The Confidence Booster
Dan Barouch’s lab is a bustling hive working on complex initiatives in many areas of HIV research. On the brink of a big vaccine study, Barouch reflects on the painstaking steps required to build a formidable investigative operation—and to feel sure about decisions in a field known for uncertainty.
By Michael Dumiak
A cutting wind has Boston commuters in scarves, but in a gleaming glass tower on Blackfan Circle, the 10th floor laboratory is orderly and calm. This is where Dan Barouch, in just a short time, has built a formidable investigative operation. Barouch oversees a 57-person research group active in HIV pathogenesis and basic science, vaccine research and development, and cure research, as well as working on emerging pathogens such as Ebola and, more recently, Zika.
Barouch is a clinician and Harvard Medical School professor. He’s director of the Center for Virology and Vaccine Research (CVVR) at Beth Israel Deaconess Medical Center and a founding member of the Ragon Institute, which is a joint research endowment of Massachusetts General, Harvard, and the Massachusetts Institute of Technology.
“He’s built a very large group that covers all the bases in terms of taking things from the concept stage to early testing and all the way forward into clinical trials,” says Bruce Walker, Ragon Institute director. “I don’t really know of another example of someone who’s done this in such an independent way. Certainly nobody at that young age.” Barouch just turned 43.
Barouch’s always been on the young side among his peers. It’s a topic that comes up often, though, he chuckles, it won’t always be like this. “I’ve often found myself to be younger than a lot of the people I sit with in various conference rooms.” He didn’t get to this point overnight. It just feels that way.
Barouch finished his clinical fellowship as an infectious disease specialist 14 years ago at Brigham & Women’s Hospital. Becoming a doctor was a lifetime goal—he still does hospital rounds once a month. Before this Barouch went to Oxford to pursue a doctorate in microbiology. He finished it in two years, then turned to medical school.
While doing his medical school residency Barouch began working with Norm Letvin, the late Harvard HIV scientist renowned for research with nonhuman primates. With Letvin, Barouch was able to publish papers and establish a research record. Barouch applied for career development grants from the US National Institutes of Health (NIH) and the Doris Duke Charitable Foundation. He got them both. Letvin gave Barouch space in a fusty prewar building that Beth Israel was renting from Emmanuel College. He was able to hire two technicians with the grants. Just like that, Barouch started his own research lab at 29 and began working toward developing an HIV vaccine.
When he’s home in suburban Newton, Barouch begins every day practicing the violin with his two young daughters. When he’s traveling, Barouch’s wife, Fina, an ophthalmologist and retinal surgeon at Lahey Hospital & Medical Center, plays music with the girls. Eight-year-old Susanna is learning Beethoven’s Minuet in G minor, while five-year-old Natalie is learning “Twinkle Twinkle Little Star.” As someone who started playing the violin when he was four, Barouch can relate.
He arrives at work on this windblown March day in a good mood. “It’s all fun. Every morning I look forward to coming to work. There’s not a single day that goes by uneventfully. Every day something new happens. It’s exciting,” he says, with a slight sniffle, sipping coffee at eight in the morning. Bustling and cluttered inside, the 10th-floor CVVR sits in an arching glass building designed by Boston architects Tsoi / Kobus & Associates. The sleek complex also houses Pfizer’s research and development facilities and the Wyss Institute for Biologically Inspired Engineering. Barouch’s broad desk at the CVVR comes with a view over the Harvard Institutes of Medicine. The Ionic columns of Harvard Medical School’s Gordon Hall are just visible. With close-cropped dark hair and strong eyebrows, Barouch is focused and crisp. He is youthful and friendly, but not overly so; he seems intent on keeping composure at all times, but is not anxious.
A little anxiety wouldn’t be surprising given the amount of activity Barouch is keeping track of. Over the spring he and his team readied for what could be some of its biggest exploits yet. The CVVR recently initiated an in-house Phase I HIV vaccine clinical trial, showing the lab’s capability to take its own basic science work into translational human research. Barouch is also part of a collaboration that’s just been awarded a US$42 million grant for HIV vaccine and cure research that will make extensive use of the team’s nonhuman primate program, with Barouch and Louis Picker of the Oregon Health & Science University as co-primary investigators.
