Aids and Beyond:
Defining the Rules for Viral Traffic

Stephen S. Morse

The lesson of AIDS demonstrates that infectious diseases are not a vestige of our premodern past; instead, like disease in general, they are the price we pay for living in the organic world. AIDS came at a time of increasing complacency about infectious diseases. The striking successes achieved with antibiotics, together with widespread application of vaccines for many previously feared viral diseases, made many physicians and the public believe that infectious diseases were retreating and would in time be fully conquered. Although this view was disputed by virologists and many specialists in infectious diseases, it had become a commonplace to suggest that infectious diseases were about to become a thing of the past and that chronic, noninfectious diseases should be our major priorities.^[1]

Donald A. Henderson, M.D., Dean Emeritus, School of Public Health, Johns Hopkins University (personal communication, 1989) recalls a speech at Johns Hopkins University in 1969 in which the surgeon general of the United States Public Health Service, expressing the optimism typical of this period, assured his audience that infectious diseases were now of marginal interest in the United States and that we should thus shift our focus of attention to the chronic diseases.

Rudely jolted back into an awareness of infectious diseases by AIDS, we now find ourselves in a period of great uncertainty, poised for the AIDS of the future. We cannot help but wonder what other catastrophes are waiting to pounce on us. In this essay I consider what we now know about the "AIDS of the future." In particular, I discuss the origins of "new" viruses and the question of whether their emergence can be anticipated and prevented.^[2]

Emerging viruses and viral evolution were the subject of a conference that I chaired in May 1989 ("Emerging Viruses: The Evolution of Viruses and Viral Diseases"), the first ever held on this subject. One purpose of the conference, which was sponsored by the National Institutes of Health with The Rockefeller University, was to unite historical, epidemiological, and molecular approaches. I am editing a book containing contributions by the participants. Several popular summaries of this conference have recently appeared; see, for example, Julie Ann Miller, "Diseases for Our Future," BioScience 39 (1989): 509-17. For a more technical summary, see Stephen S. Morse and Ann Schluederberg, "Emerging Viruses: The Evolution of Viruses and Viral Diseases," Journal of Infectious Diseases 162 (1990): 1-7.

I argue that AIDS and HIV are novel but that biological antecedents and parallels can be found in nature. The novelty of AIDS therefore probably reflects our imperfect knowledge of the natural world rather than a diabolical new development in viral evolution. It is of note, though, that the conditions favoring rapid dissemination of the virus were comparatively recent social developments of great importance. In essence,

---

they served as highways to expedite "viral traffic," from animal sources to humans and from small or isolated human populations to larger groups. This "viral traffic," as I call it, is central to the origin of most epidemics of viral disease. Most "new" or "emerging" viruses are the result of changes in traffic patterns, which give viruses new highways. Perhaps most important, human actions often precipitate viral emergence. Apart from such obvious human factors as the role of behavior in HIV transmission, many episodes of emergence have been the result of agricultural or environmental changes brought about by human intervention. We therefore bear greater responsibility for emergence, and may have greater ability to influence it, than has been supposed.

The emphasis placed by scientists and the public on the diversity of viruses—of which the stress on the novelty of HIV is one example—may have made us oblivious to these recurrent patterns and common features shared by many emerging viruses. Most "new" viruses are of zoonotic (animal) origin and are not really new; instead, they are existing viruses that have been given new opportunities or new settings.^[3]

This point has also been discussed in a very readable essay by Edwin D. Kilbourne, "Are New Diseases Really New?" Natural History 92 (December 1983): 28-32.

Viral evolution, while a fascinating phenomenon to scientists, has generally been less important per se as a mechanism of viral emergence than this transfer of existing or slightly modified viruses to new hosts. The optimistic message is that the possibly unpredictable path of viral evolution need not necessarily be fully charted before we can anticipate new diseases like AIDS. The central problem concerns the changing relationships between viruses and human society, reflecting changes in relationships between humans and their environment.

In this regard, focusing on the uniqueness of AIDS has tended to obscure the many features that this virus shares with other viruses.^[4]

See Howard M. Temin, "Is HIV Unique or Merely Different?" Journal of Acquired Immune Deficiency Syndromes 2 (1989): 1-9.

AIDS is unquestionably unusual, and its viral cause, human immunodeficiency virus (HIV), has many novel features. Nothing in our knowledge of viral disease prepared us for the unique features of AIDS. Much was known about interactions of viruses with the immune system, but a virus that caused human disease by depleting the cells responsible for specific immunity was unprecedented. AIDS was also one of the few documented examples of what appears to be a truly new virus entering the human population to cause a previously unknown disease (but see below). Most notably among its unusual properties, HIV has a predilection for T lymphocytes (and other cells) bearing the surface protein called CD4 (or T4). Various types of T lymphocytes are responsible for orchestrating and regulating all immune responses, as well as for carrying out certain types of immune functions known collectively as cell-mediated.

---

Their roles are determined by specific proteins on the cell surface, which serve as recognition markers. T lymphocytes bearing the CD4 protein, colloquially known as "T4 cells," are generally responsible for turning on and amplifying immune responses. Without these CD4⁺ T cells, the body is unable to mobilize an immune response, so that the host becomes vulnerable to the opportunistic infections that are the hallmark of AIDS.

As was shown several years ago, the predilection of HIV for CD4⁺ T cells is due to the fact that the CD4 protein is a receptor for the virus.^[5]

Angus G. Dalgleish et al., "The CD4 (T4) Antigen Is an Essential Component of the Receptor for the AIDS Retrovirus," Nature 312 (1984): 763-67; David Klatzmann et al., "T-Lymphocyte T4 Molecule Behaves as the Receptor for Human Retrovirus LAV," Nature 312 (1984): 767-70; P.J. Maddon et al., "The T4 Gene Encodes the AIDS Virus Receptor and Is Expressed in the Immune System and the Brain," Cell 47 (1986): 333-48.

That is, the virus enters T cells by attaching specifically to CD4 on the cell surface. Other viruses have specific receptors; what made HIV tragically unique was that its receptor was CD4 rather than some other protein on the cell surface. This allowed HIV access to the CD4⁺ T cell that is so crucial in the immune response. However, for reasons to be discussed, it seemed improbable that the property of infecting and killing CD4⁺ cells would be found in only one virus and not in any of its relatives.^[6]

It is hypothetically possible that a particular virus could be the only surviving member of an extinct group possessing a distinctive characteristic (in which case one might be forced to conclude that the characteristic would not have been of much survival value to the virus); however, the tendency in nature is usually the opposite: vestigial characteristics are often retained long past any apparent utility. Virtually nothing is known about viral "extinction," or even whether it occurs, except for the intentional case of smallpox. On the other hand, it is also possible, and actually not improbable, that a new characteristic could arise as a small change by mutation from an existing virus. Thus, CD4 tropism of HIV could have arisen by a fortuitous mutation in the env (viral envelope) protein required for attachment to the appropriate cell receptor for virus entry; such a mutation would enable the protein to attach to CD4 on cells. This was the view originally held by many people, including many virologists. Although HIV itself did not arise this way, for the very reasons discussed, one would suppose that an ancestor of HIV could have arisen in this fashion. Howard M. Temin has discussed the role of mutation; see "Is HIV Unique or Merely Different?" (referred to in note 4) and "Evolution of Cancer Genes as a Mutation-Driven Process," Cancer Research 48 (1988): 1697-1701. Alternatively, the virus could conceivably have acquired the capability for CD4 binding by picking up a host cellular gene for this property; retroviruses are well known for their propensity to exchange genetic information with host cells. It is hard to make any predictions about how important mutation is as a way of generating new viruses. My personal feeling is that, for statistical reasons, it is less important in human disease. Ours is only one of many mammalian species, and many other species are more numerous; if there is a finite probability of the critical step's happening in any particular species, the numerical chance of its happening first in humans is therefore compatatively small. However, such an event could someday happen, although perhaps at a lower frequency than would suit our own anthropocentrism.

