Technologies for Making New Vaccines

Classic Strategy for Viral Targets

Recombinant Technology-Based Strategies for Viral Targets

Specific modifications or deletions can be made in viral genes so that the virus is more stably attenuated. The increased stability of the attenuated phenotype results from modification(s) extensive enough that reversion through back-mutation is impossible or highly unlikely. In contrast, attenuated viruses derived by classic strategies may have only point mutations and therefore retain the capability to revert. In an attempt to generate a new vaccine, herpes simplex virus (HSV) was genetically engineered in three ways : (a) to induce attenuation, (b) to provide antibody markers of vaccination that differ from those of wild-type HSV infection, and (c) to protect against both HSV-1 and HSV-2. In another example, a particular HSV glycoprotein gene was deleted, resulting in an HSV strain that cannot replicate; this recombinant virus is produced in vitro in a cell line that supplies the deleted glycoprotein gene in trans , and the resultant virus can initiate infection in the host without further replication (termed replication-defective ).

Another molecular strategy for attenuation relies in deoptimized codons with the goal to generate a replication-defective strain. Polioviruses with deoptimized codons in the capsid region produce less infectious virus and are more highly attenuated, yet still efficiently elicit neutralizing antibodies that are protective in a mouse challenge model.

Recombinant technology also has been applied to develop live vaccines in which foreign polypeptide antigens from pathogens have been engineered into carriers. This strategy aims at presenting the foreign antigen to the immune system in the context of a live virus infection and stimulating the immune system to respond to the foreign polypeptide antigen as a live immunogen, with the expectation of developing a broader immunity (humoral and cellular). The foreign polypeptide is expressed within the infected cell and it is transported to the cell surface or broken down into peptides, which are transported to the cell surface where they may elicit cell-mediated immune responses. This strategy has also the potential advantage of amplifying of the immunogenic signal when the live vector initiates multiple rounds of replication (nonreplicating vectors are discussed in “Delivery Systems” below). If the vector virus is a commonly used vaccine, one could immunize simultaneously against the vector virus and another pathogen, ideally in a single dose. However, the preexisting or induced immune response to the live viral vector may limit the effectiveness of revaccination.

DNA Viruses

Vaccinia virus was the prototype viral vector in which dozens of different foreign polypeptides have been expressed. Immunizations of animals with this virus resulted in protection in a wide range of infections and oncology models. Recombinant vaccinia virus expressing HIV-1 gp160 has been tested clinically for prophylactic and therapeutic applications.

Given the known adverse events of vaccinia virus observed in the worldwide eradication program—which are more serious in immunocompromised persons—this virus has been engineered to reduce its virulence without compromising its efficacy as a live viral vector. ^, By passaging vaccinia virus several hundred times in chicken embryo fibroblasts, the highly attenuated Modified Vaccinia Ankara (MVA) was developed. This virus was safe in animals after experimental immune suppression, which led to its use in clinical trials with recombinant HIV vaccines, including MVA expressing monomeric gp120. ^, MVA has also been used to express antigen 85A from Mycobacterium tuberculosis which was well tolerated and immunogenic, but not efficacious, in infants and adults. ^, Two other members of the poxvirus family, fowlpox and canarypox viruses, were being developed as naturally attenuated live virus vectors that can infect human cells but not produce infectious viral progeny. Recombinant vaccinia and fowlpox viruses expressing carcinoembryonic antigen have shown clinical activity against tumors bearing this antigen.

RNA Viruses

RNA viruses can also be engineered into vectors for expressing foreign polypeptides. Lymphocytic choriomeningitis is one example of an RNA virus used as a vector. Sindbis and other alphaviruses have received extensive attention owing to their broad host range, ability to infect nondividing cells, and potential high-level expression per cell. Sindbis has been developed into a viral vector-DNA-based vaccine (see “Nucleic Acid-Based Vaccines” below). Vesicular stomatitis virus (VSV) vectors expressing filovirus proteins have been shown to protect against lethal challenge by Ebola or Marburg viruses, and alphavirus expressing cytomegalovirus (CMV) glycoprotein has been developed as a candidate CMV vaccine. The versatility of alphavirus vectors as a vaccine strategy is further manifested for noninfectious-disease targets, in that a vector encoding a grass pollen allergen is able to prevent allergic sensitization in a mouse model.

