|A Long Shot on Cytomegalovirus
An unlikely vaccinologist fights an unlikely foe.
Gerd Maul was frustrated. It was 1989, and the German-born researcher at Philadelphia's Wistar Institute couldn't develop antibodies to the nuclear pore proteins that had been a focus of his study for nearly a decade. With funding getting tight, he consulted a colleague, Sergio Jimenez, then across the street at the University of Pennsylvania, who worked with patients with the autoimmune disease scleroderma. Maul knew that the human immune system could develop antibodies to very conserved proteins that rabbits (the species he was using) would never develop. So, he asked for samples.
Describing ND10 posed a challenge; generating interest in them posed another. Maul could explain their resistance to nuclease or their disappearance and reappearance during the cell cycle, but little else. So he hit them with UV light, heat shock ... "I bombarded the cells with everything." Nigel Fraser, a Wistar virologist at the time, helped Maul infect the spotty cells with herpes simplex virus. It caused the dots to disappear, as did several other viruses. "We now had a connection that was fundable," Maul says.
Throughout the early 1990s Maul and others tracked the components of ND10 biochemically and genetically, discovering three interferon-induced proteins: sp100, Daxx, and promyelocytic leukemia protein (PML) in 1994.2 With a link to cancer via PML, dozens of research teams began to study the dots. That essentially solved Maul's problem with the moniker: They're now called PML oncodomains (PODs) or PML bodies.
Maul continued to study viral interactions with these spots as a way to probe their function. In the early 1990s, he convinced virologist Roger Everett of the Medical Research Council Virology Unit that PML bodies were important in viral infection. "He was our first convert," Maul says. Maul joined Everett in Glasgow, and the two began to query the activity of a herpes simplex virus, HSV1, as it interacts with PML bodies. An immediate early gene of HSV1 - Vmw110, which is required for waking the virus from latency - shares a zinc-binding motif with PML and also colocalizes with the PML bodies during infection, redistributing their protein content.3
©2006 R.D.R. - Custom Medical Stock Photo
PML bodies seemed to be one of the cell's intrinsic defenses against viruses, but the viruses appeared to have evolved adaptations for dealing with them. Herpesviruses, as well as other DNA viruses such as adenoviruses and papova-viruses, appear to initiate replication and transcription at the periphery of these bodies. Maul found that cytomegalovirus (CMV) expresses an immediate early gene, pp71, which derails PML body activity by targeting the Daxx subunit for destruction.4
Several antagonistic reactions take place as CMV enters the nucleus (see "Developing a Hybrid Vaccine,"), and Maul's nuclear dots served as proving grounds for many of these molecular battles. Transcription takes place at their periphery, and the PML bodies disappear at the beginning of viral replication. Moreover, in cell types for which the virus cannot complete its life cycle, PML bodies appear to be the barrier.
The specificity of the interactions hints at a long evolutionary arms race. "These are huge viruses; they have a large arsenal of defenses against the defenses of the humans," says Maul. Human CMV has a genome roughly 235 kilobases long and encodes an estimated 165 genes.5 For comparison, poliovirus has 11 gene products, and some retroviruses have only three. "Since fish, we have lived with this evolution-wise," Maul adds, and parallel battles between CMVs infecting different species have made the systems highly specialized.
If Maul has his way, this specificity will cause the downfall of the virus. This past June, Maul wrote a grant to the National Institutes of Health to fund his first foray into vaccine development. The project's goal might seem a bit quixotic: to rid the world of CMV, a virus most people caught years ago and will likely never notice. Depending on locale and socioeconomic status, between 50% and 90% of the general population has antibodies against CMV, but few have complications.
Complications can arise, but with long-shot odds: the 0.1% born with human CMV disease. Of four million live births in the United States annually, 1% pass congenital CMV, and 5%-10% of that subset, as many as 4,000 infants annually, develop catastrophic problems including microencephaly, encephalitis, seizures, deafness, varying degrees of mental retardation, or death.5
The slim percentage doesn't mean that the stakes are low. "Add [4,000 cases] every year, and that is then a cost over a lifetime, and the attendant costs of taking care of them, the attendant costs of the caregivers even, who are less productive in other ventures," says Maul. The US Centers for Disease Control and Prevention lists CMV as one of the top causes of birth defects, on par with Down syndrome, fetal alcohol syndrome, and neural tube defects. Beyond congenital cases, CMV is also a major cause of morbidity and organ rejection in transplant recipients, and it can be lethal to people infected with HIV.
In 1999, the National Academy of Sciences tallied the bill. For an infant with congenital CMV, costs include $7,000-$12,000 in initial hospital expenses, plus a lifelong drain of $8,000 per year for special schooling and $225 a day for long-term care and lost productivity. For a transplant patient, severe CMV disease can tack $14,000 onto a hospital bill, plus the cost of follow-up care.