Thus, the discovery of this mechanism led to a search, ultimately successful, for other examples of viruses that infect or kill CD4 T cells. HIV belongs to the Lentivirus subfamily of retroviruses. One might expect that relatives of HIV among other lentiviruses would behave similarly; indeed, Luc Montagnier has pointed out that most, if not all, primate lentiviruses have a predilection for CD4⁺ lymphocytes of their hosts.^[7]

Luc Montagnier, "Origin and Evolution of HIVs and Their Role in AIDS Pathogenesis," Journal of Acquired Immune Deficiency Syndromes 1 (1988): 517-20.

In addition to these, some lentiviruses of other species, such as the bovine and feline immunodeficiency viruses, also appear to attack similar targets in their respective species.^[8]

Matthew A. Gonda et al., "Characterization and Molecular Cloning of a Bovine Lentivirus Related to Human Immunodeficiency Virus," Nature 330 (1987): 388-91; Niels C. Pedersen et al., "Isolat on of a T-Lymphotropic Virus from Domestic Cats with an Immunodeficiency Like Syndrome," Science 235 (1987): 790-93. These viruses were all characterized after the discovery of the CD4 tropism of HIV as researchers became alerted to the possibility that related viruses with this property might exist. The example of mouse thymic virus, discussed below, was the first case of an unrelated virus shown to cause a similar effect.

But the ability to infect and kill CD4⁺ T lymphocytes may not even be unique to retroviruses. Herpesviruses are DNA-containing viruses unrelated to HIV. In my laboratory we have found that a mouse herpesvirus, mouse thymic virus (MTLV; murid herpesvirus 3), can specifically kill CD4⁺ T lymphocytes developing in the thymus of young mice.^[9]

Stephen S. Morse and Jay E. Valinsky, "Mouse Thymic Virus (MTLV): A Mammalian Herpesvirus Cytolytic for CD4⁺ (L3T4⁺) T Lymphocytes," Journal of Experimental Medicine 169 (1989): 591-96. The virus has been known since 1961. Rather than suggesting a common evolutionary relationship with HIV, which seems unlikely, this similarity probably indicates that unrelated organisms can evolve similar ways to go about a particular process. This apparent convergence is likely due to such limitations of the viral life-style as dependence on host cells.

T cells not possessing CD4 are not affected. Recent reports suggest that the recently described human herpesvirus 6 (HHV-6; also called HBLV) is probably T lymphotropic as well.^[10]

Dharam V. Ablashi et al., "HBLV (or HHV-6) in Human Cell Lines," Nature 329 (1987): 207; Carlos Lopez et al., "Characteristics of Human Herpesvirus-6," Journal of Infectious Diseases 157 (1988): 1271-73. HHV-6, although cytolytic for CD4⁺ T cells, probably does not enter the cell via a CD4 receptor.

In cell culture at least, HHV-6 can infect and kill cells bearing CD4.^[11]

Ablashi et al., "HBLV (or HHV-6) in Human Cell Lines."

Despite these worrisome properties, these viruses have never been associated with AIDS-like disease and are probably not responsible for any serious illnesses in mice or humans, although HHV-6 has been suggested as a possible cofactor for AIDS.^[12]

S. Z. Salahuddin et al., "Isolation of a New Virus, HBLV, in Patients with Lymphoproliferative Disorders," Science 234 (1986): 596-600.

The mouse virus does not appear to cause overt disease, even though individuals remain infected for life and chronically secrete virus, probably from T lymphocytes.^[13]

Wallace P. Rowe and Worth I. Capps, "A New Mouse Virus Causing Necrosis of the Thymus in Newborn Mice," Journal of Experimental Medicine 113 (1961): 831-44; Sue S. Cross et al., "Biology of Mouse Thymic Virus, a Herpesvirus of Mice, and the Antigenic Relationship to Mouse Cytomegalovirus," Infection and Immunity 26 (1979): 1186-95; Stephen S. Morse, "Mouse Thymic Necrosis Virus: A Novel Murine Lymphotropic Agent," Laboratory Animal Science 37 (1987): 717-25, and "Mouse Thymic Virus (MTLV; Murid Herpesvirus 3) Infection in Athymic Nude Mice: Evidence for a T Lymphocyte Requirement," Virology 163 (1988): 255-58. This does not rule out more subtle effects; we have recently found an association with autoimmune disease.

The human virus seems to cause roseola,

---

a mild childhood disease, and may be one of the commonest of all human viruses.^[14]

K. Yamanishi et al., "Identification of Human Herpesvirus-6 as a Causal Agent for Exanthem Subitum," Lancet 1 (1988): 1065-67; Lopez et al., "Characteristics of Human Herpesvirus-6."

I shall say more about HHV-6 later. We do not know why, unlike HIV, these apparently T lymphotropic infections rarely if ever cause severe disease and do not appear to result in AIDS-like syndromes. Although they might someday become the cause of new AIDS-like diseases, that is unlikely; in distinction to HIV, these viruses were probably in their respective host species for many generations and appear well adapted to their hosts.

The lesson from such findings is that infectious agents do not develop in a vacuum but are the result of an ongoing evolutionary process. Most life forms existing today evolved from organisms already in existence, and viruses appear to be no exception. Appearance of viruses de novo seems extremely rare, for the same reasons that other species rarely arise de novo . Thus, "new" viruses are likely to come from existing viruses, and, in general, viruses of today have antecedents and relatives. In a sense, viruses have "parents" just as we do. As Luc Montagnier put it, "We're boarding a train that's already in motion. New species aren't being created. We're seeing the old ones evolve."^[15]

Interview with Thomas Bass, Omni 11 (1988): 102-6, 128-34; remark quoted, p. 130.

It has taken us a long time to assimilate this lesson, and I am not sure that even now we fully grasp its implications. In the words of Joshua Lederberg, "the historiography of epidemic disease is one of the last refuges of the concept of special creationism."^[16]

Personal communication; see also note 2 above. For other comments by Lederberg on this subject, see Joshua Lederberg, "Medical Science, Infectious Disease, and the Unity of Humankind," Journal of the American Medical Association 260 (August 5, 1988): 684-85.

We still tend to think of each infectious agent as if it arose in a vacuum, and not as the result of an ongoing evolutionary process.

