For 25 years, researchers have engineered immune cells to hunt down HIV. The approach, known as chimeric antigen receptor (CAR) T-cell therapy, extracts a patient’s T cells, engineers them to recognize a specific target on infected cells, and infuses them back into the patient. The engineered cells circulate, find their target, and kill. In fact, the challenge has never actually been about whether a target could be killed; it lies in getting the newly engineered cells to last long enough to finish the job.

The first generation of anti-HIV CAR T cells, developed in the late 1990s, showed only modest efficacy against the virus in early clinical trials. Later designs improved: multispecific targeting domains, costimulatory signaling architectures, and constructs engineered to resist viral escape.

To date, none of the tested anti-HIV CAR T cells have prevented viral rebound after antiretroviral therapy (ART) is discontinued. A 2021 phase 1 trial of CAR T cells targeting HIV after interruption of ART saw all six participants rebound at a median of just over 5 weeks.

Cell Manufacturing Approach

A growing number of research groups have begun to question whether the problem was never the weapon but the forge. Now, a study published in Science Advances, led by Erin B. Cole, PhD, in Harris Goldstein’s laboratory at Albert Einstein College of Medicine in the Bronx, New York, takes the idea further than anyone else.

Instead of redesigning the CAR construct, the group changed how the cells are manufactured, replacing standard T-cell activation with HCW9206, a protein scaffold developed by Hing C. Wong, PhD, at HCW Biologics in Miramar, Florida. HCW9206 links three cytokines: interleukin (IL) 7, an IL-15 superagonist, and IL-21.

The resulting CAR T cells were enriched more than 14-fold for T-memory stem cells (TMSCs), defined by expression of the surface markers CD45RO-, CCR7+, and CD95+. TMSCs are the longest-lived, most self-renewing subset in the human immune system. The 14-fold expansion is compared to less than 5% with conventional production.

Goldstein, senior author of the study and director of the Einstein-Rockefeller-CUNY-Mount Sinai Center for AIDS Research, has an ambition that goes beyond incremental improvement. His model for a functional HIV cure is the way the immune system already handles viruses such as cytomegalovirus and varicella zoster, which persist latently in the body for decades but are held in check by circulating T cells, without requiring any drugs.

“We’re not taking drugs for CMV [cytomegalovirus], we’re not taking drugs for varicella zoster, but our immune system is suppressing it,” Goldstein said. “It’s because they have T- stem cell precursors that can continuously replenish and provide new T cells.”

Lighter Fluid vs Charcoals

To build a CAR T cell, researchers must first insert the CAR gene into the T cell. To do that, the cell needs to be activated to accept the gene. The standard method uses antibodies against two surface molecules on T cells, CD3 and CD28, which together trigger powerful activation. Though this activation makes the cells receptive to the CAR gene, it also forces them to mature rapidly into short-lived effector cells that flare hot and die fast.

Goldstein compared the result to lighter fluid on a charcoal grill: a fast, intense flame that burns out before the job is done. What he wanted instead were “white hot, long-lasting charcoals to give this sustained heat.”

The new scaffold tested by Goldstein’s group bypasses T-cell receptor activation entirely. Instead, it delivers three cytokine signals simultaneously: IL-7 for survival, IL-15 superagonist for memory expansion, and IL-21 for effector function, over 7 days before the CAR gene is inserted.

During the study, a humanized mouse model of HIV was treated with the scaffold-manufactured CAR T cells, and splenic HIV was suppressed by 92%, compared to 50% for standard cells, despite equivalent numbers reaching the spleen. 

The group also tested the scaffold with T cells from six people living with HIV and found comparable transduction efficiency and a 15-fold enrichment in memory stem cell markers compared to standard manufacturing. Both products suppressed HIV infection equivalently in an in vitro assay using the donors’ own cells, by more than 75%. 

The most compelling experiment involved leukemia. The researchers tested both scaffold-manufactured and conventionally manufactured CAR T cells in mice bearing NALM-6, a human B-cell leukemia line. Both products cleared the initial tumor. But when the mice were injected with a second round of leukemia cells 3 weeks later, only the scaffold-manufactured cells responded. Five of five mice in that group cleared the rechallenge completely. On the other hand, five of six mice treated with conventionally manufactured CAR T cells failed. 

