The Recombinant Decoy in Baker's Yeast

Baruch Blumberg, Maurice Hilleman, and the First Anti-Cancer Subunit Vaccine

Under transmission electron microscopy, serum drawn from a chronic carrier of Hepatitis B virus (HBV) reveals an extraordinary structural disproportion. Alongside sparse 42-nanometer double-shelled infectious virions—known as Dane particles, each enclosing a partially double-stranded circular DNA genome and a reverse-transcriptase polymerase—the bloodstream swarms with $10^{11}$ to $10^{13}$ particles per milliliter of smaller, solid 22-nanometer spheres and tubular filaments. Outnumbering infectious virions by a factor of one thousand to ten thousand to one, these 22-nanometer subviral particles consist exclusively of host-derived endoplasmic reticulum lipids and the viral small surface envelope protein (HBsAg), containing zero nucleic acid and zero infectious capacity.

This viral excess was first detected in 1965 at the Institute for Cancer Research in Philadelphia by geneticist Baruch Blumberg and hematologist Harvey Alter. They identified an unfamiliar precipitin band in Ouchterlony double-immunodiffusion agarose gels when reacting serum from an Australian Aboriginal donor against serum from a multiply transfused hemophilia patient. Initially termed the Australia antigen, the protein was linked by 1967–1968 to post-transfusion serum hepatitis (HBV). A decade later, between 1975 and 1981, epidemiologist R. Palmer Beasley established the oncogenic stakes of chronic HBV infection through a prospective cohort study of 22,707 male civil servants in Taiwan. Whereas HBsAg-negative men rarely developed primary liver tumors, chronic HBsAg carriers suffered a relative risk of hepatocellular carcinoma (HCC) more than two hundred times higher. That risk was driven by decades of immune-mediated hepatocyte necroinflammation, cirrhosis, and insertional mutagenesis of viral DNA into the host genome.

Because Hepatitis B virus could not be propagated in cell culture and infected only humans and chimpanzees, conventional live-attenuated or whole-inactivated cell-culture vaccines were impossible. In 1969, Blumberg and microbiologist Irving Millman proposed harvesting the virus's own 22-nanometer non-infectious HBsAg decoys directly from the blood plasma of chronic human carriers to serve as a subunit vaccine. Turning human plasma from high-risk donors into a safe injectable required an uncompromising biophysical and chemical purification gauntlet, engineered at Merck by Maurice Hilleman and his team during the 1970s. Hilleman's process subjected pooled HBsAg-positive plasma to ammonium sulfate precipitation, isopycnic sodium bromide density-gradient ultracentrifugation, and rate-zonal sucrose centrifugation. Those steps separated the small $22\text{ nm}$ spheres (buoyant density $1.20\text{ g/cm}^3$) from the heavier $42\text{ nm}$ infectious Dane particles. To guarantee that no residual Dane particle, slow virus, or unknown blood-borne pathogen survived, Hilleman subjected the purified spheres to a sequential three-step chemical inactivation protocol: digestion with pepsin at pH 2.0 at $37^\circ\text{C}$ for eighteen hours, unfolding in 8 molar urea for four hours to denature retroviruses and prions, and cross-linking in 1:4,000 formalin for seventy-two hours. Licensed in 1981 as Heptavax-B, the plasma vaccine was free of infectious contaminants—including the newly emerging Human Immunodeficiency Virus (HIV)—and achieved over 90 percent protective efficacy.

Despite its safety, the plasma vaccine remained constrained by the finite supply of chronic carrier donors, a sixty-five-week manufacturing and chimpanzee testing cycle, and clinical hesitation over blood-derived products during the early HIV epidemic. To sever production from human blood altogether, biochemist William Rutter at the University of California, San Francisco, Pablo Valenzuela at Chiron Corporation, and geneticist Benjamin Hall at the University of Washington turned to recombinant DNA engineering. Early attempts in 1979 to express the HBV surface antigen (S) gene inside Escherichia coli failed to produce an effective immunogen. Bacterial cells could not perform eukaryotic lipid assembly or form the intermolecular disulfide bonds across the cysteine-rich "a" determinant loop (amino acids 124 to 147) of HBsAg, yielding misfolded monomeric chains.

In 1981, Rutter, Valenzuela, and Hall spliced the 678-base-pair viral S gene downstream of a strong yeast glycolytic promoter inside a shuttle plasmid and transformed it into baker's yeast, Saccharomyces cerevisiae. Fermented in stainless-steel vats, the eukaryotic yeast cells translated the 226-amino-acid hydrophobic polypeptide and inserted it into their endoplasmic reticulum membranes. There the protein oxidized its disulfide bridges into the native conformational "a" epitope and spontaneously budded uniform 22-nanometer lipoprotein spheres identical to those in human plasma. Because only the single surface gene was present in the yeast, infectious virus could never exist in the fermenter. Approved by the Food and Drug Administration in July 1986 as the first recombinant human vaccine—and deployed in Taiwan's universal newborn immunization program starting in July 1984—the HBsAg vaccine reduced pediatric chronic carriage in Taiwan from nearly 10 percent to under 0.6 percent and cut hepatocellular carcinoma incidence in vaccinated cohorts by 75 to 80 percent.