Viruses show great variety, and in addition many of the viruses of greatest concern mutate rapidly and unpredictably.^[17]

Esteban Domingo and John J. Holland, "High Error Rates, Population Equilibrium and Evolution of RNA Replication Systems," and Manfred Eigen and C. K. Biebricher, "Sequence Space and Quasispecies Distribution," both in RNA Genetics, ed. E. Domingo, J. J. Holland, and P. Ahlquist (Boca Raton, Fla.: CRC Press, 1988), 3: 3-36, 211-45; David Steinhauer and John J. Holland, "Rapid Evolution of RNA Viruses," Annual Reviews of Microbiology 41 (1987): 409-33. As these reviews show, HIV is highly variable but is not unique in this respect.

Because previously unrecognized viruses are involved, mechanisms of viral emergence must mirror the unpredictability of these mutations in the genotype. It was usually assumed that most emerging viruses had to arise through the evolution of a new variant, and the emphasis on variation in the viral genome may have engendered a widespread feeling that the significance of viral evolution is to generate unpredictable or unexpected new variants. As we cannot foretell the future, and thus cannot predict the future evolution of any organism, the problem of emerging viruses has always appeared insoluble because it seemed to require predicting the course of viral evolution—an impossible task.

The valuable implication of evolutionary theory for viral origins is that if "new" viruses must arise from closely related preexisting viruses, it is not really essential to answer the question in those terms. While variability has undoubtedly contributed to the success of many of the

---

most troublesome viruses, including influenza, HIV, and many others, the more germane question is how an existing virus that normally infects one host species would be able to cross over into humans to become a human disease problem.^[18]

This essay deals with emergence of human disease, but the principles apply equally to other species.

When restated this way, the seemingly insoluble problem of viral origins thus reduces to a more manageable (although not trivial) question of viral traffic, and attacking the problem includes better understanding and appreciating the viruses that already exist in nature, including some viruses not yet discovered. Even more usefully, however, by focusing attention on viral traffic, especially between species, this concept shifts attention to more approachable questions concerning conditions, or the "rules of the road" for viral traffic. What conditions, for example, on the part of the virus or of the host or in the environment, will permit a virus to infect people? Novelty will evolve—even new mechanisms of pathogenesis, as was the case with AIDS. But we may have some advance warning in nature if we know where to look (see also note 6). On the other hand, the factors leading a virus, perhaps as yet unseen in nature, toward emergence can be more more readily predicted and studied. In addition, some emerging viruses may already be in a human population, but they may be geographically isolated.

This explanation allows us to consider viral emergence as a process in two major steps. The first step is the advent of what may at first seem to be (or, rarely, actually is) a "new" agent and its initial introduction into the human population. Depending on the virus, this step could have occurred recently or long ago, or it may even have occurred repeatedly before a successful infection. I have made this one step rather than two, because, for the reasons discussed above, a "new" agent is just as likely to be an "old" agent of another species. The virus may perhaps sometimes be slightly altered, although that is usually not necessary. The second step, dissemination in the human population, occurs once a virus infects its first human being. This model, then, presumes that emergence is simply a matter of a virus's getting into the human population and then spreading within the population. Many viruses may never achieve this second step. Although this simplification covers a multitude of sins, it provides a conceptual framework with which to begin. One consequence of this model is that—without requiring detailed advance knowledge of the virus—it permits us to analyze emergence by considering what contributes to each of these steps and what conditions could affect each step, the "traffic laws."

For the first step, even apparently new viruses, such as HIV, have

---

usually left tracks; often we have just failed to spot them in time. The conceptual problem of viral evolution is also in the first step. But because the requirements of evolution constrain novelty somewhat, there are only a few ways a "new" virus can arise. It can be a truly new virus, a major evolutionary variant, arising by genetic processes such as (for example) mutation or recombination; it can be an existing virus of another species, introduced virtually unchanged or with minor variations into humans from the other species; or it may be an existing human virus of limited scope. The "truly" new virus or major variant is possible but, for the reasons discussed above, is likely to be a rare event. While it is unlikely that we can predict its occurrence, we fortunately will not often be required to do so.^[19]

Howard Temin provides evidence that: number of neutral mutations can accumulate and lead in time to a virus possessing drastically different properties from its "parents." Therefore, he argues, it is impossible to predict what new variants will emerge. For an exposition of this argument, see his papers cited in note 6. On the other hand, other factors, such as the limited number of routes by which a virus can enter the body, impose certain constraints. It appears likely that the rules I have described here for the first step are the most important factors in the short run; on a longer time scale, over hundreds or thousands of years, genuine evolutionary change might occasionally be significant.

Several different factors can influence this first step profoundly. Many of the important changes responsible for new viral traffic are made by humans. Just as with other kinds of traffic, viral "traffic" has its traffic indicators, stop-and-go signals, and rules of the road. For example, certain types of environmental changes may be "go" signals for viral traffic. They act by increasing chances or frequency of introduction, or by favoring spread of a natural host or carrier (vector) for a virus. Deforestation and agricultural practices are among the factors most often responsible. To illustrate, I will sketch several instances of viral emergence. Although the examples may appear exotic at first, they will eventually come closer to home.^[20]

Because of their number, I have not provided specific references for most of the examples mentioned. For some references, see also Morse and Schluederberg, "Emerging Viruses." Many examples were discussed at the May 1989 conference on emerging viruses, and a forthcoming volume (see note 2) on emerging viruses will describe many of them in greater detail. I am grateful to Karl M. Johnson for providing much of the information on the hemorrhagic fever viruses; to Drs. Robert G. Webster, Peter Palese, and Edwin D. Kilbourne for information on influenza; and to Dr. Thomas P. Monath for information on the arthropod-borne viruses. For the reader desiring additional detailed scientific information, at a more advanced level, and references, the following can be recommended: For background on viruses, viral diseases, and immunology, a general textbook of medical microbiology, such as Bernard D. Davis et al., Microbiology, 4th ed. (New York: Lippincott 1990), can be consulted. For a general treatment of infectious organisms and their hosts and of viruses as the causes of disease (principles of pathogenesis), see Cedric A. Mims, The Pathogenesis of Infectious Diseases, 3rd ed. (New York and London: Academic Press, 1987); see also Sir Macfarlane Burnet and David O. White, Natural History of Infectious Disease, 4th ed. (London and New York: Cambridge University Press, 1972), which has become a classic. Finally, for specific viruses, the most detailed reference is Bernard Fields, ed., Virology, 2nd ed. (New York: Raven Press, 1990).