Given the Ebola outbreak that started in Africa in May 2014, global vaccine efforts were accelerated leading to the entry of two vector-based vaccine candidates into efficacy trials in humans. One is derived from a chimpanzee adenovirus engineered to express glycoproteins of ebolavirus, and the other is based on the VSV vector that has been engineered to express Ebola virus glycoproteins. Clinical trial results from a Phase III trial using the vaccine based on the VSV vector showed an efficacy of 100%. In 2019, the rVSV-ZEBOV vaccine (Erbevo) was licensed.

Another growing threat that has made the development of an effective vaccine an international priority is dengue disease. Vaccine development programs in both academic and industry settings have investigated various technologies, and the only successful dengue vaccine to date (reporting positive Phase III efficacy data) is a recombinant, live, attenuated, tetravalent dengue vaccine (CYD-TDV). This vaccine consists of four recombinant dengue vaccine viruses (CYD 1 to 4), each constructed by replacing the genes coding for the premembrane and envelope proteins of the yellow fever 17D vaccine virus with those from the wild-type dengue viruses, generating chimeric complementary DNAs that are individually transcribed to RNA, and then RNA transfection in Vero cells. CYD-TDV (Dengvaxia) was licensed in 2016 with 56–61% efficacy against virologically confirmed dengue. ^, In 2017, the vaccine was recommended to be used only in people who have previously had a dengue infection, as outcomes may be worsened in those who have not been previously infected. Recently, two other chimeric dengue vaccines entered phase III clinical trials. DENVax/TAK-003 is composed of a live attenuated DENV-2 and chimeric DENV-1,3,4 on DENV-2 backbone, and demonstrated an overall vaccine efficacy of 80.9% against symptomatic dengue in the Asia-Pacific and Latin America regions. Similarly, Butantan-DV/TV003 vaccine, composed of live attenuated DENV-1,2,3 and chimeric DENV-4 on DENV-2 backbone, reported seroconversion to all strains in over 70% of subjects in phase II trial, while a phase III trial is ongoing.

Classic Versus Recombinant Technology-Based Strategies for Bacterial Targets

It has been difficult to develop live attenuated bacterial vaccines by classic strategies because attenuation is difficult to achieve without reversion. Also, bacteria could stop expressing virulence factors in vitro , then turn on expression when in the host.

Serial Passage

Early in the 20th century, a strain of Mycobacterium bovis , also known as bacille Calmette-Guérin (BCG) was attenuated through 231 successive in vitro subcultures over 13 years. There are many strains of BCG vaccine available worldwide that are derived from this original strain. These vaccines vary in terms of tolerability, immunogenicity, and rate of protective efficacy for tuberculosis in clinical trials (range: 0–80% protection) for reasons that may relate to the actual vaccine strains or to differences in clinical trial designs or populations. BCG vaccines have been inoculated into billions of people worldwide, have a generally acceptable tolerability profiles, and have an important role in protection against childhood tuberculosis. BCG vaccines have also been administered orally. ^, Following the observation that patients with tuberculosis have reduced rates of cancer, BCG vaccine was developed into the first cancer vaccine, being indicated for the treatment of superficial bladder cancer by intravesicular instillation. BCG vaccine has also been used as an adjuvant for the priming doses of certain cancer vaccines and been shown to induce non-specific immune effects as outlined in Chapter 3 of this edition.

Mutagenesis

Chemical mutagenesis followed by selection was used to derive the attenuated Ty21a strain of Salmonella typhi. This vaccine was licensed for preventing typhoid fever based on its record of safety and efficacy (approximately 60% for several years) after a regimen of three or four doses.

Recombinant DNA (rDNA) technology can be applied to attenuate any new bacterial strain that holds promise to be a live attenuated vaccine. Therefore, it seems unlikely that a new live bacterial vaccine attenuated by a classic strategy alone will be developed in the future.

