reduced the incidence and severity of relapse in the mouse model.

“The results of this study show that this technology has the potential to tolerize to GlialCAM, a myelin molecule that has molecular similarity to the Epstein Barr Virus that triggers MS,” said Steinman.

Lawrence Steinman, MD, PhD, whose earlier work led to the development of Tysabri, a drug approved to treat patients with MS and Crohn’s disease, is now developing an inverse DNA vaccine for MS.

Viral connections

The definitive cause for MS is unclear, however, many studies have found B lymphocytes infected with EBV (Epstein Barr Virus) in the brains of MS patients, indicating EBV could trigger MS. Yet nearly 95% of all adults carry EBV but do not develop MS.

“Although EBV is the trigger and is necessary, fortunately, it’s not sufficient to trigger MS in most of us,” explained Steinman.

The direct cause is difficult to demonstrate for a disease like MS, but an epidemiological study led by Albert Ascherio, PhD, professor of epidemiology and nutrition at the Harvard School of Public Health, demonstrated that the risk for MS increases 32-fold upon infection with EBV and not upon infection with other similar viruses. Of 10 million U.S. military veterans included in the study, 801 had MS and 800 of them were positive for EBV. Moreover, 35 individuals who did not have any evidence of EBV infection when they entered the military, showed the presence of EBV in blood before they were diagnosed with MS, further supporting EBV’s causative role in MS.

An editorial on Asherio’s study in Science that Steinman co-wrote mentions, “These findings provide compelling data that implicate EBV as the trigger for the development of MS.”

A few days after Ascherio’s article in Science, Steinman’s group published an article in Nature (“Clonally expanded B cells in multiple sclerosis bind EBV EBNA1 and GlialCAM“) that showed a human cell adhesion protein found in myelin called GlialCAM is attacked in MS. The study provided structural and functional evidence that a domain in GlialCAM that resembles an EBV transcription factor called EBNA1 (EBV nuclear antigen 1) plays a pivotal role in MS pathogenesis. The rationale underlying the development of Pasithea’s tolerizing vaccine for MS, PAS002, is based on these findings.

“The tolerizing approach advanced by Pasithea represents a new paradigm in the treatment of MS, which has application to other autoimmune, neuroinflammatory, and neurodegenerative diseases. Their advance was made possible by the discovery that EBNA1 is a mimic of GlialCAM and serves as an immune target in MS,” said Scott Zamvil, PhD, a professor of neurology and immunology at the University of California, San Francisco, and chair in MS research. “DNA vaccinations targeting GlialCAM are unique and provide a robust clinical benefit. I can see this approach could be beneficial to patients both within the early inflammatory and secondary neurodegenerative phases of MS.” (Zamvil was not involved in this study).

Tracking pathogenesis

A hallmark of MS is the production of antibodies by the clonal expansion of B cell-derived plasmablasts in the brain.

“There are only a few diseases where if you do a spinal tap and look in the cerebrospinal fluid, you find high levels of immunoglobulins that are clonal,” Steinman said. “That was the map we used to home in on clonal antibodies that are directed to the piece of EBV that mimics GlialCAM.”

These antibodies bind GlialCAM that is expressed on astrocytes and oligodendrocytes in the white matter, resulting in the destruction of myelin that envelops axons, and eventually the disintegration of the neuronal axons.

“The majority of MS patients have a clonal antibody in the spinal fluid that cross-reacts between GlialCAM and EBNA1, and 25 to 30% of patients have that marker in the blood,” said Steinman. “We don’t know if that’s a snapshot in time, or if longitudinally studies would reveal increases at some points. Those studies are ongoing.”

B cells and their plasmablast progeny express a cell surface molecule called integrin a4, that enables these cells to travel from the bone marrow into the systemic circulation and finally cross the blood-brain barrier into the brain.

“One of our remarkable discoveries is that the main homing molecule alpha4 integrin that allows plasmablasts that make the clonal antibody to enter into the cerebrospinal fluid (CSF), is lost once the plasmablasts are in the brain,” said Steinman. “This means the plasmablasts can enter [the brain], but they cannot escape!”