Totaling the damage, the National Academy of Sciences (NAS) reached the startling conclusion that this supposedly benign virus actually costs the US economy $4.4 billion each year. An NAS report lists CMV as one of seven top-priority vaccine targets, and places CMV alongside more infamous diseases such as insulin-dependent diabetes and group B streptococcus.
Conquering CMV won't be easy. The virus stymied Stanley Plotkin, the Wistar vaccinologist who developed highly successful rubella and rotavirus vaccines. His attenuated strain of CMV appeared to induce strong immune responses, but it failed to protect patients against infection in a clinical trial. "Is antibody sufficient, or do you need cellular immune responses? Or, are cellular immune responses sufficient and you don't need antibody? Those are unresolved issues that have somewhat hindered CMV vaccine development," says Plotkin.
Some say that with the low hanging fruit all but gone, vaccinology has slid into a relatively dormant period. "I think historically the vaccines that were easy to make using simple technologies have been made. I think the reason why the vaccine market has not expanded is because it didn't have the technology," says Jeff Chulay, chief medical officer at AlphaVax in Research Triangle Park, NC. But buoyed by the NAS report and other developments (see "An Economic Boost for an Unlikely Target"), several pharmaceutical companies have begun to revitalize their CMV vaccine programs (see "CMV in Pharma's Sights," below), using various strategies.
Just as Maul's path to the virus was circuitous, his strategy to defeat it is less than conventional. He speaks of his work in meandering way, articulating parallels that don't sound appropriate even when they are. In an excitable German accent, he spouts phrases like "star spangled sky," familiar yet strange. He puzzles over words that he's sure don't have an English counterpart and beams delightedly when he finds one. With a calmer tone he'll relay distressing news bluntly, but not without feeling. His lab quite recently had eight people, he says, as he leads the way through perhaps 65 square meters of quiet lab space. "There are now three," he says with regret. But mostly, he is roused by the prospect of discovery. "We jump to a thing that challenges us," and from there, he says, energy expended elsewhere is waste.
Francois Nosten points across the Moei River into Burma.
|"These are huge viruses; they have a large arsenal of defenses against the defenses of the humans."
- Gerd Maul
Maul's current challenge is to determine whether murine CMV (MCMV) - which can get into human cells, but can't replicate lytically - could be the basis of a human vaccine. The plan, which underlies his grant proposal, is to "humanize" the mouse virus enough to work as an attenuated human CMV vaccine.
In principle, researchers could add as many as 10 to 15 genes to MCMV's large genome, allowing it to express other viral epitopes. Its poor infectivity in human cells should make it safe, but before proceeding Maul wants to be certain that the species barrier is robust.
Conveniently, work on PML bodies provided a critical new tool for studying the barrier. Disrupting the Daxx gene in mouse cells makes the cells much more sensitive to CMV infection than in wild-type cells. Using these hypersensitive cells as their assay, the researchers tried to pinpoint the block to infection and extrapolate the precise step where mouse virus replication becomes derailed in human cells.
To their surprise, they did not find a specific step. Murine CMV can complete each step of its life cycle, from entry to the release of infectious particles, in human cells. But the virus gets through the whole cycle on only rare occasions, which is why previous studies had not uncovered this leak in the species barrier. By adding several proteins from the human virus, including immediate early protein 1 (IE1, a counterpart to HSV1's Vmw110), replication of MCMV in human cells is boosted dramatically.6
While probing the species barrier, Maul and his collaborators were also analyzing the mouse CMV genome.7 Fully sequenced more than a decade ago, murine CMV contains 170 predicted open-reading frames (ORFs), but nobody had done a comprehensive study to prove that all the ORFs are expressed. Using microarrays and RT-PCR, Maul and his colleagues found 172 transcribed genes, including seven that hadn't been predicted in the ORF analysis.
Many of CMV's extra genes appear to have evolved to deflect host defenses, while others form the Byzantine regulatory network that governs latency and reactivation. "The virus has proteins [that] actually segregate or sequester histone deacetylases," says Maul, which may help CMV block some host defenses at the transcriptional level.
The virus also has internal blocks. Induction of the IE gene is essential for viral production. The splice variant IE1 turns on production of IE2, which in turn inhibits IE1. Furthermore, the early protein 112/113 inhibits IE2, causing a long delay (nearly 24 hours in humans) between immediate early gene transcription and replication of the viral genome. Meanwhile, the virus gathers massive quantities of proteins where PML bodies dissipate, without producing viral DNA. "It doesn't need that much protein," Maul muses. Why does it do this?
