In the late 1980s, Arthur Krieg, then a fellow at the National Institutes of Health, set out to understand what happens inside B lymphocytes when they become activated by antigens. To answer this question, he decided to use a newly available strategy - treating cells with custom-designed antisense oligonucleotides to block one or another type of messenger RNA. He wanted to see if he could figure out which of the newly translated proteins are critical for B cell activation. As often happens in science, his first experiment worked like a charm. Both of his antisense oligonucleotides (ODNs) stimulated B cell proliferation and neither negative control oligo had any such effect. He wrote up his results and published them in 1989.

From Antisense to Nonsense . . . and Beyond

Then things became confusing. Krieg found he could stimulate B cell proliferation with many of the synthetic ODNs he designed, not just the supposed antisense ones. At first, he thought that the powerful stimulatory effects coming from certain negative control ODNs might stem from the fact that these really were antisense sequences, complementary to mRNAs that hadn't yet been identified. However, his attempts to find such RNAs, by screening Northern blots or searching through DNA databases, came up negative. This made Krieg wonder whether the ODNs might act by some mechanism other than antisense inhibition of specific mRNAs.

At this point, Krieg had taken a faculty position at the University of Iowa College of Medicine. Cautioned by senior faculty about the risk of chasing a potential artifact, he wrote his first grant proposal on antisense mechanisms of action. Still, he says, "these ODNs were just so potent that I couldn't believe that the finding wasn't important . . . for some reason."

Spurred on by the fact that several of the ODNs had greater potency than well-known B cell mitogens such as endotoxin, and that certain ODNs also sped up natural killer (NK) cell proliferation, Krieg started hunting for common elements among the stimulatory ODNs. He toyed with, but then dismissed, the idea that stem-loop structures were critical. Eventually he saw that the activating ODNs all had cytosine-phosphate-guanine (CpG) sequences in them. Not all the CpG ODNs had stimulatory activities. Context seemed to matter, but the dinucleotide sequence was necessary.

Might these tiny bits of synthetic DNA be mimicking some natural process in their effects on immune cells? Looking through the literature, Krieg found evidence in a paper by Yamamoto et al. that bacterial DNA had strong immunostimulatory effects [1]. These researchers attributed the immunostimulatory effect to palindromic sequences in the bacterial DNA and in the synthetic oligonucleotides they tested, but Krieg's own data failed to support the need for palindromes in the DNA. However, the data in the Yamamoto et al. paper agreed with Krieg's notion that immunostimulatory DNA requires CpGs.

Breaking the Invader's Code

Krieg, as well as other investigators, found that vertebrate DNA lacked the immunostimulatory activity of bacterial DNA. Then, a great "Ah-ha!" moment came when he recalled that CpGs occur in microbial genomes at about 1 in 16 sequences, whereas vertebrates have only 25% of that frequency of CpGs in their genomes. Moreover, the vast majority of cytosines (Cs) in CpGs are methylated in vertebrate DNA, but not in microbial DNA.

Piecing together the puzzle, Krieg guessed that his ODNs activated the immune cells because they mimicked microbial DNA. He tested the idea by making methyl-CpG versions of his most potent immunostimulatory ODN. Sure enough, they lacked the immunostimulatory effects that their unmethylated counterparts had. "That was an incredibly exciting time in my lab!" Krieg exclaimed.

Getting Down to Business

Krieg sensed that the actions of CpG ODNs on immune cells in culture might translate into interesting effects in vivo. For example, they might be useful as "adjuvants and biological response modifiers," as he noted at the end of his 1995 paper in Nature [2]. (Adjuvants are substances which, when mixed with antigenic material, heighten the immune response to the antigen.) Initially, technology transfer officers at the university suggested that Krieg start a new company, but he said he had no interest. Therefore, they considered licensing the rights to the CpG ODNs to a pharmaceutical company, and Krieg went to give a seminar on his work at the pharmaceutical firm. "As I was talking with them, I realized that they didn't really understand it, and if the university licensed it to them, it would be out of our hands and, therefore, it might never be developed," he stated.

After that meeting, Krieg told the university officials that he had changed his mind; he wanted to start a new company and was happy to have the help they had originally offered him. Important support came from Qiagen, a German biotechnology company that initially hoped to license the rights for veterinary applications. Qiagen gave seed money for proof-of-principle studies that showed that immunostimulatory effects of certain ODNs in mice required the CpG sequence and also showed that the ODNs worked on cultured human cells.

Qiagen business development head Joachim Schorr also introduced Krieg to Canadian vaccinologist Heather Davis, professor in medicine at the University of Ottawa Faculty of Medicine and principal investigator at the Loeb Health Research Institute in Ottawa, Canada, who was conducting DNA vaccine research and therefore quickly grasped the potential significance of Krieg's observations.