Cartoon representing the close interactions between EBNA1 and the autoreactive antibody. [adapted from Lanz et. al., Nature, 2022]

Steinman’s team collaborated with a team led by William Robinson, MD, PhD, a professor of immunology and rheumatology at Stanford University. “Robinson’s patented technology (US Patent 2013:WO2012148497 A2012148493) for deconvoluting heavy and light chain clonal antibody expansions was critical in identifying the monoclonal antibody that binds EBNA-1 and GlialCAM,” said Steinman.

The team identified the cross-reactive antibody through single-cell sequencing of plasmablasts in blood and CSF collected from MS patients, and protein microarray assays of CSF-derived antibodies against MS-associated viruses. Using sequence analysis, interferometry affinity measures, and X-ray crystallography, the team determined the crystal structure of the EBNA1–peptide epitope bound to a fragment of the autoreactive antibody. This enabled the team to track the development of the GlialCAM cross-reactive antibody.

Molecular mimicry and its boosters

A foreign antigen may resemble a protein produced in the body. This phenomenon called “molecular mimicry” provokes T cells to attack the body’s own cells that express the mimicking protein, particularly in people with certain HLA types—molecules that act as billboards for immune-provoking protein fragments. [Steinman]

When parts of a host’s own protein resemble parts of a viral protein, the “molecular mimicry” can induce immune cells into attacking cells bearing such molecular mimics. Ordinarily, the body’s T lymphocytes are adept at recognizing foreign proteins (antigens), parts of which are presented by cell surface HLA (human leucocyte antigen) that act as molecular billboards on the cellular landscape.

“People with a type of HLA molecule called HLADRB*1501 or HLRDR2 are more susceptible to the consequences of molecular mimicry,” said Steinman.

Post-translational modifications of the shared domain enhance the mimicry between GlialCAM and EBNA1. The molecular mimic includes serine residues which when phosphorylated causes the cross-reactive antibody to bind more intensely.

The EBV viral protein EBNA1 and the human protein GlialCAM share a similar piece that triggers the generation of cross-reactive antibodies where antibodies against EBNA1 bind to GlialCAM. [Steinman]

Steinman said, “One of the genetic risk factors for susceptibility to MS are kinases that can phosphorylate proteins. That may be one of the reasons that although most of us have antibody to EBV in our blood, we do not have antibody in our CSF that is cross-reactive between EBNA-1 and GlialCAM.” The research community is looking at modulating kinases that control disease susceptibility by determining whether the protein is phosphorylated.

Modeling the disease

Steinman’s team used the EAE (experimental autoimmune encephalomyelitis) mouse model in their preclinical studies. This model was generated by Thomas Rivers nearly 90 years ago to understand why the smallpox vaccine was causing disseminated encephalomyelitis—a rare autoimmune disease marked by inflammation in the brain and spinal cord—in a small percentage of individuals vaccinated for smallpox.

EAE is the most common experimental model for MS. Studies in the EAE model have led to the development of three blockbuster drugs for MS: Copaxone made by Teva, Tysabri (natalizumab), a monoclonal antibody against integrin-alpha4 from Steinman’s lab, and Mayzent (siponimod), an oral drug for progress MS from Novartis.

“If I scored three holes in one, I would be considered a pretty good golfer and not just lucky,” Steinman said, emphasizing the importance of the EAE model in MS. “No one’s going to invest in a drug for MS unless it works in EAE.”

Inspired by Koch’s postulates in establishing the cause of a disease, Steinman’s team injected the EAE model with EBNA1 and found this increased the severity of paralysis associated with MS.

“The EBNA1 was associated with noncoding CpG sequences of DNA that activate the innate immune system. The CpG given at the dose in Lanz’s paper worsened paralysis in EAE, compared to giving the same peptide mixed with CpG that was scrambled. This in some ways is a fulfillment of Koch’s postulates,” explained Steinman.

Treating MS without shutting down the immune system

The therapies that are currently approved for MS are drugs that seriously modulate the immune system. Steinman said, “Tysabri blocks the ability of the immune system to traffic into the brain. It’s very good in shutting down MS, but if your brain has to fight an infection and your immune system is blocked, you’re susceptible.” Antigen-specific tolerance is the holy grail of an immune therapy—it is the therapeutic gap that Pasithea seeks to fill with its tolerizing vaccines.

“If we know what one of the main triggers of the disease is, and we could shut down the response to it, we may have one of the long sought after goals that you would want to see for an immune therapy: blocking the cause of the disease and leaving the rest of your immune system free to fight infections,” said Steinman. “We don’t want to shut down the immune system, if we can avoid it.”