The scaffold-made cells had recognized the returning cancer, expanded, and killed again.

“That was the most important observation,” Goldstein said. “The sustained capacity to give you that protection, whereas the standard cells didn’t.”

‘Decoration Markers on Cells’

Rafick-Pierre Sékaly, PhD, an immunologist at Emory University in Atlanta, who studies T-cell memory and HIV persistence, acknowledged the need for long-lived cells but questioned whether this study proves it has found them.

Are cells that acquire memory stem cell surface markers such as CD45RO-, CCR7+, and CD95+ genuinely stem-like, meaning they can produce copies of themselves indefinitely, or are they merely dressed for the part? “Phenotypes don’t tell you anything…they’re just decoration markers on cells,” Sékaly said.

True stemness, Sékaly argued, requires gene expression data, epigenetic confirmation, and functional self-renewal validation. Goldstein noted that to look beyond surface markers, his group performed RNA sequencing on the CAR T cells from two donors. 

“We did the transcriptome analysis, and you definitely see that you have selective upregulation of the genes that are associated with T stem cells,” Goldstein said. 

Goldstein also said epigenetic profiling and confirmation are underway, but those results are not yet available. “That should really be very definitive,” Goldstein said of the pending epigenetic work. He did acknowledge that functional demonstration is the gold standard but said it is rarely achievable in preclinical work. The rechallenge in his study, he argued, provides exactly that proof: cells that proliferated and killed on second encounter with their target.

The Virus Hides

Even if the stemness question is resolved, a more fundamental problem looms.

Timothy Henrich, MD, an HIV cure researcher at UCSF, who has conducted treatment interruption and gene therapy trials, said persistence is a real limitation in the CAR T field but not the only one.

The target that Goldstein’s CAR T cells recognize, the HIV envelope protein gp120, is primarily expressed during active viral replication, when virions are budding from the cell membrane. On suppressive antiretroviral therapy, there is almost no envelope to be seen. 

Recent data presented at the Conference on Retroviruses and Opportunistic Infections in February showed that even latency-reversing agents, compounds designed to flush the reservoir into visibility, did not meaningfully increase gp120 surface expression. Without active viral replication, Henrich said, there simply isn’t enough target for CAR T cells to find.

This creates a timing dilemma. Without an antigen, even long-lived CAR T cells have nothing to engage. But when a patient stops antiretrovirals, viral rebound can explode within days. Whether memory CAR T cells can respond fast enough during the brief smoldering period before breakout is unknowable from mouse data. 

Whether memory CAR T cells can respond fast enough during that narrow window can’t be gleaned from mouse data.

“The obstacle is really the timing of when to give a CAR T cell,” Henrich said. “You need enough antigen to proliferate, to act, and to keep the CAR Ts around, but you also don’t want too much to overwhelm them. Striking that balance in HIV is going to be a big challenge.”

Still, Henrich found the approach promising, particularly for the estimated 30% of people living with HIV who cannot tolerate or access long-term antiretroviral therapy, for whom any form of cure would be better than the alternative.

Goldstein said HCW Biologics is planning to move scaffold-manufactured CAR T cells into a cancer clinical trial within a year, and discussions are underway to incorporate the manufacturing method into a future iteration of the ongoing phase 1/2a dual-CAR HIV trial at UCSF and UC Davis.

For now, the question is whether the charcoals hold their heat long enough, and whether, in HIV, there is anything left to cook.

“If you can overcome that one obstacle, then you have fewer obstacles to overcome,” Henrich said. “Will it work? I don’t know. But this wouldn’t hurt, that’s for sure.”

The study was funded by the National Institutes of Health (grants R01AI172607, R01AI174275, T32AI007501, R01CA263079) and the Einstein-Rockefeller-CUNY Center for AIDS Research (P30AI124414). Goldstein is a co-inventor on a patent application related to this work filed by HCW Biologics Inc. Henrich and Sékaly disclosed no financial conflicts of interest.