Recombinant Bacteria

The engineering of bacteria for attenuation is much more complex than for viruses because bacterial genomes are approximately 100-fold larger than those of viruses. The strategy for attenuation, like that for viruses, is to identify the gene(s) responsible for the virulence of bacteria or for their ability to colonize and survive in host tissue(s) and then eliminate the gene or abolish or modulate its in vivo expression. As is the case for modification of viruses, there must be a balance between the virulence of a bacterial strain and its activity as a vaccine, and it is possible to over-attenuate a bacterial strain to the point that it no longer replicates sufficiently in vivo and is no longer immunogenic. Vibrio cholerae strains have been developed as live cholera vaccines; attenuation has been accomplished by rDNA-directed deletion of genes encoding virulence factors (e.g., cholera toxin [CT]). To ensure attenuation by reducing the probability of reversion, deleting at least two independent genes or genetic loci that contribute to virulence is highly desirable. There have also been attempts to develop live Shigella vaccines through particular chromosomal or plasmid-based gene mutations to reduce pathogenicity.

Recombinant Bacterial Vectors

Suitable bacterial strains have been engineered into live vectors for expressing foreign polypeptides. The most common development has been to engineer enteric pathogens to induce mucosal immunity against the foreign polypeptide following oral delivery. S. typhi vectors have been the most commonly developed strains in terms of immunology, molecular design, and clinical testing. Strains of V. cholerae , Shigella flexneri , and Listeria monocytogenes have also been engineered. The challenges for these vectors are to retain sufficient virulence to replicate in the gut while achieving sufficient attenuation and expressing appropriate levels of foreign polypeptides. The BCG vaccine strain has also been engineered as a live vector to express foreign polypeptides. BCG offers additional protective effects via epigenetic reprogramming of innate immune cells and induction of cytokines, a phenomenon called “trained immunity”, which is associated with protection from nonmycobacterial infections such as yellow fever. The first clinical trials of a recombinant BCG vaccine (rBCG30) were initiated with BCG30 (Phase I), which is a recombinant BCG Tice strain derived from the Mycobacterium–Escherichia coli shuttle vector pSMT3 and expresses approximately 5-fold more M. tuberculosis 30-kDa protein than the parental BCG strain. rBCG30 was found to be well tolerated and more immunogenic than BCG. Recombinant BCG strain ΔureC expressing listeriolysin O pore forming protein (hly) is also in phase II–III clinical development (NCT03152903), after showing superior protection over parental BCG strain in mice. BCG has also been engineered to express recombinant Plasmodium falciparum CSp or Toxoplasma gondii ROP2 antigens, inducing specific immune responses following mice immunization. ^, One aspect of the versatility of BCG is its efficacy by different routes of administration (oral, intranasal, and intradermal), which allows the development of alternative vaccination strategies. ^,

Streptococcus gordonii , a gram-positive commensal (nonpathogenic) bacterium, has been engineered to express foreign polypeptides on its surface by genetic fusion to attachment sequences of S. gordonii M protein. One significant challenge is whether the foreign polypeptide will be sufficiently immunogenic, given that this bacterial species does not naturally elicit a strong immune response.

Bacteria that replicate intracellularly can be engineered to deliver plasmid DNA into host cells for expressing vaccine antigens. S. flexneri has been attenuated by a deletion mutant in the asd gene, an essential gene. While such a strain can be propagated in vitro in the presence of diaminopimelic acid and can invade cells (as long as it maintains a plasmid encoding invasion- associated polypeptides), it cannot replicate in vivo , where diaminopimelic acid is unavailable. A plasmid harboring a eukaryotic promoter and recombinant gene was transformed into this strain. The resultant recombinant S. flexneri strain was able to invade mammalian cells in vitro and to express the plasmid-encoded protein as a vaccine antigen. Because S. flexneri replicates in the intestine and stimulates mucosal immunity, this vector may be delivered orally to transfer DNA into host cells where mucosal immunity is stimulated. Other bacterial species that invade mammalian cells but not divide may also be able to deliver recombinant plasmids to different cell types for expressing recombinant proteins as vaccine antigens.

