第14讲 Vaccines
HEADS UP!
The purpose of a vaccine is to "trick" the immune system into making memory B and T cells which can defend against a future attack by the real thing. The requirements for generating memory helper T and B cells are different from those for generating memory CTLs.
INTRODUCTION
During many "natural" infections, memory B and T cells are generated which can provide protection against a subsequent attack. However, a natural infection can be quite devastating – even lethal. If there was a safe way to trick the immune system into thinking it had been attacked, and to get it to produce memory B and T cells that are appropriate to defend against the anticipated attacker, then a person could be protected against a real infection.
That, of course, is what a vaccination does.
A vaccination is the immunological equivalent of the war games our armed forces use to prepare troops for combat. The goal of these "games" is to give soldiers as realistic a simulation of battle conditions as is possible without putting them in great danger. Likewise, a vaccination is intended to prepare the immune system for battle by giving the system as close a look at the real thing as is possible without exposing the vaccine recipient to undue risks. Indeed, the generals who plan war games and the scientists who develop vaccines have a common aim: maximum realism with minimum danger.
Vaccines have been extremely useful in controlling infectious diseases. For example, before a diphtheria vaccine was available, the number of new cases of diphtheria in the United States reached over 350,000 per year. Now, as a result of widespread vaccination against diphtheria, usually fewer than five cases are reported annually.
GENERATING MEMORY HELPER T AND B CELLS
When we are first exposed to a pathogen, dendritic cells at the battle site ingest the attacker or fragments of the attacker, and travel to nearby lymph nodes. There they use class II MHC molecules to present peptides derived from the invader's proteins. If a helper T cell has receptors which recognize these peptides, it can be triggered to proliferate. Eventually, some of these helper T cells become memory cells which can help protect against a subsequent attack. So for memory helper T cells to be generated, all that is required is for dendritic cells to collect "debris" from the battle scene (e.g., viral coat proteins or part of a bacterial cell membrane) and present peptides derived from this debris to helper T cells.
Likewise, when a B cell's receptors recognize an attacker or a fragment of an attacker which has been transported to the secondary lymphoid organs by the lymph or the blood, that B cell can be activated. After a period of proliferation, if T cell help is available, some of the resulting B cells will become memory cells. So as with helper T cells, even a bit of battle debris is enough to activate a B cell and generate memory B cells. The important point here is that memory B and helper T cells can be produced efficiently even when no immune system cells have been infected by the attacker.
GENERATING MEMORY KILLER T CELLS
Memory killer T cells also can be produced during a microbial attack, but for this to happen, the microbe must infect an antigen presenting cell. For example, if a virus infects a dendritic cell, that virus will commandeer the cell's bio-synthetic machinery, and use it to make viral proteins as part of its reproductive strategy. Some of these proteins will be chopped up into peptides and loaded onto class I MHC molecules. As a result, killer T cells whose receptors recognize the virus' peptides will be activated, and if assistance is available from helper T cells, memory killer T cells will be produced.
So the requirements for generating memory helper T and B cells are different from those for generating memory CTLs. Memory helper T cells and B cells can be produced even when an invader does not infect an antigen presenting cell. In contrast, for memory killer T cells to be made, the attacker must infect an antigen presenting cell.
Under certain experimental conditions, antigen presenting cells can use class I MHC molecules to present antigens taken up from outside the cell – antigens that normally would be presented by class II MHC molecules. This phenomenon is termed cross-presentation, and it might allow virus-specific CTLs to be generated even when the virus does not infect antigen presenting cells. Currently, it is not known how important cross-presentation actually is for the normal functioning of the human immune system. Indeed, no antiviral vaccine has been devised that uses cross-presentation to generate protective CTL memory in humans. Of course, it is possible that cross-presentation may eventually be used to produce such a vaccine. However, at this time, the rule seems to be that for a vaccine to efficiently generate memory CTLs, antigen presenting cells must be infected. In this lecture, we'll stick to that rule.
STRATEGIES FOR VACCINE DEVELOPMENT
A number of different approaches have been employed to develop the vaccines currently used to protect against microbial infections. In addition, innovative, new vaccine designs are being tested. One important feature of a vaccination is that its efficacy does not depend on the recipient altering his level of hygiene or his lifestyle. Consequently, many believe that a vaccine against HIV-1 – a virus that currently infects about 6,000 people per day – may be the best way to stop the spread of AIDS. Because this disease is such an important health issue, as we discuss different types of vaccines, we will ask whether any of them might be suitable to use as a vaccine that would protect against an HIV-1 infection. In the end, I think you will agree that designing a safe and effective AIDS vaccine is a difficult challenge.