Argentine hemorrhagic fever is caused by Junin virus. It emerged from obscurity to cause about 400–600 cases annually over an area of 100,000 square kilometers (up from the original 16,000 square kilometers of 1958). The emergence of Argentine hemorrhagic fever was precipitated by agricultural changes as people cleared the pampas for agriculture and began to plant maize. A natural host for this virus is a mouse, Calomys musculinus; infected individuals of this species chronically shed virus in their urine. Although this rodent was always in the Argentine pampas, it began to flourish when natural grassland was cleared and maize was planted, so that ultimately it outnumbered the other rodents. Studies show an enormous difference in numbers of this mouse in cornfields, as opposed to natural grasslands,^[21]

Gloria de Villafañe et al., "Dinámica de las comunidades de roedores en agro-ecosistemas pampásicos," Medicina (Buenos Aires), 37, Suppl. 3 (1977): 128-40.

and the first recognition of Argentine hemorrhagic fever (1953) corresponds to increased corn planting in the region. Additional data corroborate the association. The rodent population fluctuates in a three- to five-year cycle, as does the incidence of Argentine hemorrhagic fever cases, and percentage of infected mice is higher in areas with many human Argentine hemorrhagic

---

fever cases. Bolivian hemorrhagic fever is caused by a related virus (Machupo virus) with a similar story; here the rodent is Calomys callosus . For various economic reasons, agriculture, primarily cattle raising, increased in the affected areas of Bolivia over the past thirty years. Calomys callosus adapted well to the new conditions, with the result that more people came in contact with the virus carried by this rodent. Increasing agriculture caused increasing cases. In the 1960s about 1,000 cases were reported, with 20 percent mortality. A program of rodent control, trapping and killing infected mice, has been very effective; as a result, there have been no new cases since 1974. This decrease further indicates that the putative association of rodent and disease was correct. An Old World relative of these two viruses, the notorious Lassa fever of Africa, follows an almost identical pattern.^[22]

The recognition of Lassa fever, and its astonishing mortality, was the subject of a popular book about fifteen years ago, Fever! The Hunt for a New Killer Virus, by John Fuller (New York: Reader's Digest Press, 1974). A recent incident was reported in the press (Lawrence K. Altman, "When an Exotic Virus Strikes: A Deadly Case of Lassa Fever," New York Times, February 28, 1989, p. C3).

The major natural host of this virus is another mouse, Mastomys natalensis , which adapts readily to humans, thriving on the food people leave and sharing human habitation. It unwittingly sheds Lassa fever virus, and humans become infected by contact.

The unrelated Korean hemorrhagic fever (Hantaan) falls into the same pattern. The natural host is Apodemus , basically a field rodent, and people come in contact with infected animals during rice harvesting. Increased rice planting has provided food for Apodemus as well as for people, and prevalence of Korean hemorrhagic fever has increased accordingly.

Not all of these viruses originate in rodents, although a remarkable number do. A number of important disease-causing viruses are also transmitted by arthropod vectors, such as insects or ticks. Most of these are viruses that can infect both mammals and the arthropod vector, a rather rigorous requirement that bespeaks evolutionary intimacy, on the part of the virus, with both invertebrates and vertebrates. They cause diseases that usually have long histories, and the arthropod vectors (really, arthropod hosts) serve primarily to disseminate a virus or to transport it into new individuals from a natural zoonotic (animal) source. Factors encouraging the arthropod vector can be important in disease emergence, as is demonstrated by several arthropod-transmitted diseases that have emerged recently. Rift Valley fever, found in Africa, caused serious outbreaks in Egypt in 1977 and more recently in Mauritania; the infection is characterized by a fever, usually with hemorrhaging; is naturally transmitted by various mosquitoes; and normally infects ungulates, such as sheep. Because the larvae of most mosquitoes involved in virus transmission develop in water, the addition of large open sources of water

---

often increases the mosquito population and has a major impact on transmission. In Egypt, although it is not known for certain why the virus emerged, the Aswan high dam was a possible factor precipitating emergence. The factors were more clearly defined in the Mauritanian Rift Valley fever outbreaks. Here the human cases occurred near areas along the Senegal River where large dams (for hydroelectric power) had recently been constructed. Similarly, sporadic outbreaks in other parts of Africa were usually associated with unexpectedly heavy rains.

There have been several incidents of Oropouche fever, first seen in Trinidad around 1957, in the Amazon region of Brazil. Appearance of the disease coincided with the introduction of cacao as a cash crop to the Amazon region. The vector, a biting Culicoides midge, breeds well in empty cacao hulls discarded after harvesting. The virus is also widespread in Panama, where a number of cases have been reported since 1989, and in Venezuela, where a notable outbreak occurred in early 1990. I will have a few words shortly about Lyme disease, not viral but also arthropod borne.

In all these cases, expanding agriculture, often accompanied by deforestation, played a major role in precipitating emergence—that is, in introducing a zoonotic (animal) virus into a new population. The role of agriculture seems logical on consideration. After all, if many "new" viruses are zoonotic, how would people come in contact with animal species bearing unfamiliar zoonotic viruses? Agricultural practices, as well as increased human habitation, may change the ecology of an area to allow a previously minor species to proliferate, as in the cases above. Expanding human habitation in a region, which may include or be the result of clearing land for agriculture, may also put people in direct contact with new animal species (and thus their viruses), as in the example of monkeypox. Monkeypox is an African virus that is related to smallpox but causes a milder form of illness upon infecting humans. It has often been named as a possible successor to smallpox following the recent eradication of the human virus. Monkeypox is so called because it was first identified in infected monkeys, but it is actually a virus of rodents, especially squirrels. People become infected when they develop settlements at the edge of the rainforest and, encroaching on the forest, come in contact with infected rodents inhabiting the forest.^[23]

Z. Jezek and Frank Fenner, Human Monkeypox, Monographs in Virology, vol. 17 (Basel: S. Karger, 1988). Fortunately, monkeypox has only a limited ability to spread from person to person and therefore is probably not a major threat, at least in its present form. Frank Fenner considers human monkeypox a transient phenomenon in areas undergoing transition from forest to cleared land. He points out that risk of exposure to monkeypox increases as people begin to encroach on the forest but that the risk decreases considerably after deforestation is largely complete in an area.

More remarkably, the same principles also often apply to viruses whose emergence can clearly be ascribed to viral evolution. Influenza A virus is one of the few known examples (aside from some arguable cases, such as HIV, it may be the only example) of such a virus. Every twenty

---

years or so, influenza A undergoes a major antigenic shift in one key protein, known as the hemagglutinin (H) protein, and a pandemic results.^[24]

Robert G. Webster and R. Rott, "Influenza Virus A Pathogenicity: The Pivotal Role of Hemagglutinin," Cell 50 (1987): 665-66. Robert Webster (personal communication) calls influenza the oldest emerging virus that is still emerging.

Although most changes in influenza virus H proteins occur by so-called antigenic drift, involving the accumulation of random mutations (this drift can lead to the smaller influenza epidemics seen every few years), new pandemic influenza viruses arise by a different route, that of major antigenic shifts. These invariably involve a reassortment of viral genes carried by different influenza strains. Thus, the important event in generating new pandemic influenza strains has, oddly, been not mutational evolution but a reshuffling of existing genes. Where do the genes come from? It has recently been found that most influenza genes are maintained in wildfowl; every known subtype of the H protein can be found in waterfowl. A number of virologists believe that pigs are an important "mixing vessel," allowing influenza virus to make a transition from birds to humans.^[25]

H. Kida, K. F. Shortridge, and R. G. Webster, "Origin of the Hemagglutinin Gene of H3N2 Influenza Viruses from Pigs in China," Virology 162 (1988): 160-66. In contrast, Chinese scientists believe that the pig is the recipient (getting virus from people) rather than the donor; see Zhu Ji-ming, "Human Virus Diseases in China: Research and Control," Impact of Science on Society 150 (1988): 137-47.

Every major flu epidemic known has originated in south China, which has also long practiced a traditional and unique form of integrated pig-duck farming.^[26]

Christoph Scholtissek and Ernest Naylor, "Fish Farming and Influenza Pandemics," Nature 331 (1988): 215. For further information on the traditional Chinese pig-duck agriculture systems, see K. Ruddle and G. Zhong, Integrated Agriculture-Aquaculture in South China: The Dike-Pond System of the Zhujiang Delta (New York: Cambridge University Press, 1988). I thank Wallace Parham, U.S. Congress Office of Technology Assessment, for valuable information on this subject.