Dendritic Cells (DCs) as Autologous Vaccines

Dendritic cells can be exploited as potent professional antigen-presenting cells for tumor-specific antigens by isolating these from the patient and ex vivo transferring mixed antigens from a person’s tumor or alternatively with a recombinant immunostimulatory protein. The DCs process the antigens and re-express them on the DC surface. The cells then are reintroduced into the patient for antigen presentation and immune stimulation of T cells. DCs loaded with prostate-specific antigen genetically linked to the granulocyte-macrophage colony-stimulating factor cytokine have shown evidence for clinical efficacy in immunotherapy of prostate cancer and were licensed in 2010 (Sipuleucel-T). Over the last decade, new technologies provided improvement of DC therapy for prostate cancer, although no new product has been approved yet. The mixed-antigen approach has shown promise for glioblastoma. ^, Because autoimmune diseases are characterized by the loss of immunological tolerance to self-antigenic determinants, it may be possible to exploit DCs as antigen-presenting cells to achieve tolerance, as has been demonstrated in animal models of diabetes, multiple sclerosis, and autoimmune encephalitis. Clinical trials are being conducted to study DC-based treatments for diabetes, rheumatoid arthritis, multiple sclerosis and Crohn’s disease. Initial results of phase I trials indicated that DC therapy was safe and showed signs of clinical improvement.

Whole Pathogen

In the early 20th century, the only technically feasible approach was the use of the whole pathogen since there was essentially no knowledge of antigens and their role in immunity. The inactivation of whole bacteria or whole viruses had the objective of eliciting the formation of antibodies to many antigens, from which some antibodies presumably would neutralize the pathogen or its virulence determinants.

Whole Human Cells

Some of the designs for cancer vaccines are based on the use of individual tumor-specific antigens, either purified or expressed in a vector. A concern regarding using individual antigens in a vaccine is that the induced immune response following vaccination or administration of antibodies may stimulate antigenic modulation and diminished expression of the target antigen, loss of major histocompatibility complex class I expression or loss of other T-cell receptor-associated signaling, thereby reducing efficacy. The advantage of inactivated allogeneic whole-tumor-cell vaccines is the opportunity to present a wide array of tumor-specific antigens to the immune system. Tumor cell lines are expanded in vitro , pooled, and inactivated. A melanoma vaccine derived from two tumor cell lines and given with a synthetic adjuvant is licensed in Canada, but further development for the European and U.S. market was discontinued. In addition, autologous tumor cells modified with the hapten dinitrophenol and adjuvanted with BCG have shown evidence of clinical benefit for melanoma and were initially licensed in Australia but subsequently withdrawn and eventually discontinued because of financial constraints and subsequent decision to discontinue manufacturing. ^,

Tumor cells can also be genetically modified to express immunomodulators that may enhance antigen-specific antitumor immune responses. Irradiation of cells transduced with the recombinant vector renders them unable to proliferate, yet still able to secrete the immunomodulator. Autologous melanoma cells are transduced with a recombinant adenovirus expressing granulocyte-macrophage colony-stimulating factor; the cells are then irradiated and reinfused to patients with melanoma, resulting in an antitumor effect. Likewise, autologous plasma cells transduced with recombinant adenovirus expressing interleukin-2 are irradiated and reinfused to patients with multiple myeloma.

Viral Targets for Protein-Based Approaches

The first protein-based vaccines relied on natural sources of antigens. In this regard, the first hepatitis B vaccine is unique in using a human source (plasma) for the vaccine antigen. Liver cells of persons chronically infected with hepatitis B virus (HBV) shed excess hepatitis B surface antigen (HBsAg) into the blood. The HBsAg, a 22-nm lipoprotein particle antigen, has protective epitopes. To develop a vaccine, plasma was harvested from long-term chronic hepatitis B carriers, HBsAg purified, and the final preparation subjected to one to three inactivation techniques (depending on the manufacturer) to kill HBV and any other adventitious human agents possibly present in the starting plasma. This vaccine was well tolerated and highly efficacious, but it is no longer used as the vaccine can now be made by recombinant technology.