He's fascinated by the puzzles, but confides that he's fearful he won't have the money to investigate them further. Maul notes that most conversations with colleagues drift into commiseration about the state of federal funding. "It usually comes in by about the sixth sentence." As we talked on a warm November afternoon, Maul confided that he had received somber news that morning: His NIH grant to develop a vaccine was rejected.
This is the third time in his career that funding has bottomed out, but Maul is undeterred. He's looking into the prospect of reversing his hybrid virus plan. Plotkin, who serves as an advisor to industry, has been persuading him to develop a human virus that will infect mice, a project that could gain some interest from the CMV vaccine developers looking to test their wares in a preclinical model. Now, Maul seems resigned to try it, and of course, he'll resubmit his grant, even if it seems like a long shot.
Maul says his murine CMV model has about a 50% chance of working. Optimism returning, he smiles. "Many people would definitely go to Atlantic City if they had a 50% chance of winning."
AlphaVax in Research Triangle Park, NC, is one of a handful of companies taking up the cytomegalovirus (CMV) vaccine challenge. The company's approach is to insert CMV antigens, glycoprotein B and pp65, into the genome of a modified alphavirus. "Because we're [using] a single-cycle vector that does not persist, we're safer than many other types of vaccines, such as live attenuated vaccines. And because it's an RNA vector ... there's no risk of integration to the host DNA," says Jeff Chulay, chief medical officer at AlphaVax.
The company's CMV vaccine is still at the preclinical stage, where one of the major challenges is the lack of a good animal model. Mice, guinea pigs, and many other animals have their own species of CMV, but as research from Gerd Maul at Wistar Institute and others have shown, the viruses and the hosts' responses to them have evolved in distinct ways, making it difficult to draw clear parallels between different systems (see story above). AlphaVax, for example, has shown that a vaccine carrying antigens of guinea pig CMV can protect the animals from passing the virus to their offspring, but it's unclear whether that's an accurate prediction of clinical success.
At least one company, Vical of San Diego, has already cleared the preclinical hurdle. The Vical vaccine is a DNA vector carrying glycoprotein B and pp65. Once injected into a muscle, the vector expresses the viral genes, and in a Phase I trial, patients developed both antibody and T-cell responses against the antigens. Vijay Samant, president and CEO of Vical, is optimistic about the product, and the company is now recruiting volunteers for a Phase II trial.
Vical is focusing initially on a small subset of CMV disease. "The early applications are people with bone marrow transplants, [in whom] it's a leading cause of graft rejection," says Samant. Clinical complications of CMV are relatively common in transplant patients, allowing the company to get statistically meaningful results from a relatively small trial population. In the trial, donors of hematopoietic cells will get the experimental vaccine before bone marrow extraction, in the hope that CMV disease can be prevented in the immunosuppressed cell recipients.
Stanley Plotkin, who developed several vaccines while at the Wistar Institute in Philadelphia, and who now advises Sanofi Aventis, is guarded but optimistic about Vical's work. "I like the approach, but of course DNA vaccines in human medicine have not so far been adopted," he says. "At this stage it's a good thing to have a number of different approaches."
1. C.A. Ascoli, G.G. Maul, "Identification of a novel nuclear domain," J Cell Biol, 112:785-95, 1991.[Pubmed]
2. J.A. Dyck et al., "A novel macromolecular structure is a target of the promyelocyte-retinoic acid receptor oncoprotein," Cell, 76:333-43, 1994.[Pubmed]
3. R.D. Everett, G.G. Maul, "HSV-1 IE protein Vmw110 causes redistribution of PML," EMBO J, 13:5062-9, 1994.[Pubmed]
4. R.T. Saffert, R.F. Kalejta, "Inactivating a cellular intrinsic immune defense mediated by Daxx is the mechanism through which the human cytomegalovirus pp71 protein stimulates viral immediate-early gene expression," J Virol, 80:3863-71, 2006.[Pubmed]
5. R. Khanna, D.J. Diamond, "Human cytomegalovirus vaccine: time to look for alternative options," Trends Mol Med, 12:27-33, 2006.[Pubmed]
6. Q. Tang, G.G. Maul, "Mouse cytomegalovirus crosses the species barrier with help from a few human cytomegalovirus proteins," J Virol, 80:7510-21, August 2006.[Pubmed]
7. Q. Tang et al., "Experimental confirmation of global murine cytomegalovirus open reading frames by transcriptional detection and partial characterization of newly described gene products," J Virol, 80:6873-82, July 2006.[Pubmed]
Visit www.the-scientist.com this month for info on Gerd Maul's recent art show in Philadelphia, a podcast interview on his work with murine CMV, and more.