Agriculture may play the leading role in emergence of this virus as well. Here, too, viral traffic—reassortant viruses from the mixing of animal influenza strains and the transmission of the resulting virus to humans—is more important than viral evolution for human disease.

Human immunodeficiency virus is a more difficult case. Where did HIV come from? We do not know the origin of HIV, but a probable primate origin is often suggested and appears highly plausible, at least for HIV-2. The origin of HIV-1 is more problematic. The existence of animal lentiviruses with a predilection for CD4⁺ T lymphocytes strongly suggests the possibility of a zoonotic origin for HIV at some time in the past. What is currently unknown is how and when the virus was first introduced into humans.^[27]

It is also unknown how many times the virus may have been previously (but unsuccessfully) introduced to humans. A great deal has been written on the origins of HIV. For an excellent discussion of several aspects of this question, see Mirko D. Grmek, Histoire du SIDA (Paris: Payot, 1989; English trans., History of AIDS [Princeton, N.J.: Princeton University Press, 1990]). A chapter by Gerald Myers in my forthcoming volume on emerging viruses (note 2) will consider the origin and spread of HIV. For an earlier discussion of these views, see T. F. Smith et al., "The Phylogenetic History of Immunodeficiency Viruses," Nature 333 (1988): 573-75 (and accompanying "News and Views" commentary by David Penney, "Origins of the AIDS Virus," 494-95). See also S. Conner and S. Kingman, The Search for the Virus (London: Penguin Books, 1988); and Montagnier, "Origin and Evolution of HIVs." An alternative hypothesis, that HIV-1 is an ancient virus in humans, has been suggested, notably by Montagnier.

The principles involved in these examples apply to all types of infectious diseases in all parts of the world. For example, it is hard to imagine a part of the world more heavily populated and thoroughly explored than the northeastern United States, but Lyme disease, the media's star disease of the 1989 and 1990 summer seasons, follows the same principles. Although bacterial rather than viral, Lyme disease is also zoonotic, being naturally found in several other mammals and probably originating in wild mice, and is transmitted by a tick. It is not clear why Lyme disease has recently emerged, but conditions favoring increased contact of people and infected tick vectors are likely to be principal reasons. These conditions appear to include changes in forestland around

---

houses. Malaria, a major cause of death worldwide, is caused by a protozoan parasite and not a virus. It is so widespread that one can hardly consider it emerging. But the recent completion of a new highway (SR 364) through the Amazonian rainforest of Brazil resulted in a massive increase in malaria cases in the region.^[28]

Thomas Lovejoy, personal communication, May 1989; Richard House, "Malaria Spreads in Brazil as Development Opens Up the Amazon," Washington Post, July 18, 1989, p. 5.

The second step, dissemination within the human population, is obviously crucial. Not only for newly introduced viruses, but also for many viruses long established in humans, emergence is the result of increased or accelerated dissemination. For this step, there is a vast epidemiological literature, which it is beyond the scope of this essay to review. Instead, I will mention a few recent developments, extending the metaphor of viral traffic by adding traffic in the more familiar sense. Modern transportation offers rich possibilities for rapid dissemination of new or exotic viruses. Recently a man who contracted Lassa fever while visiting Africa became sick after returning to the United States.^[29]

Altman, "When an Exotic Virus Strikes." This example demonstrates how comparatively easy it is for a disease to be spread rapidly by travel.

As another example, HIV undoubtedly traveled along the Mombasa-Kinshasa highway and came to the United States presumably through travel.^[30]

I thank Gerald Myers for this example. See also Conner and Kingman, The Search for the Virus, pp. 212ff, and Peter Piot et al., "AIDS: An International Perspective," Science 239 (1988): 573-79.

I have already mentioned malaria and the Brazilian highway. Of course, this is hardly a new phenomenon in infectious diseases, as witness the classic example of bubonic plague. In this vein, the dissemination of dengue and yellow fever, both transmitted by the same species of mosquito, is a particularly instructive example. The viruses and the mosquitoes were both probably spread by the African slave trade. It has been suggested that the mosquitoes that spread these diseases were inadvertently carried to the New World in the large open water containers on slave vessels. The mosquitoes lay their eggs in water, where the larvae hatch and develop; availability of water is therefore a major factor in population growth for many mosquito species. Plus ça change: A new, and more aggressive, mosquito, Aedes albopictus (the Asian tiger mosquito), was recently found in the United States and is now established in seventeen states. An effective vector for dengue and several other mosquito-borne viruses, the mosquito was introduced into the United States in 1985 in containers of used tires imported into Houston, Texas, from Asia. Wet tires are known as excellent breeding grounds for several species of mosquitoes, and have been shown to harbor many more tiger mosquito larvae than dry tires. Thus, carriers of disease are still themselves carried, however inadvertently, in commerce.

Human population movements are of obvious importance for disseminating viruses. Migration to cities from remote areas may pose a particular challenge. People in a remote area may come in contact with

---

an isolated virus, as in the examples of monkeypox and Lassa fever. If they move to a city, an increasingly common event, they bring their diseases with them. The population growth strains the city's infrastructure and can cause serious problems, as shown by the impressive expansion of dengue virus (a mosquito-borne infection). In many tropical cities open water storage is used increasingly as the city enlarges; as a result, additional breeding grounds are provided for mosquito vectors. At the same time, the high density of these urban areas places infected people and susceptible people in close contact, so that a cycle of infection is established.

Public health measures—such as mosquito control programs, health certification of travelers, and health inspection of imported livestock—have traditionally been directed to combating this stage, which has generally been the most vulnerable to attack. These programs have been instrumental in containing many potential threats, but they also have several drawbacks. Their success with the targeted diseases depends on vigilance and assiduity. Sadly, even when adequate weapons to combat disease are available, we may fail to use them effectively. As a case in point, the recent resurgence of measles in some U.S. cities seems largely due to the failure to ensure that all children are adequately immunized early in life. Efforts may fall victim to their own success, being prematurely relaxed or abandoned, usually to save money; as a result, the conditions that precipitated the program in the first place may reestablish themselves. Many mosquito control programs have met with this fate after initial partial success. These programs are reactive and can generally succeed only with known diseases, although some programs may confer broader benefits. Most of these programs also cannot contain viruses that can spread efficiently from person to person, such as influenza. Present strategy with influenza is to attempt to track emerging new strains and to immunize when feasible.

The modern world also offers additional gateways for viral traffic. For example, as has tragically been demonstrated with HIV, such medical procedures as blood transfusion and tissue transplantation offer the donor's viruses direct access to new hosts. Since many viruses, including HIV, are not able to spread efficiently from person to person, these procedures circumvent the lack of effective means of transmission. As these lifesaving procedures become more widely used, and as the scarcity of donors forces medical centers to look farther afield, it is reasonable to expect more instances. Agriculture again provides an interesting analogy. Viroids, small pieces of genetic information that lack the protein

---

coat normally needed by viruses to infect host cells, are spread, as far as we know, entirely by mechanical transmission on agricultural implements such as pruning knives and harvesters. It is speculation, but the evolution of viroids could very likely have been shaped, unbeknownst to its human agents, by these human activities.