To achieve such antigen-specific tolerance, Steinman’s team obtained promising results with an inverse DNA vaccine for MS that encoded the full-length myelin basic protein (MBP), in an earlier study. The first-generation vaccine reduced levels of autoantibodies against myelin and showed improved MRI brain lesions in a Phase II trial on 267 MS patients. However, with stronger therapies like Tysabri available, investors were reluctant to invest in an inverse MBP DNA vaccine, and the trials were discontinued.

“Now with GlialCAM linked to the EBNA1, which is a definitive trigger of MS, investigators like myself, are ready to push ahead again with DNA vaccines to GlialCAM, and perhaps with RNA vaccines like the one developed by BioNTech,” said Steinman.

In addition, the earlier vaccine encoding MBP included immune-stimulatory CpG sequences in the plasmid vector that posed a hindrance. “Testing the newer versions of the DNA plasmid with no CpG and with the coding region targeting GlialCAM promises to be more efficacious than tolerizing to MBP,” Steinman added.

A problem with both RNA and DNA vaccines is that they can stimulate the immune system via toll-like receptors. Recent work from BioNTech shows changing uridine to pseudouridine prevents RNA from stimulating the immune system. Eliminating CpG sequences from the vector backbone accomplishes a similar effect for DNA vaccines.

“There are some very interesting comparisons in what we’re trying to do with DNA and what BioNTech is doing with RNA. So far with traditional immunizing vaccines, RNA beat DNA. We have Moderna, Pfizer, BioNtech producing RNA vaccines, but we don’t have too much traction with DNA vaccines,” said Steinman. “We’ll have to see if DNA vaccines are better at tolerizing than RNA vaccines.”

Pasithea’s inverse DNA vaccine

Pasithea’s tolerizing vaccine PAS002 aims to suppress encephalomyelitis and MS. Its CpG-less plasmid vector encodes the EBNA1 peptide that mimics GlialCAM. Preclinical results in the EAE model show that PAS002 significantly lowers GlialCAM antibodies.

“The data show a reduction in the intensity of the initial attack and subsequent relapse in the EAE model. Initial results are telling us PAS002 does have foundations for moving this ahead,” said Steinman. “We’re somewhat comfortable as are the regulatory agencies with the idea of a vaccine where the DNA is modified and we’re encoding an antigen. It’s very bold and we want to make sure it’s safe and efficacious.”

“The tolerizing vaccine developed by Steinman based on his GlialCAM discovery is a major advance in our attempts to develop an immunologically specific, nontoxic therapy for multiple sclerosis,” said Howard Weiner, PhD, a professor of neurology at Harvard Medical School. (Weiner was not involved in Steinman’s study). “The beauty of using an inverse vaccine encoding a protein already present, is that it takes advantage of the body’s own regulatory system to fight disease.”

The vaccine can be administered subcutaneously or intramuscularly and does need to be designed to cross the blood-brain barrier. Unlike biologics such as monoclonal antibodies, tolerized immune cells cross the blood-brain barrier efficiently. “We want to tolerize the immune system outside the brain. If they see inflammation, tolerized immune cells can get into the brain and broadcast tolerance rather than inflammation.”

Steinman does not expect PAS002 to be a one-shot wonder. “One thing we’re learning on the other side of the equation with nucleic acid vaccines is that the durability of the effects isn’t as long as we had hope for,” said Steinman. Preclinical titers indicate the vaccine would need to be administered periodically but the regimen remains to be worked out in future studies.

In a serendipitous turn of events, Steinman found PAS002 could potentially play a role in vaccine development efforts against monkeypox. Pox viruses, including monkeypox, contain the same GlialCAM sequence that is present in EBNA1. Although the first line of defense against an infection such as monkeypox is still a conventional vaccine that boosts immune responses against the pathogen, Pasithea’s tolerizing vaccine could reverse uncontrolled immune responses in individuals who develop encephalomyelitis following monkey pox infection or vaccination. “But this will need further studies,” said Steinman.

Bolstered by the promising preclinical results, Steinman is determined to see PAS002 advance to the clinic for MS. His team is also developing biomarker assays to study the effect on anti-GlialCAM antibodies that they plan to deploy in the upcoming clinical trials.