Hurrah for Daria Hazuda! In pioneering work done in the 1990s & 2000s she opened up a whole new class of antivirals for treating HIV, Integrase Strand Transfer Inhibitors (InSTIs). By relentlessly going after a viral enzyme people told her was “undruggable” she was able to lead the development of the first InSTI, raltegravir (Issentress), which was FDA-approved in 2007 and is now a staple of combination HIV therapies, included in the WHO’s listing of essential medicines.
Hazuda didn’t stop there. She continued climbing up the ladder at Merck and currently serves as Vice President for Infectious Diseases Discovery for Merck and Chief Scientific Officer of their MRL Cambridge Exploratory Science Center. She’s gotten a lot of awards, but she hasn’t gotten enough attention outside the HIV field – she didn’t even have a Wikipedia page. So, I made one for her and I hope it will help her get some more appreciation. You can check it out to learn more about her and today I want to tell you the story of raltegravir, because it involves beautiful biochemistry and highlights Hazuda’s scientific skills and determination.
HIV stands for Human ImmunoDeficiency Virus and it’s the virus that causes the disease AIDS (Acquired Immune Deficiency Syndrome). According to UNAIDS, 38 million people around the world are infected with HIV, 1.7 million people became newly affected in 2019, and 690,000 people died from AIDS in 2019. https://bit.ly/2BRQMD5
First reported in 1981, HIV/AIDS clearly remains a huge global problem. However, while there is no cure (yet) for AIDS, decades of research by scientists including Hazuda has turned HIV from a death sentence into a manageable chronic condition (if treatment is accessible…)
HIV is a retrovirus. Instead of “retro” in the sense of discos and roller skates, it’s retro in the sense that it “goes back” from RNA to DNA, violating the central dogma of molecular biology that says that genetic information flows from DNA (original instructions) to RNA (recipe copies) to protein. HIV travels with its genome (genetic blueprint) in the form of RNA, but then “reverse transcribes” that RNA into DNA once it gets into cells. Why? This allows it to sneak into our genome, which is written in DNA.
HIV infects a kind of immune cell called a T cell, and, once integrated it can either hang out and not really do much damage, or it can produce a lot of new virus (through viral replication) that can go infect a lot more T cells. Most treatments aim to prevent T cells from infecting new cells and thus keeping them in a silent state.
Problem is, HIV mutates really quickly, so it’s often able to develop resistance to single drug treatment regimens (monotherapies). For that reason, patients with HIV are usually treated with multi-drug “cocktails” containing several different drugs. Often, one of these drugs is a drug that Hazuda led the development of, Raltegravir, or a related “second generation” Integrase Strand Transfer Inhibitor (InSTI). But, before that, it was just the more “conventional” targets.
In the first decades of HIV treatment, before Hazuda’s groundbreaking work, anti-HIV drugs went after the HIV reverse transcriptase (the enzyme HIV uses to make a DNA copy of its RNA) or the HIV proteases (protein cutters that HIV uses to process its proteins). They’d work for a time, but resistance would occur. Scientists realized they should look for another target to add into the mix. So some of them, including Haruda, turned to a different viral protein, an enzyme (reaction mediator/speed-upper) called integrase.
Integrase is responsible for sneaking HIV DNA into a cell’s DNA. I say “sneak,” but it’s not like it can just slip in. Our genome is made up of long paired-up strands of DNA called chromosomes. So, if the virus wants to get its genome in there, it has to cut open one of those chromosomes, push the strands apart, insert itself and stitch up the seams. The virus relies on the cells’ own stitchers, but it makes a viral protein called integrase that does the DNA cutting and viral DNA insertion.
Integrase has 2 main functions
- it has to process the viral DNA ends to make them more “attacky” (this step is referred to as 3’ end processing)
- it has to break open the cell’s DNA & insert itself (this step is called strand transfer)
After reverse transcriptase does its job making a DNA copy of the HIV RNA, it’s integrase’s turn.
In the “assembly” step, Integrase assembles on specific sequences (kinda like code words) that are on the ends of the HIV-1 DNA in its “Long Terminal Repeats” (LTRs). Integrase acts as dimers (2 enzyme copies bound together) that each work on a strand and the dimers can join together to make a tetramer. This integrase + viral DNA combo is called the viral PreIntegration Complex (PIC). Once this PIC forms, integrase gets to work.