Bacterial Targets for Protein-Based Approaches

Proteins purified from cultures of B. pertussis have been combined to formulate acellular pertussis vaccines, which have replaced whole-cell pertussis vaccine for routine pediatric vaccinations in most developed countries. Depending on the number of different protein antigens, these licensed acellular pertussis vaccines are referred to as one-, two-, three-, four-, or five-component vaccines. These vaccines all contain pertussis toxoid, which when combined with other pertussis antigens, comprises the acellular pertussis vaccines. One preparation of the pertussis toxoid component is described in “Genetic Inactivation” below.

Tumor Cells

Proteins isolated from individual tumors have been used as cancer vaccines. In one application, individual tumors were isolated, the autologous gp96 heat-shock protein was purified from each tumor, and the heat-shock protein preparation was injected into the respective subject, resulting in stimulation of CTL responses.

Chemical Inactivation

Many bacteria produce protein toxins that are responsible for the pathogenesis of disease, including tetanus and diphtheria. The toxin molecules are purified from bacterial cultures ( Clostridium tetani [T], and Corynebacterium diphtheriae [D]), and then detoxified by incubation with a chemical such as formalin or glutaraldehyde. These detoxified toxins, referred to as toxoids, are included in the diphtheria, tetanus, and pertussis (DTP) combination vaccine. ^,

Genetic Inactivation

The chemical “toxoiding” procedure has the disadvantages of potential alteration of protective epitopes (with ensuing reduced immunogenicity) and the potential for reversion to a biologically active toxin. However, rDNA technology can be used to produce a stable toxoid. For example, pertussis toxin was mutated (in two locations to ensure stability) to reduce its enzymatic activity. The altered gene was substituted for the native gene in B. pertussis , resulting in an immunogenic but stably inactivated pertussis toxoid. This double-mutant pertussis toxoid (which also is treated with formalin under mild conditions to improve its immunogenicity and stability) is a component of one particular acellular pertussis vaccine, which is not in use clinically at this time, while other acellular pertussis vaccines use nonrecombinant pertussis toxoid and are described in more detail in Chapter 45 . A genetic approach to derive a diphtheria toxoid also was successful following the mutation of C. diphtheriae cultures and screening for enzymatically inactive yet antigenic toxin molecules. This genetic toxoid (CRM ₁₉₇ ) is the protein carrier for a licensed Haemophilus influenzae type b (Hib) and pneumococcal conjugate vaccine (see sections below). This technology also has been applied to V. cholerae toxin and ETEC toxin to produce candidate mucosal adjuvants (see later section).

Recombinant Polypeptides

The expression of recombinant polypeptides as subunit vaccines arguably has been the most extensively used and widely accepted application of rDNA technology for the development of new vaccines and these approaches are summarized.

Particle-Based Vaccines

The first application of rDNA technology to the production of a vaccine was for hepatitis B through the expression of the HBsAg gene in baker’s yeast Saccharomyces cerevisiae , giving rise to 22-nm HBsAg particles. In contrast, expression of the HBsAg gene in E. coli gave rise to HBsAg polypeptides but not HBsAg particles. The yeast-derived HBsAg virus-like particles (VLPs) were very similar in structure to HBV virions. The purified HBsAg is adjuvanted with aluminum salts. This vaccine supplanted the equally efficacious and well-tolerated plasma-derived vaccine that was initially approved by the U.S. Food and Drug Administration in 1981 and discontinued in 1990. With the goal of an edible vaccine, HBsAg also has been expressed in transgenic potato tubers and has been shown to be safe and immunogenic in clinical trials.

Particles are almost always more immunogenic than individual polypeptides ( Fig. 67.1 ). Furthermore, particles (including VLPs) usually elicit antibodies to conformational epitopes on the particle (and on the respective virus), whereas isolated surface polypeptides of the particle might not. Examples of such particle immunogens are hepatitis A virus virions (immunogenic in humans at dosage levels as low as 50 ng) and VLPs for human papillomavirus (HPV). The HPV virion is a highly ordered structure whose major protein is L1 and when expressed in eukaryotic cells, results in the formation of VLPs that elicit HPV-neutralizing antibodies. Bivalent, quadrivalent, and nonavalent HPV vaccines have been shown to be highly effective and have been licensed.