Then there is the lab’s role in bringing its leading vaccine candidate into efficacy trials in collaboration with Johnson & Johnson (J&J), the National Institute of Allergy and Infectious Diseases (NIAID), the HIV Vaccine Trials Network (HVTN), and multiple clinical partners. The vaccine candidate is already being tested in a series of Phase I/IIa clinical trials, and pending the results, Phase IIb efficacy trials could start next year. If effective, the vaccine candidate is on a path to potential development and production, given J&J’s industrial manufacturing capabilities. This would be a rarified stage—only a handful of HIV vaccine candidates have made it this far.
Barouch’s weekly team meetings begin in a boisterous atmosphere. But he quickly turns to business. The schedule of ongoing studies run down every Friday represents a broad amount of scientific effort and dozens of precisely tracked ongoing experiments. On shelves next to Barouch’s desk are hundreds of plain manila folders. “These are just the data from primates,” Barouch says offhandedly.
Barouch did not set out to build a big research group, but the work required it. NIH funding provides the bulk of money to the lab. Barouch has administrative staff to help manage these large and complex grant budgets, but he still does his own grant writing. In the last five years the Barouch lab has received more than $80 million in NIH funding, averaging about $15 million a year. The group also gets substantial funding from the Department of Defense, amfAR, the Ragon Institute, and other industry and philanthropic sources such as the Bill & Melinda Gates Foundation.
Down the narrow hallway from Barouch’s office, past a shared shotgun office dubbed ‘Techtopia’ with a half-dozen lab techs poring over data, the wing opens up into lab space. In one corner the protein group is assembling. In another area, two researchers from the virology group, Peter Abbink and Michael Boyd, are opening up the vacuum manifold on a Promega filtering unit. Boyd will use the unit along with a Zymo research kit to purify RNA for the assays used to measure HIV viral loads in experiments and clinical trials. Boyd is a research assistant; Abbink, a Dutchman, is the virology lab manager, the Barouch lab’s longest-serving member, and is introduced, somewhat jokingly, as ‘The Master of the Vectors.’
This refers to his work in developing the viral vector used in the Barouch lab’s vaccine candidate now in clinical trials. Abbink and the virology group make vectors based on adenovirus serotype 26 (Ad26), a strain of the common-cold virus. Barouch’s experiments with Ad26 are one of the defining stories of his lab and career so far.
Interest in adenovirus vectors took him to the Netherlands in the summer of 2004. He’d just finished his medical studies and Letvin knew that Barouch was interested in adenoviruses and their potential in HIV vaccine research. Letvin introduced him to Jaap Goudsmit, who at the time was chief medical officer of a biotech company called Crucell, located in the old Dutch university town of Leiden.
“I said, ‘you want to work with adenoviruses? Come spend a couple months with us,’” Goudsmit recalls. So started what would become a key partnership even when Crucell was eventually bought by J&J’s pharma subsidiary, Janssen. Janssen is now a vital partner of Barouch’s and the network of collaborators testing their Ad26-based candidate in combination with other candidates in clinical trials. “Janssen is directing late-phase manufacturing and clinical development of this vaccine. That’s what we need. In the HIV vaccine field, we need more industry involvement, not less. There is no way an academic group can conduct all the activities for late-phase development. It requires industry,” Barouch says.
Of course back when Barouch first went to Leiden he didn’t know what was coming. He just wanted to learn. “I learned how to grow and clone adenovirus vectors with my own two hands,” he says. It was a productive stay. Crucell had an intern that helped Barouch that summer. By the time Barouch went back to Boston, he’d made an agreement with Crucell: they would provide the viral particles, cell lines, and the DNA plasmids necessary for producing new vectors. Barouch also offered the Crucell intern a job in Boston. That’s how Peter Abbink became the Vector Master.
An arduous time for adenovirus
For many years now researchers have harbored hopes of utilizing attenuated viruses as vectors to deliver HIV antigens. Using a live-attenuated version of HIV itself is not feasible as there’s too much risk that the virus could mutate and regain its pathogenicity. Several different viral vectors are being investigated, including adenovirus, and there are many different strains or variants of adenovirus, called serotypes. Adenovirus serotypes 5 (Ad5) and 26 are two strains that have been tested extensively as vectors.