As Davis noted, learning about Krieg's work meant making "a huge mental shift in the way we think about the immune system and the things that might turn it on and control it. Also, his findings explained a previously unexplained event. DNA vaccines, at least in mice, work very well when only small amounts of antigen are produced. In contrast, if you give the antigen itself, you must give a whole lot more to get a response unless you add an adjuvant. Once Krieg's discovery came out, it became really clear that the DNA vaccines had their own adjuvants."

In 1997, Krieg, Schorr, and Davis cofounded the new company, CpG ImmunoPharmaceuticals. In 2000, the company was renamed Coley Pharmaceutical Group to honor turn-of-the-20th century surgeon William B. Coley, who pioneered the use of bacterial extracts as a novel means of treating cancer by stimulating the immune system.

Cancer Treatment Comes First

Since both Krieg and Davis recognized that their academic research careers had not prepared them to run a company with such enormous potential, they recruited Robert L. Bratzler, formerly cofounder of Sepracor and president and CEO of ChiRex, as CEO. Krieg and Davis assumed that the new company would emphasize the development of adjuvants for infectious disease vaccines, but Bratzler, according to Krieg, suggested exploring the contribution they might make to cancer therapies. "Even though [the CpG ODN strategy] was likely to work well for infectious disease vaccines, to build a company on that is hard. It takes a long time to get regulatory approval for new vaccine components. We may actually add more value sooner in the cancer field than we would have in infectious diseases. The time for product approval is much shorter in that field," said Krieg.

Considerable headway using CpG ODNs is already being made in cancer treatment, as Krieg noted in his talk at a recent cancer vaccine meeting. CpG ODNs may have therapeutic potential as cancer vaccine adjuvants to boost responses to known tumor antigens. The CpG ODNs may also have utility, along with monoclonal antibodies, as enhancers of cellular cytotoxicity. Finally, they can be used either in monotherapy or in combination with chemotherapy to shrink bulky tumors. Results from preclinical animal studies obtained with all these approaches show strong beneficial effects of CpG ODNs, and human clinical trials are ongoing, according to Krieg.

Infectious Disease Treatments Follow Closely Behind

CpG ODNs as adjuvants for infectious disease vaccines are by no means languishing in the background at Coley Pharmaceutical. For example, animal studies quickly showed that CpG ODNs, given with a recombinant hepatitis B virus (HBV) vaccine, produced stronger cellular and humoral (antibody) responses more rapidly and with fewer doses than did the vaccine alone [3]. Then, initial clinical trials using CpG ODNs, in conjunction with a HBV vaccine, gave similar antibody responses in humans.

A vaccine-CpG ODN combination has already helped stave off a potentially devastating HBV epidemic among endangered orangutans [4]. As Davis explained, in the late 1990s, forest fires in Indonesia forced a large number of these animals from their native habitats. The Indonesian government moved many of the threatened orangutans to a rehabilitation preserve that prepares them to be released back into the wild. When an HBV outbreak killed several orangutans in the preserve in 1998, preserve officials hoped to spare the rest through vaccination. The commercial vaccine, which works well in humans and chimpanzees, successfully immunized fewer than 10% of the vaccinated orangutans. Because the vaccine/CpG ODN combination had by then passed all preclinical toxicology testing, Davis and the Indonesian government officials decided to give the orangutans that mixture. 100 percent of the vaccinated orangutans responded. Now, all orangutans receive the combination vaccine when they arrive at the rehabilitation preserve.

Coley Pharmaceutical already has nine distinct clinical trials underway, according to Davis, and many more seem on the horizon. For example, laboratory animal data point to CpG ODNs becoming effective adjuvants not just for injected vaccines, but also for those given orally or nasally. Many infectious agents first contact our bodies at mucosal surfaces such as the respiratory or reproductive tracts or the gut, prompting vaccine developers to find ways to activate the mucosal immune system. Previous efforts to make potent oral or nasal vaccines have been thwarted by the lack of good adjuvants for these delivery routes, but recent reports from Coley Pharmaceutical show that the CpG ODNs work well in this capacity [5].

Sensing and Responding to Invaders

Krieg's detective work on the immunostimulatory actions of CpG ODNs didn't stop at the cellular level. He and his colleagues have led the field in deciphering the subcellular pathway used by these bacterial DNA pretenders. As recently reviewed [6], CpG ODN stimulation of innate immune responses involves activation of intracellular signaling pathways that sense a variety of biochemical components of microorganisms, including lipopolysaccharides and high mannose proteins. CpG ODN activation of these pathways results in a suite of cell type-specific responses. For example, B cells show increased antibody production and ignore apoptotic suicide signals whereas certain antigen-presenting cells start to produce more cytokines that in turn activate T cells. These findings explain how CpG ODNs and the microbial DNA they mimic have such potent stimulatory actions in the adaptive (antigen-specific) immune system.

In a broader sense, this tale serves as a reminder that in experimental science, the data will speak for themselves, and the scientist will succeed by paying careful attention to what the data are saying.