image: SOSIP trimer team members celebrate Rogier Sanders’ professorship appointment in Amsterdam, 2017. From Left: Ian Wilson, John Moore, Rogier Sanders, Andrew Ward. Courtesy of Dr. John Moore
Credit: Dr. John Moore
When SARS-CoV-2, the coronavirus that causes COVID-19, began spreading worldwide in 2020, many research teams immediately set to work developing a vaccine against it. Building on decades of previous work on mRNA technology and on other viral vaccines, including HIV, they achieved their goal within the year. The most widely used mRNA vaccine design contains the genetic instructions for the body to make the spike protein that the virus uses to enter cells. The resulting immune response protects against infection and, more importantly, disease and death. However, developing a vaccine for HIV has proven much more difficult.
“The COVID-19 vaccines were an enormous achievement but the spike protein on SARS-CoV-2 was like low-hanging fruit for vaccinologists,” said Dr. John Moore, professor of microbiology and immunology at Weill Cornell Medicine and part of an international team that has brought biomedicine closer than ever to an HIV vaccine. “It behaves like its counterparts on viruses for which vaccines are relatively easy to develop, such as influenza. Unfortunately, we learned back in the 1990s how hard it is to make an HIV vaccine.”
Building a stable env protein
The goal of immunization with a viral protein, or some portion of it, is to limit infection by teaching the body to generate neutralizing antibodies that bind to these viral proteins and block their interaction with the receptors found on the cell’s surface. These antibodies can also flag virus-infected cells for destruction by other immune system components.
For SARS-CoV-2, this viral target is called the spike protein; its counterpart on HIV is the envelope (Env) protein trimer. But HIV researchers attempting to target Env in the 1990s discovered that when the three-subunit Env protein is produced in the laboratory it promptly falls apart. To create vaccine candidates for HIV, and later SARS-CoV-2 and respiratory syncytial virus (RSV), it was critical to engineer this kind of multi-subunit vaccine to be more stable.
In 1998, with funding from the National Institutes of Health, Dr. Moore launched an HIV vaccine project to tackle this problem. The challenge was engineering an Env protein trimer that was hardier but still resembled the original closely enough to elicit appropriate antibody responses in test animals, and then people. Dr. Moore was soon joined by Rogier Sanders, a graduate student who came from Amsterdam to work on the project as part of his dissertation. The first advance, published in 2000, involved engineering a new chemical bond that helped key trimer components to stick together without distorting their overall structure. The second key development, in 2002, was swapping one amino acid for another in one of the trimer subunits to fix another major source of instability.
Over the next decade, Dr. Sanders, working with Dr. Moore after he returned to Amsterdam, made several more modifications to the Env protein that enabled them to eventually build a truly stable trimer. They named it SOSIP.664, a term reflecting the nature of the successful modifications.
A collaboration with structural biologists Dr. Ian Wilson and Dr. Andrew Ward at Scripps Research in La Jolla provided critical insights by showing what the new trimer designs looked like when viewed by electron microscopy. The project also involved what Dr. Moore refers to as “sheer grunt work”. To find the best mimic of the Env protein as it appears on the surface of HIV, the team obtained genetic information for about 100 different HIV strains from around the world and then synthesized SOSIP.664 trimers from all of them. A battery of laboratory tests and, above all, structural analyses by the Scripps team enabled the researchers to find the genetic sequences that produced the best Env trimer.
This optimal sequence, designated BG505, was isolated from an infant born with HIV in Kenya by Dr. Julie Overbaugh of the Fred Hutch Cancer Center and her colleagues at the University of Nairobi. To help further HIV research, they had shared the information with the International AIDS Vaccine Initiative (IAVI), a co-funder of Dr. Moore’s team at that time.
A final breakthrough occurred when electron microscopy images showed how the assembled trimers were attracting fat molecules, causing them to aggregate into useless clumps. Once the researchers removed that part of the protein, they had the stable, engineered Env protein they wanted. They named it BG505 SOSIP.664.
Eliciting broadly neutralizing antibodies
Another major challenge in developing an HIV vaccine is that the virus mutates rapidly to evade detection by the immune system. Thus, people living with HIV around the world carry different versions of the Env protein. “It’s akin to what we saw with the COVID-19 variants, but much, much worse,” Dr. Moore said. An effective HIV vaccine must coax the immune system to make "broadly neutralizing antibodies" (bNAbs) capable of attacking many forms of the virus. “We know these antibodies exist, because some infected people make them, and we could show they bound to our SOSIP trimers,” added Dr. Moore. He and his colleague, Dr. P.J. Klasse, professor of research in microbiology and immunology at Weill Cornell Medicine, have been studying HIV neutralizing antibodies for over 25 years.
But could BG505 SOSIP.664 and other trimers the team soon made stimulate the production of bNAbs? Early tests in animal models showed that the BG505 trimers elicited antibodies specific for the infant’s strain, but not the bNAbs that neutralize a broad sample of viruses. The quest continued, now guided by ever-increasing knowledge of the underlying immunology.
Now, leading investigators are pursuing a multi-step immunization process known as “germline-targeting” to generate a lasting HIV vaccine response. This strategy involves activating the antibody-producing cells that make precursors of the broad neutralizers, then coaxing those antibodies along a path to full activity. A germline targeting SOSIP trimer, re-designed by the Sanders’ team and designated GT1.1, is in human trials supported by the Gates Foundation. A recent paper reported success in generating the desired bNAb precursors in a group of healthy volunteers. In an accompanying editorial, Weill Cornell professors Drs. Sallie Permar and Patrick Wilson outline why this approach to an HIV vaccine is so promising. Follow-up clinical trials in Africa are in progress or being planned also. The Moore/Sanders team is continuing its multi-year collaboration with the Permar group to further evaluate the GT1.1 trimer at the pre-clinical stage, as the accrued information can inform clinical trial design.
Progress in jeopardy
Projected decreases in NIH support for vaccine research and development, and other reductions in federal spending, could jeopardize these promising advances. Private philanthropy, including from the Gates Foundation, is vital, but can’t fully compensate for federal funding.
“The NIH has funded the basic design and development work for SOSIP trimer vaccines for over 20 years,” said Dr. Moore. “These were competitive grants. Everything is at risk.” But whatever the future holds, he notes how the COVID vaccines used the same principle of engineering stability into the spike protein. “So indirectly, our work on HIV helped make the COVID mRNA vaccines work as well as they did.”