How do we assess the nature of a viral threat? Before we address this question, let us put aside one class of emerging viruses that occasionally make the news (and rightfully so) but that probably would not represent a major threat. These are viruses that have only recently been identified because of advances in diagnostic technology, but have probably been with us a long time. Two recent examples are non-A, non-B hepatitis and human herpesvirus 6, both viruses that have been discovered within the last three years. In the case of non-A, non-B hepatitis, there is considerable evidence that the virus had been a major cause of post-transfusion hepatitis for years, but the virus itself remained elusive. The application of molecular technology, using DNA cloning, finally made it possible to identify the virus.^[31]

Qui-Lim Choo et al., "Isolation of a cDNA Clone Derived from a Blood-Borne Non-A, Non-B Viral Hepatitis Genome," Science 244 (1989): 359-62.

I have already mentioned human herpesvirus 6 (HHV-6).^[32]

Another T lymphotropic human herpesvirus, dubbed "human herpesvirus 7" (HHV-7), has recently been described (N. Frenkel et al., "Isolation of a New Herpesvirus from Human CD4⁺ T Cells," Proceedings of the National Academy of Sciences (USA) 87 (1990): 748-52).

HHV-6 was originally reported from Robert Gallo's laboratory under the name "human B lymphotropic virus" (HBLV); it was discovered fortuitously when it interfered with the growth of HIV isolates in tissue culture. At first thought rare, HHV-6 was later associated with the very common childhood disease called roseola. Since roseola has been known for many years,^[33]

J. Zahorsky, "Roseola Infantilis," Pediatrics 22 (1910): 60.

it is likely that HHV-6 has been a ubiquitous virus for decades, probably centuries. Many known diseases in search of causes can be placed in this category. There are likely to be many surprises here, but few threats, because the viruses are already widely disseminated. On the other hand, the importance of technological advances in making these discoveries cannot be overemphasized. These viruses became apparent because the means were developed to demonstrate their existence.^[34]

A few words on the importance of technology in disease recognition may be appropriate here. The recognition of HIV was dependent on the previous development of methods for growing T lymphocytes in culture, including key methods that were developed in Gallo's laboratory. For an excellent discussion of the history of HIV and of the role of technology in the discovery of HIV, see Grmek, Historie du SIDA. In a more general sense, the introduction of tissue culture methods, in the 1940s, was a major breakthrough in the study and characterization of viruses. The identification of non-A, non-B hepatitis virus (now called hepatitis C virus), described above, is another example of the successful application of technology. It can be expected that new tools for detection will uncover new viruses. In particular, many new avenues are opened by the recent development of an exceptionally sensitive technique—the polymerase chain reaction, or PCR (Randall K. Saiki et al., "Primer-Directed Enzymatic Amplification of DNA with a Thermostable DNA Polymerase," Science 239 (1988): 487-91; Chin-Yih Ou et al., "DNA Amplification for Direct Detection of HIV-1 in DNA of Peripheral Blood Mononuclear Cells," Science 239 (1988): 295-97). PCR is capable of detecting one HIV-infected cell in a hundred thousand (Richard A. Gibbs and Jeffrey S. Chamberlain, "The Polymerase Chain Reaction: A Meeting Report," Genes and Development 3 (1989): 1095-98). Because PCR can detect and amplify DNA in minuscule amounts of sample, and is comparatively undemanding, it is rapidly finding favor in many applications. PCR has great potential for disease archaeology and the study of evolution. One difficulty with conventional virological methods is that they often require a sample that has been carefully handled. It is often difficult to detect viable virus, for example, in fixed tissues or in samples that have been stored carelessly or for long periods. By PCR many otherwise intractable samples can now be tested, even mummified human bodies 7,000 years old (Svante Pääbo, "Ancient DNA: Extraction, Characterization, Molecular Cloning, and Enzymatic Amplification," Proceedings of the National Academy of Sciences (USA) 86 (1989): 1937; S. Pääbo, R. G. Higuchi, and Allan C. Wilson, "Ancient DNA and the Polymerase Chain Reaction: The Emerging Field of Molecular Archaeology," Journal of Biological Chemistry 264 (1989): 9709). This technique also has possibilities for detecting viral genetic information in ancient samples. There have been many speculations about the antiquity of HIV and AIDS. Mirko Grmek (personal communication, Paris, July 1989) has suggested using PCR to test for HIV or HIV-like viruses in century-old tissues preserved in pathological museums, and this would be quite a feasible way to determine whether HIV infection might have existed then. Even older samples can be tested for specific viruses. PCR techniques are available now for detecting entire families of viruses based on limited genetic resemblances and offer powerful tools for studying viral "paleontology" and evolution (see, for example, David H. Mack and John J. Sninsky, "A Sensitive Method for the Identification of Uncharacterized Viruses Related to Known Virus Groups: Hepadnavirus Model System," Proceedings of the National Academy of Sciences (USA) 85 (1988): 6977-81; Andy Shih, Ravi Misra, and Mark G. Rush, "Detection of Multiple, Novel Reverse Transcriptase Coding Sequences in Human Nucleic Acids: Relation to Primate Retroviruses," Journal of Virology 63 (1989): 64-75. As a result, one can now "go fishing" for viral ancestors and relatives in formerly untestable samples. PCR makes multiple copies of the specific piece of DNA it detects, and the resulting product can be analyzed and compared with known viruses. This capability is a boon for the study of viral evolution over long periods of time. Such study was previously impossible because, as has often been remarked by viral evolutionists, "viruses have left no fossil footprints" (quote from Darryl C. Reanney, "Evolutionary Virology: A Molecular Overview," in The Human Herpesviruses, ed. André J. Nahmias, W. R. Dowdle, and Raymond R. Schinazi (New York: Elsevier, 1981), p. 519; for a more recent review of molecular evolutionary studies of viruses, see Adrian Gibbs, "Molecular Evolution of Viruses: 'Trees,' 'Clocks' and 'Modules,'" Journal of Cell Science Supp. 7 (1987): 319-37, and earlier forms could only be inferred from vestigial genetic information in existing viruses.

It is conceivable that a change in some critical condition could cause one of these already widespread viruses to become a threat, but such an occurrence appears unlikely.

Predicting the greatest threats is a more difficult task, made more difficult by significant gaps in our knowledge. Of course, we cannot foretell the future. Many would say that the only sure bet is that the next threat will be one not on any list today, as was the case with AIDS. That is why I think it is much more important and useful to emphasize the general principles underlying viral emergence, rather than to attempt to compile a list. However, since I am flinging about a plethora of virus names as examples, I list here some of the viruses that might be

---

perceived as future threats. Several have already been discussed. Most lists would probably include the following: influenza; the hantaviruses (Hantaan, Seoul, and related viruses); Rift Valley fever; yellow fever; dengue fever; Junin (Argentine hemorrhagic fever); Lassa fever; Marburg and Ebola viruses (members of the family Filoviridae); and various encephalitides, all arthropod borne, such as Japanese encephalitis, Venezuelan equine encephalitis, and Eastern equine encephalitis.