One major obstacle to producing an AIDS vaccine is that it isn't certain which types of memory cells are needed.
The results of trials with vaccines that only produce memory B cells and antibodies have not been very impressive.
Moreover, individuals who are infected with HIV-1, but whose immune systems resist the virus for long periods of time, usually have inherited particular class I MHC molecules – suggesting that presentation of antigens to killer T cells is important for resistance. Consequently, most immunologists believe that an effective AIDS vaccine must generate memory killer T cells. Unfortunately, the production of memory CTLs requires that the agent used as a vaccine be capable of infecting antigen presenting cells – and this puts severe restrictions on the types of AIDS vaccines that might be safe to use.
Non-infectious vaccines
Many vaccines are designed not to infect the vaccine recipient. The Salk vaccine for polio is an example of such a "non-infectious" vaccine. To make his vaccine, Dr. Salk treated poliovirus with formaldehyde to "kill" the virus.
Formaldehyde acts by gluing proteins together, and the result of this treatment is that the virus looks to the immune system very much like a live poliovirus – but it cannot infect cells because its proteins are non-functional.
This treatment is the molecular equivalent of the parking police applying a "boot" to the wheel of a car. The car may look quite normal, but because the wheels can't turn, the vehicle is disabled. The common flu vaccine is also a killed virus vaccine, and a similar strategy has been used to make vaccines against disease-causing bacteria. For example, the typhoid vaccine is prepared from bacteria that have been grown in the lab and then treated with chemicals such as formaldehyde.
Although the chemicals used to kill these microbes certainly will incapacitate most of them, the procedure is not guaranteed to be 100% effective, and some of them may survive. Now if a vaccine is intended to protect against a virus like influenza, which otherwise will infect a large fraction of the population, the presence of a few live viruses in the vaccine preparation is not a major concern – because without vaccination, many more people would contract the disease. In contrast, if it is intended to protect against a virus such as HIV-1, in which infection is usually preventable (at least for adults in developed countries where blood supplies are carefully screened), a vaccine that has even a small probability of causing the disease could not be used to vaccinate the general public.
Some bacteria produce proteins called toxins that actually cause the symptoms associated with the bacterial infection. In a few cases, these toxins have been used as non-infectious vaccines. To prepare such a vaccine, the toxin is purified, and treated with aluminum salts to produce a weakened form of the toxin called a toxoid. When injected into a recipient, the toxoid mobilizes B cells that produce antibodies which can bind to and inactivate the harmful toxin during a real attack. Vaccines made from diphtheria or tetanus toxins are examples of this type of non-infectious vaccine.
Some non-infectious vaccines use only certain parts of a pathogen. The idea here is to retain the portions that the immune system needs to see for protection, while discarding the parts that cause unpleasant or dangerous side effects. An "acellular" vaccine for pertussis is made in this way. The original pertussis vaccine was prepared from whole, killed pertussis bacteria, and about half of the infants inoculated with that vaccine had an adverse reaction to it. Fortunately, almost all these side effects were mild when compared with the life-threatening possibility of contracting whooping cough. The acellular vaccine, which has a much lower rate of adverse reactions than the original pertussis vaccine, is made by growing the pertussis bacteria in culture and then purifying several of the bacterial proteins away from the rest of the bacterial components.
Viral proteins produced by genetic engineering also can be used as non-infectious, "subunit" vaccines. The highly effective vaccines against hepatitis B virus and the human papillomavirus are both made in this way. Because only one or a few "synthetic" viral proteins are used to make a subunit vaccine, there is no possibility that the vaccine will cause the infection it is designed to protect against.
A potential drawback of all non-infectious vaccines is that although they will generate memory helper T cells and B cells (which can make protective antibodies), memory killer T cells will not be made – because antigen presenting cells will not be infected. Of course, many pathogens (e.g., extracellular bacteria) do not infect human cells at all. Consequently, the lack of memory CTLs (which kill infected cells) is not an issue in designing vaccines to protect against these microbes. In addition, antibodies produced by memory B cells are sufficient to protect against some viruses which do infect human cells. Indeed, both poliovirus and hepatitis B viruses infect human cells. Nevertheless, the non-infectious Salk poliovirus vaccine and the hepatitis B virus subunit vaccine both work very well – even though neither vaccine generates memory killer T cells. So whether memory CTLs are required for protection depends on the particular microbe and its lifestyle.