A replication-deficient Ad5 vector developed by Merck was first tested in the STEP trial in partnership with NIAID and the HVTN. The study involved 3,000 volunteers from a diverse population at high risk of HIV infection, including men who have sex with men and female sex workers between ages 18 and 45. A companion to the STEP trial called Phambili tested the same Ad5 vaccine candidate in a different population of high-risk men and women in South Africa. At the time these trials started, Barouch had just returned from his sojourn in the Netherlands. Researchers were upbeat about the prospects for Ad vectors, including Merck’s Ad5 candidate.
“It was by far the most immunogenic vaccine, and to this date still one of the best vaccines for inducing T-cell responses,” says Nicole Frahm, associate director for laboratory science at the HVTN. Previous efficacy trials focused on stimulating antibody responses, but the Merck candidate was designed to stimulate only cellular immune responses. The logic was that even if a strong T-cell response couldn’t protect against infection, it might help reduce the severity of disease in vaccinated individuals who still became HIV infected.
But three years into STEP its safety committee stopped immunizations. Phambili was halted too. The data indicated that in STEP there was a higher infection rate in the vaccinated group than in placebo recipients.
Frahm, who had left studying HIV pathenogenesis as a postdoc in Bruce Walker’s lab to take up vaccine research at the HVTN just as STEP was closing, remembers the sadness that pervaded the field at that time. “It was depressing,” she says. “This was one of those vaccines where everybody thought you would see a positive signal because it was so immunogenic.”
A detailed analysis of the STEP results showed that the higher risk of HIV infection in the vaccinated volunteers seemed to come from a subgroup of uncircumcised male volunteers, who, due to being previously exposed to the virus, were already sero-positive for Ad5.
After STEP, a troubling hypothesis gained currency: that Ad5 vectors may recruit activated CD4 T cells to mucosal tissues, thereby increasing the number of target cells for the virus to infect and increasing the chance of HIV infection occurring.
So far no definitive data supporting this hypothesis has come to light. Frahm’s group, which is about to publish findings from studying Ad5, has found no evidence for this either. “We have looked extensively in the mucosa,” she says. The bottom line is that researchers still don’t know what happened.
While Frahm’s lab and others were racing to understand what had happened in the STEP trial, another large-scale Ad5 vaccine trial began. HVTN 505 tested a DNA prime/Ad5-based vaccine candidate boost that was similar but distinct from the one tested in STEP study. This trial also limited itself to circumcised men who have sex with men and whom did not have pre-existing Ad5 immunity at the time of enrollment. Even so, four years into the study, 505 ground to a halt because there was no difference in the HIV infection rate between the vaccine and placebo groups.
The results of these trials caused researchers to question the use of adenovirus vectors altogether. In the summer of 2013, NIAID held a meeting to discuss their future.
Through all this, Dan Barouch and his rapidly growing team watched closely. “We all took a long, hard look at the vaccine field and what made sense to do and what did not,” he says. “At that moment in time it was not entirely obvious that we were going to continue to develop Ad vectors.”
What Barouch and his team did was start investigating alternate Ad serotypes. They researched how different the various Ad serotypes looked to see if there was any evidence that the Ad26 vector the lab was zeroing in on was any better than Ad5. They also tested whether Ad26 vectors would recapitulate the safety concerns seen with Ad5.
From 2007 on, Barouch’s group published paper after paper outlining experiments with Ad vectors. These were summarized in a perspective co-authored by Picker in which, drawing on the previous experiments, they outlined the differences between Ad5 and Ad26 (Nat. Rev. Immun. 12, 765, 2014). The review outlines biological differences between the two serotypes, differences in human exposure to adenovirus strains, and differences in the innate and adaptive immune responses they induce, including in the mucosa. Based on these findings, Barouch pushed forward with Ad26.
Colonel Nelson Michael was drinking coffee in a hotel lobby in Seattle’s South Lake Union neighborhood getting ready to attend a meeting at the Gates Foundation. It was 2010 and Michael, director of the US Military HIV Research Program (MHRP), had begun to work with Barouch and a few others on monkey experiments testing the combination of Barouch’s Ad26 candidate with another viral-vector based candidate developed by MHRP scientists. Barouch was in Seattle for the Gates meeting too, staying at the same hotel as Michael. He came down from his room, laptop in hand, excited, Michael recalls. The data Barouch shared with Michael showed that together the two vaccine candidates could protect monkeys against a stringent challenge. “It was so exciting! I’ll never forget how pumped up we were,” Michael says. “In those moments Dan has a childlike excitement about data that I find endearing. He’s a fun person to work with. You don’t have to like people you collaborate with, but it’s a joy when you do.”