Influenza, of course, is familiar. The hantaviruses—Hantaan, Seoul, and related viruses—cause hemorrhagic fevers with renal syndrome (that is, fevers accompanied by severe bleeding and kidney involvement); these viruses, found in Asia, Europe, and the United States, are naturally occurring viruses of rodents (in the case of Hantaan, a rodent called Apodemus agrarius ). Seoul virus is found in rats, including urban rats in Korea as well as in Baltimore and other American cities. James Le Duc has recently found a possible association between this virus and chronic renal disease in people living in inner-city Baltimore.^[35]

James LeDuc, personal communication, May 1989.

Yellow fever, a mosquito-borne disease characterized by fever and jaundice, originated in Africa and is now widespread in Africa and South America; it probably originated as a virus of monkeys. Dengue fever, a virus in the same family that is also transmitted by the same mosquitoes as yellow fever, probably also originated in the Old World but is now in tropical areas worldwide (Africa, Asia, the south Pacific, South America, and the Caribbean). Other viruses, classified by virologists as members of the Arenavirus family, cause hemorrhagic fevers and are natural infections of rodents. These include Junin (Argentine hemorrhagic fever) and Machupo (Bolivian hemorrhagic fever), and the once infamous (because of its high mortality rate in Western medical missionaries who first came in contact with the virus in the early 1970s) Lassa fever of West Africa, which originated in the rodent Mastomys natalensis , all of which viruses were discussed above. Among viruses believed to have originated in monkeys or apes are two related African viruses, Marburg and Ebola, which cause fever with hemorrhage; the unrelated HIV also can be placed in this category. The various equine encephalitides are mosquito borne but tend to have natural animal hosts as well.

Of the viruses listed, I think that influenza, dengue, and the hantaviruses are of greatest potential importance to North America; the recent outbreaks of Oropouche in the Caribbean and Central America are also notable. All these viruses either are widening their scope (dengue, hantaviruses, and recently Oropouche) or still cause recurrent pandemics

---

(influenza). Because of its proximity, dengue might be a special concern. It is widespread in Asia and is also spreading over the Caribbean basin. A dengue outbreak in Cuba in 1981 involved over 300,000 cases. Under certain circumstances an individual who was previously infected with one variety (technically, subtype) of dengue virus can develop a severe form known as dengue hemorrhagic fever upon later infection with a different subtype. The frequency of dengue hemorrhagic fever is increasing as several subtypes of dengue virus extend their range. Aside from the viruses I have listed, there is also always a likelihood that other, as yet undescribed or presently obscure, zoonotic viruses may emerge, as did HIV and Lassa fever. That is why I have emphasized the principles and used these viruses only as examples.

Although the framework offered here identifies the essential conditions for viral emergence, there is still a great deal to learn. Consider influenza. With all the possibilities for recombination, and many human infections annually (perhaps 100,000,000), pandemic influenza strains appear only once every twenty years or so. Why? To put the question in more technical terms, what restrains the emergence of new viruses? Comparatively little is known about this fascinating question, although some patterns are beginning to appear. We also only vaguely understand what factors are required for efficient transmission of viruses in humans. Apart from influenza, many of the other viruses discussed here—such as Junin (Argentine hemorrhagic fever), Marburg, Hantaan, and Lassa fever—fortunately have limited ability to spread from person to person. They would have been devastating if they had that ability or if they were to acquire it. We also need to know more about the mechanisms and determinants of interspecies transfer of viruses. This is a complex matter involving both viral and host factors, but some information is already available and the question is susceptible to further scientific attack. Equally little is known about constraints on viral evolution. In particular, the role of natural selection in shaping or restraining viral evolution has been little explored. Certainly, constraints operate at the level of the virus—host interaction and the maintenance of the virus in nature. In order to survive, viruses must be maintained in nature in some living host. This requirement alone must impose strong selective pressures on a virus.

Even with these gaps in knowledge, we now possess, at least embryonically, the necessary intellectual foundation and tools for attacking these questions, and are faced with the challenge of dealing with the problem of disease emergence. We are not outside the problem; we are

---

learning that emerging viruses do not come as a malevolent rain from above. Human actions have influenced many of these calamities, including HIV. This is, perhaps, both the good news and the bad: We are not completely helpless; but before we can begin to do something, this issue must become a social and economic priority. In many ways, this may prove to be a harder problem than the virological one. However, the essential conclusion is that we must learn to be aware of the consequences of our own actions.

Despite its limitations, the historical record provides clues to traffic patterns. As McNeill has noted, new diseases tend to emerge when populations cross disease boundaries.^[36]

William H. McNeill, Plagues and Peoples (Garden City, N.Y.: Doubleday, 1976). His recent essay "Control and Catastrophe in Human Affairs," Daedalus 118 (Winter 1989): 1-12, giving his views on human attempts to control catastrophic events, may also be of interest.

Recurrent patterns, and near recurrences, abound in history. As pointed out by Elizabeth Fee and Daniel Fox in their Introduction to this volume, such recurrences can be misleading when they are used as analogies. However, they can also be instructive when viewed as manifestations of similar biological processes and traffic patterns that have continued throughout history. The historic association of bubonic plague with rats, and its entry at seaports, is well known.^[37]

Philip Zeigler, The Black Death (New York: Harper and Row, 1969); Robert S. Gottfried, The Black Death (New York: Free Press, 1983). As a non-historian, I have sometimes idly wondered whether the history of Europe would have been different if the people of this period had known sooner about this association and if the simple metal rat-catcher that prevents rats from leaving ships had been available in the Middle Ages.

Hence the concern when a virus resembling Hantaan was found only a few years ago in rats living around Baltimore harbor. Although an imperfect analogy, the discovery should remind us that the historic association of rodents, ports, and disease dissemination is not an antiquarian oddity.

More recently the epidemiology of AIDS itself—although, tragically, not its effects—could have been inferred from what was known about hepatitis B, which had a remarkably similar epidemiology. Long before the viral etiology of AIDS was defined, epidemiologists had demonstrated the similarity of transmission patterns for AIDS and for hepatitis B, with identical high-risk practices and risk groups.^[38]

W. Thomas London and Baruch S. Blumberg, "Comments on the Role of Epidemiology in the Investigation of Hepatitis B Virus," Epidemiologic Reviews 7 (1985): 59-79.

This information began pointing the way toward suitable precautions to limit spread.

If we had a science of traffic patterns, part biology and part social science, we might have made these inferences more readily, with many lives saved. Perhaps such a field may be struggling to emerge, and among those in the forefront might be mentioned Joshua Lederberg, Baruch Blumberg, Frank Fenner, Edwin Kilbourne, Karl M. Johnson, Thomas Monath, Christoph Scholtissek, Luc Montagnier, D. A. Henderson, Mirko Grmek, William McNeill, the late Fernand Braudel, and Daniel Fox.

Enthusiasm, of course, must be tempered by reality. Even if such a science were developed, and even if there were universal agreement on an agenda, it will probably never be possible to anticipate or prevent every episode of disease emergence. Aside from human factors, our

---

knowledge and ability to act will always be imperfect: not every consequence can be anticipated, the world is too complex, and the generation of new pathogens by viral evolution, however constrained and rare it may be, is still possible. At the moment perhaps the most we can hope for realistically is to begin making inroads into the problem.