Another disadvantage of non-infectious vaccines is that the protection they confer generally is not as long-lasting as the protection produced by vaccination with a live microbe. That's why, for example, the tetanus toxoid vaccine must be "boosted" about every ten years for it to be effective.
Attenuated vaccines
Another strategy for producing a vaccine is to use a weakened or "attenuated" form of the microbe. Virologists noticed that when a virus is grown in the laboratory in a cell type which is not its normal host, the virus sometimes accumulates mutations which weaken it. The Sabin polio vaccine, for instance, was made by growing poliovirus, which normally reproduces in human nerve cells, in monkey kidney cells. This strategy resulted in polioviruses which were still infectious, but which, in their weakened condition, could not cause the disease in healthy individuals. Most children in the United States receive attenuated virus vaccines for measles, rubella, and mumps.
Attenuated virus vaccines usually provide long-lasting immunity because they replicate to a limited extent in the host, thereby mimicking a natural infection.
An attenuated vaccine can be tested on animals to get a general idea of whether the attenuation procedure has worked. However, to be sure a crippled microbe can stimulate the production of memory cells, yet not cause disease, it must be tested on humans – usually volunteers who expect to be at risk for contracting the disease.
In this regard, it is interesting to note that by the time Dr. Sabin was ready to test his attenuated virus vaccine, most people in the United States had already received the Salk polio vaccine. So at the height of the Cold War, Dr. Sabin took his vaccine to Russia and tested it there.
Polio was such a dreaded disease that the Russians were delighted to be "guinea pigs" for Dr. Sabin's made-in-the-USA vaccine.
One important feature of attenuated virus vaccines is that they can produce memory killer T cells. This is because the crippled virus can infect antigen presenting cells and can stimulate the production of CTLs before the immune system has had a chance to destroy the weakened "invaders." However, because an attenuated vaccine contains a microbe that is infectious, there are safety issues. When a person has recently been vaccinated with an attenuated virus vaccine, he may produce enough virus to infect some of the people with whom he comes in contact. This can be an advantage if those people are healthy, because it "spreads the immunization around," producing what immunologists call herd immunity.
However, a person whose immune system is weakened (e.g., by chemotherapy for cancer) may not be able to subdue the attenuated virus. After all, the attenuated microbe in the vaccine isn't dead. It's just weak. So for those who are immunosuppressed, this type of gratuitous vaccination can have serious consequences.
A second potential safety concern with an attenuated virus vaccine is that before the recipient's immune system subdues the weakened virus, the virus may mutate, and these mutations may restore the strength of the virus. Although this is not a very likely scenario, some healthy people who received the Sabin vaccine contracted polio – because the weakened virus mutated and regained its ability to cause disease.
Carrier vaccines
Some newer vaccine preparations use genetic engineering to introduce a gene (or genes) from a pathogenic microbe into a virus that doesn't cause disease. This engineered virus can then be employed as a "Trojan Horse" to carry the gene of the pathogenic microbe into human cells. The idea here is that if the carrier infects the vaccine recipient's antigen presenting cells, these cells will produce the pathogenic microbe's protein in addition to the carrier's own proteins. As a result, inoculation with a carrier vaccine should generate memory killer T cells that can protect against a future attack by the real pathogen. Importantly, there is no chance that this vaccine will cause the disease it is designed to protect against – because only one or a few of the pathogen's many genes is "carried" by the vaccine.
It might seem that this approach would be perfect to use to prepare an AIDS vaccine, and vaccines of this type are being tested. Most recently, a vaccine trial in Thailand used a canarypox virus (a cousin of Jenner's cowpox virus) as a Trojan Horse to carry in several genes for HIV-1 proteins. This carrier virus vaccination was then "boosted" by vaccinating the same individuals with a subunit vaccine containing a synthetic version of one of the same HIV-1 proteins produced by the carrier virus.
The people receiving these vaccinations, plus roughly an equal number of individuals who received a placebo vaccination, were followed for a period of three years to determine how many in each group subsequently became infected with HIV-1 as a result of risky sexual behavior.
