By Deborah Borfitz 

March 12, 2024 | Researchers at University of California San Diego School of Medicine have found that the rules of vaccinology need to be reversed for Staphylococcus aureus (SA), the common and sometimes dangerous bacteria that humans encounter shortly after birth and are continuously exposed to for the remainder of their life. Targeting the “subdominant” antigens that initially trigger a weak immune response could yield more protective antibodies than the dominant ones targeted by SA vaccines that have consistently failed in clinical trials, according to George Liu M.D., Ph.D., professor in the department of pediatrics. 

This is the key finding of a newly published study in Cell Reports Medicine (DOI: 10.1016/j.xcrm.2023.101360), which also corroborated an earlier discovery that prior exposure to SA leaves a nonprotective imprint on mice. On the few occasions when a vaccine did work, says Liu, it was only because the mice never made an immune response to the bug in the first place and therefore had no bad response to recall.  

At last there is a plausible explanation for why SA vaccines that have performed so well in lab animals, who are not normally exposed to staph, don’t translate to people, he says. “The predictive value of this naïve mouse to human is zero.” 

The immune imprinting hypothesis, which Liu and his team first proposed in a 2022 article in Cell Host & Microbe (DOI: 10.1016/j.chom.2022.06.006), has a precedent in the “original antigenic sin” idea forwarded more than six decades ago about waning response to influenza vaccination following initial exposure to the flu virus. Although SA does not evolve like influenza viruses, it is somehow connected to past immune reactivity, he says. 

The hypothesis was first explored in Merck’s phase 3 trial of a SA vaccine targeting SA surface protein iron surface determinant B (IsdB), where participants receiving it not only did no better than those who did not, but they were more likely to die from SA if they later acquired a staph infection, continues Liu. Researchers subsequently learned that mice previously infected with SA, unlike SA-naïve animals, did mount an antibody response to vaccination but it was not protective—a good reflection of what was happening clinically. 

As it turns out, SA infection tends to induce tons of antibodies that confer no benefit, he says. “They’re just hanging around, so that when you get a vaccine... [it] preferentially recalls that non-protective antibody response.” 

With the latest study, that finding was replicated in 10 other SA vaccines that have failed in clinical trials and it held true for both active (vaccination) and passive (premade monoclonal antibody) immunization. This suggests that SA vaccines might be more practically and predictably developed based on pre-existing humoral imprint characteristics, says Liu. 

Immune Imprinting

The first staph vaccine study was published as early as 1903, says Liu. Methicillin-resistant SA (MRSA) was first observed among clinical isolates from patients hospitalized in the 1960s and since the 1990s has spread rapidly to make development of a vaccine a public health priority.   

About 30 SA vaccine clinical trials (many phase 1 studies) have been conducted over the last three decades, and there has not been a single successful vaccine, Liu continues. Among the most prominent vaccine candidates to make the translation to human studies, besides IsdB, were those targeting MntC (manganese transport protein C), ClfA (clumping factor A), Hla (alpha-hemolysin), and (LukE) leukotoxin E. 

The situation had become so worrisome and confounding that meetings were convened starting about 10 years ago to try to get at why the mouse model was not accommodating SA vaccine development efforts, as it has for so many other vaccines including those for influenza and most recently COVID-19. This was when Liu and his colleagues started thinking about how lab mice live in cages with little opportunity to encounter staph and decided to replicate the Merck clinical trial in both mice that had and had not been exposed to the bacteria.  

When expanded to multiple SA vaccine clinical trials, the surprise was that the vaccinology rule reversal principle didn’t apply to three of the vaccines. When investigators took a closer look, they found the reason to be that the mice didn’t have a signature of past reaction to the infection. When it came to the passive immunization platforms (suvratoxumab for neutralizing SA alpha toxin and tefibazumab directed against ClfA), nonprotective antibodies actively blocked the effectiveness of the monoclonal antibody, consistent with the results of their respective clinical trials. 

At least in mice, the study demonstrated that the immune imprinting principle could be used to confidently predict the outcome of a vaccine trial based on prior exposure to the antigen being targeted, says Liu. “While we cannot say this is the explanation as to why a vaccine works, the mechanism can clearly show how it works.”  

Sponsor companies have understandably grown more pessimistic about the odds of coming up with a marketable SA vaccine and have at times faulted poor preclinical study design. “Our [imprint] hypothesis has been pretty positively taken by the field so far,” notably biotechs, now it has been validated in a series of vaccines against different cell-wall-associated antigens, says Liu. 

Antibodies that attack the cell walls of SA bacteria are responsible for the phenomenon, he adds. Those that instead target the toxins produced by SA are the ones able to successfully neutralize them.  

‘Accidental’ Resistance

It is reasonable to ask if SA might bestow any benefits on the human immune system, given it is a nearly ubiquitous bacterium that lives quietly and unnoticed as a member of the skin, nasal, and gut microbiome. “I see it as more of a pathogen than a commensal bacterium,” says Liu, since it seems to do more harm than good. 

Unlike the SARS-CoV-2 virus, SA also doesn’t readily work with the body's natural defenses to enable development of an immunity-conferring vaccine. The first COVID vaccine tried was 90% successful, he points out, which is relatively unusual. 

The difficulty level of creating a vaccine for SA is on par with other multidrug-resistant bacteria including those causing malaria and tuberculosis, Liu says. “Many of those have lived with us for a long time, meaning they learned how to survive in our body and have probably found a strategy to neutralize our immune system by creating antibodies that don’t work.” The production of protective antibodies is a natural response to infection, “but bacteria have outsmarted us with this evasion mechanism ... [and] the casualty here is the vaccine itself.”  

It’s an “accidental” sort of antimicrobial resistance, says Liu. As has been postulated, people may become resistant to certain vaccines (e.g., pertussis) over time as the causal bacteria evolves to escape vaccination. 

Just as an antibiotic keeps bacteria in check, bacteria can resist a vaccine, albeit preemptively, he explains. The commensal microbiome produces quite a few of these immune evasion mechanisms, and one of them may be to disable the production of effective, SA-fighting antibody T cells. 

The imprinting principle could explain the failure of most if not all SA vaccines that have entered clinical trials to date, says Liu. That makes it a good basis for rethinking vaccine development strategies to outwit the “part-time pathogens” living within us.  

From a public health policy perspective, demand for a SA vaccine is unquestionably high to prevent infection-related hospitalizations and deaths in high-risk individuals, notably the elderly. Staph infections are “one of the most common invasive bacterial diseases with great mortality” and notorious for their ability to become resistant to antibiotics, “generating even more need,” Liu says.