Michael first got to know Barouch in 2008 at a scientific meeting run by immunologist Barton Haynes at the Duke Center for HIV/AIDS Vaccine Immunology (CHAVI). Barouch was still working with Ad5 at the time and Michael was in the midst of RV144, which tested a canarypox virus candidate known as ALVAC along with a gp120 boost. The RV144 trial would go on to show modest efficacy.
Michael was also working with another poxvirus vector: modified vaccinia Ankara (MVA). “You get more CD8 responses from adenoviruses,” Michael says, “and more CD4 responses out of MVA.” According to Michael, the MVA vector itself, even just with repetitive vaccinations, can also generate antibodies.
The two began talking about combining their appraoches, that is priming with Barouch’s Ad vector and boosting with Michael’s MVA-based candidate. The collaboration proved fruitful. First, monkey experiments that Barouch previewed with Michael back in 2010 showed protection (Nature 512, 74, 2014). Now, the combination of Barouch’s Ad26-vectored mosaic Env/Gag/Pol prime, MHRP’s MVA vectored candidate expressing the same mosaic antigens, and an HIV gp140 protein boost are being tested in the Phase I/IIa study called APPROACH. The Env protein boost was added because of the RV144 efficacy data and monkey studies showing the protective efficacy of an Ad26/gp140 regimen (Science 349, 320, 2015). This study is a big effort. It involves researchers from multiple institutions—Beth Israel, MHRP, the HVTN, the International AIDS Vaccine Initiative (IAVI), and Janssen. With the trial now fully enrolled, these institutions are collecting immunogenicity data on the experimental multivalent HIV vaccine candidates from 400 healthy volunteers at 14 sites in the US, Rwanda, Uganda, South Africa, and Thailand.
APPROACH is also a complex study: It involves eight arms, each receiving different regimens of the vaccine candidates or placebo with immunizations at 0, 3, 6, and 12 months. The regimens include: Ad26 mosaic Env/Gag/Pol candidate alone, or the Ad 26 candidate boosted by the MVA vectored candidate expressing the same antigens followed by either a high- or low-dose clade C gp140 boost with alum adjuvant. The mosaic antigens were developed as a result of another collaboration, this one between Barouch and Bette Korber of the Los Alamos National Laboratory. Mosaics are bioinformatically engineered HIV protein sequences optimized to cover the diversity of HIV around the world.
The Phase I/IIa results are expected sometime later this year, but even before the results are in, Barouch’s group has an eye toward simplifying this seemingly complex vaccine regimen.
APPROACH tests what would be a year-long, four-shot vaccine regimen, which would be challenging to administer in the parts of the world where HIV is most prevalent. This is why Barouch’s group is conducting IPCAVD 010. This is the first clinical trial the group is conducting wholly in-house. It is funded by the Ragon Institute, which provides flexible funds to support pre-clinical and clinical HIV research. The goal of IPCAVD 010 is to test whether the APPROACH regimen can be simplified: the same vaccine candidates will be administered over three or six months in IPCAVD 010 instead of the year-long administration being tested in APPROACH. If comparable immunogenicity is seen with this shorter regimen, it would be cheaper and easier to implement should the vaccine prove effective.
The APPROACH collaboration is just one example of how Barouch operates so effectively within an often competitive field. Barouch is also working with other companies and collaborators to advance efforts to test broadly neutralizing monoclonal antibodies for their ability to prevent, treat, or even help cure HIV infection.
A few years ago Barouch was talking with Dennis Burton, a professor of immunology and microbial science at The Scripps Research Institute (TSRI) in La Jolla, CA, and head of the IAVI Neutralizing Antibody Center. Burton had done protection studies with the broadly neutralizing antibody (bNAb) PGT121, isolated from an IAVI cohort, in 18 rhesus macaques. But as they talked, the two began wondering whether the antibody could be utilized in a therapeutic context. “Instead of wondering what it would do in an uninfected animal, what would it do to an infected animal? That’s what we were asking,” Burton says.