Although the authors claimed that the trial "showed a significant, though modest, reduction in the rate of HIV-1 infection," the data is not very convincing. During the study period, 56 people who received the authentic vaccine contracted HIV-1, whereas 76 members of the group which received the sham vaccine became infected with the virus. These are very small numbers on which to base a meaningful conclusion. Moreover, HIV-specific T cells could only be detected in about 17% of the people who received the vaccine. Finally, when the people who became infected were tested, there was no significant difference in the amount of virus in the blood of members of the two groups. This would suggest that the vaccination had little effect on the ability of infected individuals to resist the viral infection – not what you would expect from an effective vaccine.
WILL THERE BE AN AIDS VACCINE?
Most immunologists believe that to be effective, an AIDS vaccine must generate memory killer T cells. If this is true, non-infectious vaccines, which have been used to protect against many other pathogens, will be of little use against HIV-1. In principle, a weakened form of the AIDS virus could be used as a vaccine that would produce memory CTLs. However, because the AIDS virus has an extremely high mutation rate, there is great concern that an attenuated form of HIV-1 might mutate to become lethal again. Consequently, it is unlikely that a vaccine which uses an attenuated version of the AIDS virus could ever be used to vaccinate the general population. A carrier vaccine could generate memory killer T cells without putting the vaccine recipient at risk for a real AIDS virus infection. So far, however, this strategy has not yielded a vaccine which elicits a strong, protective immune response against HIV-1.
Even if a safe vaccine could be devised which would produce HIV-1-specific CTLs, the high mutation rate of the AIDS virus makes it an elusive target. On average, each AIDS virus produced by an infected cell differs from the original infecting virus by at least one mutation. Consequently, the body of someone infected with HIV-1 contains not just "the" AIDS virus, but a huge collection of slightly different HIV-1 strains. Moreover, if this person infects another person, that individual will usually not be infected by just a single AIDS virus, but rather by a whole "swarm" of different viral strains.
What this means is that the memory cells produced by a vaccination might protect very well against the particular strain of HIV-1 used to prepare the vaccine, yet be totally useless against the mutant versions of the virus which arise in a real infection. Indeed, the virus' ability to mutate rapidly may prove to be the most difficult problem of all to solve in making an effective AIDS vaccine.
Despite all these difficulties, immunologists are working hard to produce an AIDS vaccine that can be used to protect the public – because such a vaccine is viewed as the current best hope for controlling the spread of the AIDS virus. Recently, antibodies have been discovered in rare AIDS patients which can neutralize many different HIV-1 variants. If a vaccine could be made which would elicit these broadly neutralizing antibodies in healthy individuals, such a "universal" vaccine might be able to protect against infection – at least by many of the common HIV-1 strains. Unfortunately, experiments indicate that broadly neutralizing antibodies against HIV-1 usually arise years after the initial infection as the result of many rounds of somatic hypermutation. This finding raises the question of whether a vaccine can be invented which elicits broadly neutralizing antibodies without having to wait for extensive somatic hypermutation to take place.
And, of course, it may turn out that even broadly neutralizing antibodies are not enough, and that virus-specific CTLs really are required for protection against an HIV-1 infection.
It is important to note that HIV-1 is not the only microbe for which there is no effective vaccine. Roughly a million people die every year from malaria, yet there is no vaccine that has been shown to be generally protective against this disease. Likewise, immunologists have not been able to devise an effective vaccine against tuberculosis, a bacterial infection which kills about three million humans each year. And roughly one third of all the people on earth are infected with herpes simplex virus, yet a vaccine that protects against infection with this virus does not exist. Indeed, it is the hope of many that in trying to develop an AIDS vaccine, immunologists will discover new strategies that will make it possible to produce vaccines which will protect against some of the other pathogens for which vaccines currently are not available.
VACCINATION TO PREVENT VIRUS-ASSOCIATED CANCER
Vaccines can be used to prevent certain types of cancer. For example, a chronic infection with hepatitis B virus increases one's risk of getting liver cancer about 200-fold, and roughly 20% of long-term, hepatitis B carriers eventually develop this disease. Moreover, hepatitis B virus ranks as one of the most infectious of all viruses: Transfer of a fraction of a drop of blood is sufficient to spread the virus from one human to another. Fortunately, vaccines that protect against infection by hepatitis B virus have been available in the United States since 1982, and the current vaccine is administered not only to healthcare professionals, who routinely come into contact with blood and blood products, but also to children. This subunit vaccine gives the immune system a "preview" of a real hepatitis B infection, allowing ample time for memory B cells and the antibodies they produce to be mobilized. If infection does occur, the prepared immune system can quickly eradicate the virus, effectively preventing hepatitis B-associated liver cancer.