First, it chops of 2 DNA-letter (nucleotide) chunks from each 3’ end (DNA is asymmetric with one end (the 5’ end) having a free phosphate group, and the other end (the 3’ end) having a free hydroxyl (-OH) group. This all happens in the cytoplasm (general cellular interior), but the DNA is stored in a membrane-bound compartment of the cell called the nucleus. So the PIC now goes in there.
Once there, it binds to the cellular DNA (target DNA). Unlike the first step, where it occurs only at the viral ends, integrase can choose “anywhere” on the target DNA to integrate (although things like accessibility play a role and certain regions are preferred.
At this step, we have what’s called an “intasome,” which is the complex formed between integrase, the viral DNA, & the cellular DNA (basically it’s the group at the hand-off step). Now the hand-off has to happen.
In the strand transfer (aka strand joining) step, integrase “permanently” links this pre-processed viral DNA to the cellular DNA. How it works at the biochemical level is that the processed viral DNA attacks the backbone of the viral DNA at its phosphate. Since each letter can only bind to one other letter through its phosphate, the attacked DNA has to give up its previous partner. So the cell DNA breaks and gets joined to the viral DNA. Each viral strand attacks one DNA strand, so both you get them both stuck in there.
But there are gaps left at the non-attacked ends – integrase makes a staggered cut – attack sites are 5 nucleotides apart in the target strands. This leaves single-stranded gaps and the 2 unpaired nucleotides that are the result of the pretrimming.
So cellular DNA polymerases (DNA copiers) & ligases (DNA stitchers) fix those up.
And now HIV’s genome is permanently embedded in that cell, so all the cells that are made from it will inherit that viral genome (note that these aren’t gremline cells (egg & sperm cells) so HIV doesn’t get inherited in that sense). So all those cells will have the capacity to infect more cells, and the person will have the capacity to infect more people.
Therefore, if you could inhibit integrase, you could prevent the virus from infecting cells and/or people. Problem was, most people didn’t think you *could* inhibit integrase.Hazuda wasn’t the only person to try it. People at all the other pharma companies were trying as well. But, after lots of failure, they deemed it “undruggable.” Hazuda didn’t. She kept going, although she faced skepticism and pushback at almost every step along the way.
People thought it would be impossible to inhibit integrase for a couple of reasons. First off, the integration reaction carried out by integrase is “irreversible” in that integrase won’t just change its mind and pop that viral DNA back out. However, the integrase inhibitors they were testing were reversible in terms of their binding – they can bind and unbind and bind and unbind and bind…
The thought of many skeptics was that such reversible binders would be useless because all the unbindings would allow integrase to quickly do its thing. But Hazuda found that, although the drugs’ binding was reversible, it was “functionally irreversible.” Binding of the drug prevented the viral DNA from getting in long enough that the cell caught wind of that foreign piece of DNA hanging out and degraded it and/or turned it into circular non-functional viral. Therefore, the drug wouldn’t need to constantly be guarding every copy of viral DNA. Instead, it would only have to stall things long enough for backup to arrive.
But I jumped ahead of myself there, sorry. How did she find those drugs in the first place? An early problem she faced was that there was no good way to measure integrase activity – I mean, there were ways to measure binding and stuff, but there was no way to measure strand transfer. So she developed a way – an integrase activity assay! (“assay” is basically just a word for an experiment that measures something).
She attached pieces of DNA mimicking one of the viral ends (LTRs) to the wells of a plate. This stuck-on viral DNA would serve as the “donor DNA.” Then she added integrase. Integrase would bind to the DNA and do the 3’ processing step, chopping off 2 letters. The she added “target DNA” – a short segment of double-stranded DNA that integrase could help the donor DNA attack. Strand transfer would occur, joining the donor and target so that the target DNA is now stuck to the well as well. This target DNA was labeled at its ends with a molecule called biotin which they could later use to visualize how much target was bound in each well (and therefore how much strand transfer had occurred).
She could then use this experimental set-up to see if compounds could inhibit integrase – and by testing pre-cleaved & uncleaved viral DNA she could separate effects on 3’ processing from effects on strand transfer to help figure out *how* the inhibitors were inhibiting and target inhibitors of specific steps.
But she’d need to find inhibitors before she could figure out how they worked. Don’t worry, she’ll do both! But it will take a lot a lot of gruntwork…
She used her assay to add chemical compounds, one per well, and see if the signal decreased, indicating less strand transfer was occurring. She screened a LOT of compounds – being at Merck she had access to huge compound libraries. Problem was, you know how pharma companies love to show off their liquid-handling robots that can dispense drops of drugs into wells super quickly? Well, the robots available at the time weren’t compatible with this assay. So she and 2 assistants had to manually pipet the compounds – over 250,000 of them! over a period of a few months.