At the time there was a lot of skepticism, based on prior research results, that antibodies could be effective against established infection. But Burton and Barouch wondered whether the new crop of more potent bNAbs being isolated at breakneck speed since 2008 would do better. The two decided to give PGT121 to a group of monkeys infected with a simian immunodeficiency virus/HIV hybrid. What they saw was a dramatic therapeutic effect: virus levels in the animals were suppressed up to three logs a week after antibody administration. Even after the antibody was gone, some of the animals with the lowest viral loads at the start had undetectable viral loads. “The antibody appeared to reduce virally infected cells as well as free virus,” Barouch says. The experiment showed viral suppression for a median of 56 days even after the antibody was gone. Three of the monkeys never did rebound with infection (Nature 503, 224, 2013).
Now Barouch has a call every two weeks with the contract manufacturer Catalent, which is nearly finished manufacturing PGT121 for clinical trials funded by the Gates Foundation to test the safety, pharmacokinetics, and efficacy of the antibody in both HIV-infected and uninfected volunteers. This work is also a collaborative effort, involving researchers from Beth Israel, the Ragon Institute, TSRI, Theraclone Sciences, Gilead Sciences, and IAVI. The Ragon Institute provided the initial funding for the PGT121 manufacturing. Another bNAb, PGDM 1400, is also being manufactured. It was more recently isolated by researchers at IAVI’s Neutralizing Antibody Center at TSRI.
Rising to the top
Michael recalls something else about getting to know Barouch at the CHAVI meeting. Barouch was wondering whether he’d have an impact: there were a lot of big-time scientists in the HIV field there, and the young researcher said he didn’t know if the adenoviruses would work or not. “Even if they do,” Michael recalls him saying. “I don’t know that there will ever be room for someone as junior as me.”
Michael laughed. “I said if your data is good, there’s always going to be room.”
Barouch’s not only had good data, he’s made a point of communicating it clearly. “What’s critical for running a research group is clarity. So much about science and research is communicating the findings to others,” he says. That includes publications, grant applications, and public speaking. Barouch says he was once quite shy but made the effort to learn how to speak in front of people (mostly, by speaking in front of people). Barouch presents on his work in one form or another once a week at least. He takes teaching seriously, spending time with the graduate students in his lab. On a Monday morning at nine o’clock they’re chatting and queueing outside his door as though it was a bagel truck.
He also spends much of his time travelling—in the last month he’s been to Portugal, Marseille, Paris, San Diego, and Washington, DC. He also attended two conferences in Boston. It’s a punishing schedule, but he doesn’t stay in the lab late at night as he once did. “I leave here at six, most of the time. I spend time with my family,” he says. They ski. They go to Hawaii on vacation. He does work every evening for several hours after the girls are in bed, and also in those hours he spends travelling. “I do my best work on the plane. A good six-hour plane trip, I do like that.”
Barouch comes from a bright family: his mother Winifred was a homemaker until the children went to school, then she went back for her PhD in biochemistry and worked for the NIH. Barouch’s father, Eytan, is a math lecturer and flow lithography engineer, teaching at Clarkson University in Potsdam, New York, and started his own company, Vector Technologies. Barouch’s sister, Lili, is a cardiologist at Johns Hopkins. It’s likely they stood out growing up in Potsdam, which is so far upstate it’s possible to bike to the St. Lawrence River, the rolling border with Canada and a lumberjacking lodestar. Barouch found a place in Potsdam, though, starring in the public Potsdam Senior High School’s math and science competitions.
“The striking thing about Dan was that his experiments always worked,” says Sir Andrew McMichael, a pioneering HIV researcher who was Barouch’s doctoral supervisor at Oxford, and who introduced Barouch to Norm Letvin. “Some students run into problems here or there. They need to work things out. Three months go by.”
McMichael says some researchers are laid back and then have a rash of results. Others bash away all the time. He counts Barouch as more of a basher.
“He was in a hurry,” he says. “Very bright, wanted to get on to do things. To make a difference. To do, and push forward. He was always pushing forward.”
Michael Dumiak reports on global science, technology, and public health and is based in Berlin.