Infection with certain "oncogenic" types of the human papillomavirus (HPV) can increase the risk of cervical cancer. These viruses are spread by sexual contact, and there are now so many women infected with this virus that cervical carcinoma has become the second most common cancer in women worldwide, resulting in about 250,000 deaths per year.
Although there are about a dozen slightly different types of HPV associated with cervical cancer, two types, HPV-16 and HPV-18, are implicated in about 70% of all cervical cancer cases. Subunit vaccines made from proteins that make up the protective coats of the viruses are about 95% effective in preventing infection by both types of HPV. Interestingly, one of the vaccine formulations includes coat proteins from two other HPV types, HPV-6 and HPV-11. These two viruses are not associated with cervical cancer, but they do cause genital warts in both men and women. The thinking in including these two "extras" is that preventing genital warts might encourage boys and men to be vaccinated, since they might otherwise be reluctant to be vaccinated to prevent a disease (cervical cancer) they cannot get.
Recently, a vaccine has been developed which can protect against five additional types of HPV that are associated with cervical cancer. It is estimated that worldwide use of this new vaccine could prevent about 90% of all cervical cancer – providing that most sexually active young women could be vaccinated. Unfortunately, many of the cases of cervical cancer occur in less-developed parts of the world – where immunization via injection is problematic.
VACCINE ADJUVANTS
In order for a vaccine to mimic the invasion of a pathogenic microbe, the immune system must view the vaccine as both foreign and dangerous. This is not a problem for a vaccine which uses a crippled virus – because a crippled virus naturally provides both signals. However, for vaccines composed of only one or a few microbial proteins, providing the requisite danger signal can be a serious problem. Indeed, if a foreign protein is injected into a human, the immune system generally just ignores it – because it poses no danger.
Because of the requirement for a danger signal, it is common practice to combine vaccines with an adjuvant (derived from a Latin word meaning "help"). In fact, most of the vaccinations you have received probably contained aluminum hydroxide or "alum," which functions, at least in part, by providing that important danger signal. Other, more powerful adjuvants are now being approved for use. For example, the Cervarix vaccine, which can protect against infection by the human papillomavirus, uses MPL, a modified version of the bacterial surface protein LPS as an adjuvant. In this formulation, viral coat proteins provide the first signal-specific recognition of something foreign – and MPL alerts the immune system that there is danger associated with these viral proteins. Adding an adjuvant to a vaccine can greatly increase its potency, and can reduce the dose of vaccine which must be administered.
REVIEW
Vaccinations take advantage of the ability of B and T cells to remember invaders we have previously encountered. By introducing the immune system to a "safe" version of a microbe, vaccination prepares these adaptable weapons to respond rapidly and powerfully if a real attack occurs at some future time. The production of memory B and helper T cells does not require that an antigen presenting cell be infected. Consequently, non-infectious vaccines that elicit protective antibodies have been made from dead viruses or even a single viral protein. However, non-infectious vaccines do not produce memory killer T cells, and the protection conferred by non-infectious vaccines generally is not as long-lasting as the protection elicited by infectious vaccines.
Most immunologists believe that to protect against HIV-1, a vaccine will need to elicit memory killer T cells. To do this, a vaccine must be able to infect antigen presenting cells. Attenuated vaccines have been produced using a weakened version of a microbe that can still infect APCs, but cannot cause disease. However, a vaccine intended to protect the general population against HIV-1 must have no possibility of causing AIDS. And because HIV-1 has a very high mutation rate, there are no guarantees that an attenuated AIDS virus will not reactivate. Consequently, an attenuated form of the virus probably cannot be used to protect the public against an HIV-1 infection.
Another approach to making a vaccine that will elicit killer T cell memory is to insert one or more of a microbe's genes into the genome of a benign carrier. Then, when the carrier infects antigen presenting cells, the microbe's proteins will be produced. These proteins can be displayed by class I MHC molecules and can activate CTLs.
So far, however, this approach has not produced a generally useful AIDS vaccine. Several "anti-cancer" vaccines are now on the market. These subunit vaccines can reduce the risk of contracting either hepatitis B virus or the human papillomavirus. Infection with these viruses greatly increases the probability that a person will suffer from liver cancer (hepatitis B virus) or cancer of the uterine cervix (human papillomavirus). The potency of a vaccine can be increased by combining the specific antigen that a B or T cell recognizes together with an adjuvant. The purpose of an adjuvant is to "get the attention of the immune system" by providing a danger signal required for activation.