She found a number of hits, most containing a “1,3-diketo acid.” A ketone is where you have a carbon double-bonded to an oxygen (C=O) in the middle of a chain & “acid” here refers to carboxylic acid, which is a (C=O)-OH at a chain end.
The “DiKeto Acid” drugs proved to be effective at blocking viral integration using her assay. But the weird part was, they didn’t affect the 3’ end processing step – they were about 100-fold better at blocking strand transfer than they were at blocking 3’-processing. That was weird because other groups had shown that the same active site was responsible for carrying out both steps. So how could her drugs block one but not the other?
A key clue came when she found that the drug only bound to integrase once integrase was bound to viral DNA. Later, structures would show why. The inhibitors make crucial interactions with both the viral DNA and the viral integrase, blocking the active site magnesium and displacing the viral DNA’s 3’OH (the blades of the DNA scissors). The inhibitors also prevent the integrase/viral DNA combo from binding the DNA substrate.
But the DKAs didn’t have good drug properties in terms of solubility and “bioavailability” (how much would actually make it into your cells) and stuff. So she took those early hits and got some fellow medicinal chemists involved. These people are pros at SAR (Structure Activity Relationship). Basically, they tweak different areas of the drug in different ways and see if it makes the drug more or less effective. For example, if they replace a water-avoided (hydrophobic) group with a water-loving (hydrophilic) group and the drug stops working, it gives them a clue that maybe that hydrophobic group is important.
This, in fact, was exactly what they found. There was a part of the drug that needed to contain a hydrophobic group, which they suspected would allow the drug to bind to a hydrophobic region of integrase.
But, of course, that wasn’t all the drug needed. The hydrophobic group helped the drug bind, but in order to actually disable the enzyme, it would need to go after integrase’s “scissors.” It’s the 3’OH’s of the viral DNA ends that do the actual attacking/breaking of the cellular DNA. But they get help from a couple of magnesium ions (charged particles). The two Mg²⁺ help activate the OH’s and stabilize reaction intermediates. Their drug would need to “chelate” the Mg²⁺. “Chelate” just means bind a metal in multiple places. In this way, the drug could “hide” the Mg²⁺ from the enzyme. The diketo group of the DKAs could do just that – as could other similar groups like dihydroxypyrimidines (like what they settled on in rateglavir).
Enter skeptics again – integrase is far from the only enzyme to use Mg²⁺. In fact, a LOT of our cells’ own enzymes rely on it (as well as other metals like Zn²⁺ and Ca²⁺). People thought that the drug was just acting as a generic Mg²⁺ chelator and would therefore be nonspecific and cause more collateral damage that good.
But the other parts of the drug helped it integrase-hug! The drugs formed specific interactions with integrase (once bound to viral DNA). In addition to blocking integration in those DNA-bound-to-a-well assays, the drugs proved to be effective at blocking viral replication in a dish. And, by allowing the virus to replicate lots under selective pressure, they were able to isolate mutant viruses that were resistant to their drugs. They then sequenced the DNA of those resistant viruses and found that they all had mutations near the active site of integrase. And if you tested that mutant integrase with the compounds, even without all the rest of the virus, or if you put the mutant integrase into the backbone of a non-resistant virus, you wouldn’t get inhibition. So these experiments showed that the the drugs were specifically inhibiting integrase and that it was this inhibition that was causing the antiviral effects!
Unfortunately, problems arose when they tried to test one of their first “really promising” drugs in dogs – the dogs got sick and they had to stop the trial. They didn’t know yet, but it turned out the problem was a dog-specific one, but it had to do with the “naphthyrididne carboxamides.” It was a devastating blow. But, Hazuda isn’t one to give up, remember? And now she has more people convinced that InSTIs are possible – and they’re really close!
So they searched for alternatives – and a Merck group in Rome identified some in drugs that had been found as part of an HCV drug development program. Those drugs had been ineffective against the Hepatitis C Virus polymerase they were designed for, but they proved to be effective against HIV integrase and, with a bit of tweaking and optimization, they turned them into an effective drug, Raltegravir – initially called MK518, L-000900612 (they make and test a LOT of compounds so they don’t give them normal names until they know it’s worth it)
Raltegravir was discovered in February 2002 & approved in the US in October 2007, making it the first InSTI to get FDA-approval. in 2009, the Department of Health and Human Services (DHHS) made it a “preferred agent” for first-line treatment of HIV. Since then, additional “second generation” InSTIs have been FDA approved (elvitegravir, dolutegravir, & bictegravir) and the DHHS guidelines now recommend that an InSTI is combined with 2 nucleoside reverse transcriptase inhibitors (NRTIs) in HIV combination therapies. https://bit.ly/2BQY6Pj
Now, a bit more about Hazuda herself – copied from Wikipedia, yes, but cuz I wrote it, so…
Hazuda was raised in Hillsborough, New Jersey. Her father was an engineer and her mother worked in the regulatory-compliance division of Janssen Pharmaceutica (now a part of Johnson & Johnson) She initially pursued a premedical degree at Georgetown University, but fell in love with research during a part-time job in a lab and decided to go into drug discovery. She earned got a B.S. from Rutgers University, followed by a PhD in biochemistry from the State University of New York at Stony Brook, where she trained with Dr. Cheng-Wen Wu. She then did a post-doctoral research fellowship at Smith, Kline, and French in the department of Molecular Genetics.
Hazuda joined Merck in 1989, where she started as a Senior Research Biochemist in the antiviral research group. She was initially assigned to work on influenza, but she asked to be switched to HIV research. She continues to oversee Merck’s HIV research, which includes the development of long-acting antiretrovirals. She has also been instrumental in the development of antiviral treatments for Hepatitis C Virus (HCV), leading the development of antivirals for Hepatitis C Virus (HCV) including Elbasvir and Grazoprevir. Additionally, as Chief Scientific Officer of the Merck Research Laboratory Cambridge Exploratory Science Center she oversees research on interactions between the human microbiome and immunity. She previously served as Global Director of Scientific Affairs for Antivirals in Merck’s division of Global Human Health, as well as co-site head of basic research for the Merck West Point research facility.
Hazuda is on the editorial board of the American Chemical Society Journal on Anti-infectives Research and the Journal of Viral Eradication. She previously served on the scientific advisory boards of and the Center for Aids Research (CFAR) of UCLA and the Gladstone Institute as well as the NIH Aids Research Advisory Committee (ARAC). She is a member of the The Forum for HCV Collaborative Research, the NCI Basic Sciences Board of Scientific Counselors, and the Scientific Program Advisory Council of the American Foundation for Aids Research (AMFAR).
Hazuda has recieved the Bernie Field Lecture Award and the David Barry DART (Development of Antiretroviral Therapies) Achievement Award. The integrase inhibitor she led the development of, Isentress (ralteglavir), was awarded the Prix Galien prize in 2008. She was awarded the Italian Premio Galeno Award for this work. In 2019 she Ohio State University presented her with the Distinguished Research Career Award. In 2017, she was part of a team of chemists awarded the “Heroes of Chemistry” award from the American Chemical Society for the development of the HCV combination therapy Elbasvir/Grazoprevir. She was Elected as a Fellow to the American Academy of Microbiology in 2010.
here are links to those key papers I show in the figures:
Hazuda, et. al, Inhibitors of Strand Transfer That Prevent Integration and Inhibit HIV-1 Replication in Cells, Science, 2000 https://science.sciencemag.org/content/287/5453/646.long
Espeseth, et. al, HIV-1 integrase inhibitors that compete with the target DNA substrate define a unique strand transfer conformation for integrase, PNAS, 2000 https://doi.org/10.1073/pnas.200139397
Grobler et. al, Diketo acid inhibitor mechanism and HIV-1 integrase: Implications for metal binding in the active site of phosphotransferase enzymes, PNAS, 2002 https://doi.org/10.1073/pnas.092056199
Summa et. al, Discovery of Raltegravir, a Potent, Selective Orally Bioavailable HIV-Integrase Inhibitor for the Treatment of HIV-AIDS Infection, J. Med. Chem, 2008 https://doi.org/10.1021/jm800245z
You can also hear Hazuda tell the story in her own words. That’s how I learned about her – she gave a great talk at the CSHL History of HIV/AIDS research meeting a few years ago that is free to watch: https://bit.ly/39STMeP
and, of course, check out her Wikipedia article (and improve it!) https://en.wikipedia.org/wiki/Daria_Hazuda
more on topics mentioned (& others) #365DaysOfScience All (with topics listed) 👉 http://bit.ly/2OllAB0