Why has it taken so long to develop such thinking? For one thing, we may have been misled by our own preconceptions. Virologists and microbiologists have been concerned with the properties of the disease agents—physical and molecular—and of the diseases they cause, and have tended to concentrate on the particular, possibly to the detriment of defining features in common. Until very recently there were many examples of viral traffic and some examples of mutations, but we had not critically evaluated the contributions of each mechanism to viral emergence. It also seemed too daunting a problem. But, as the examples discussed here should demonstrate, the variations in the agents themselves may be less crucial than the traffic laws, the conditions that allow introduction and dissemination in the human population. While this idea is possibly a logical extension of the Darwinian emphasis on natural selection in the environment, it is nevertheless rarely considered. An appropriate global emphasis on conditions, as in historical thinking, would therefore be valuable in combination with the powerful molecular tools now available for virus detection.

Biological scientists may also find it difficult to believe that people themselves bear much of the responsibility for what may seem at first to be natural processes. Historians and social scientists are more accustomed than virologists and microbiologists to think in terms of the consequences of human actions, and of conditions that cause or permit certain developments, and try to infer these predisposing conditions from the results. This perspective is valuable and complements the kind of analytic and causal thinking in which biological scientists are trained. As the historian Marc Bloch put it, "The virus [sic] of the Black Death was the prime cause of the depopulation of Europe. But the epidemic spread so rapidly only by virture of certain social—and, therefore, in their underlying nature, mental—conditions."^[39]

Marc Bloch, The Historian's Craft, trans. Peter Putnam (New York: Knopf, 1953), chap. 5, quoted sentences, p. 194.

Like every other kind of traffic, viral traffic is increasing. What we are now learning about viral emergence shifts the burden to society at large, to all of us. The conditions I have described are really manmade. Consequently, we need to develop greater sensitivity to our environment and the complex ecological relationships that have evolved. In many cases viral emergence follows deforestation and is another unanticipated

---

consequence of despoiling the environment. As deforestation progresses worldwide, as human activities continue to alter the environment, as population influx into Third World cities continues unabated, as every part of the world becomes more accessible, one would expect disease emergence to accelerate. Our first line of defense is to recognize that these and similar human activities can have serious health consequences, and to anticipate these consequences. We need effective strategies to deal proactively, before their spread becomes critical, with viruses as yet unrecognized and with those that disseminate efficiently. To put it simply, if we are often the engineers of viral traffic, we need better traffic engineering. We need viral traffic studies and road maps of disease. I mean this not only literally, in the sense of medical geography, but also metaphorically. When agricultural development is desirable, it would also be wise to consider and plan for disease emergence as a possible side effect. Environmental impact surveys, conducted thoroughly and systematically, should also include consideration of the microbial and viral fauna in the region.

How might such studies be done? Until a more systematic approach can be defined, we have only the rudiments of an answer, but some generalizations can be made. Certain activities, notably the sorts of environmental changes I have discussed, should be recognized as potentially hazardous, especially in tropical regions. Surveys can test for a known virus when there is a proposed expansion of conditions favorable to viral transmission in areas where this virus is endemic. For example, in view of the history of Argentine hemorrhagic fever, plans to clear new areas of the pampas could trigger field surveys to test for the presence of Junin virus in local rodents. Plans for dam building in certain parts of Africa should bring Rift Valley fever to mind, with appropriate field surveys for this virus.

Unknown viruses will present more of a challenge. One can work from analogy with known examples, and search for viral relatives in similar environments. We do not know what other viruses exist in nature, but using the biotechnological tools now available (such as PCR, discussed in note 34) for broad and rapid testing, we can make more systematic efforts to find out. Fenner has noted that our knowledge of arbo (arthropod-borne) viruses increased dramatically in the 1950s and 1960s largely because the Rockefeller Foundation funded a program to screen for arboviruses in the field by the methods then available.^[40]

Frank Fenner, "Keynote Address," in Viral and Mycoplasmal Infections of Laboratory Rodents, ed. Pravin N. Bhatt et al. (Orlando, Fla.: Academic Press, 1986), p. 21; amplified in personal communication, November 1989.

This screening was done mostly by simple biological assays in which samples of ground-up mosquitoes were injected into mice in order to detect viruses

---

pathogenic for mammals that might be present in the mosquitoes. Crude as this sort of screening was, it yielded many new viruses. However, such screening has many disadvantages and is not cost-effective. It is also difficult to evaluate the actual human or animal disease potential of the viruses discovered in this way. Present technology makes it much easier to identify families of viruses in human, primate, rodent, or arthropod populations by broad-based PCR and serological techniques.

Industrialized nations must learn to assist the Third World in financing the needed planning and protective measures to accompany development projects. In many cases relatively simple measures could help greatly, if they are chosen well and if there is sufficient global resolve to implement them. For example, rebuilding water systems in tropical cities to reduce or eliminate open water sources could have a real impact on dengue. Perhaps such projects could someday become priorities before their need reaches the crisis stage. In the intellectual arena our best strategy may be to encourage expanded attempts to find answers to the scientific questions mentioned above, and to forge stronger alliances between molecular virology and such organism-based approaches as field biology, evolutionary biology, and pathogenesis (mechanisms of disease and host-virus interactions). Several scientists have expressed concern that field virologists will soon be in critically short supply; with a paucity of training programs and trainees, the outlook for the future is bleak. In the Third World, for example, there are few well-staffed and thoroughly equipped field laboratories, and their number is decreasing.

The tragedy of AIDS has spurred us on to a consideration of these issues, but it is an unfortunate comment on human nature that such adversity is required before these issues are considered at all. Even with AIDS, the virological problems were largely secondary to a social problem: the failure to recognize the threat and mobilize responses in a timely way.^[41]

Randy Shilts, And the Band Played On: Politics, People, and the AIDS Epidemic (New York: St. Martin's Press, 1987), describes many of these failures and their disastrous consequences.

The "moral equivalent" of war, as many leaders have learned to their chagrin, is a poor substitute for the actual thing. It is hard to sustain fervor without a visible adversary. There are also insufficient economic incentives to mobilize concerted action in advance of a crisis. One can only hope that the value of doing so will be appreciated before we are in the throes of another crisis. Ironically, the costs are likely to be small in comparison with major military projects. As Henderson has pointed out, the eradication of smallpox, a landmark in infectious disease control, was accomplished at a total cost of about $300 million.^[42]

Donald Henderson, personal communication May 1989, February 1991; and D. A. Henderson, in F. Fenner et al., Smallpox and Its Eradication (Geneva: World Health Organization, 1988). Dr. Henderson directed the world smallpox eradication program.

---

It is unlikely that infectious diseases will ever be totally eliminated. Our desire to believe that they can be eliminated may reflect the irrational feelings of terror and loss of control that thoughts of these diseases inspire.^[43]

The subject of much literature, from antiquity to now. Susan Sontag, AIDS and Its Metaphors (New York: Farrar, Straus and Giroux, 1988), demonstrates the power of these images even now.

We have no recourse but to confront these feelings and to deal constructively with them. Even in this highly technological age, we cannot control our biological milieu. At the same time, we must recognize the role that we ourselves play in shaping this milieu. We are part of a complex, interlinked world that we can alter but do not fully control, and science is the study of this complexity. The periodic appearance of "new" infectious diseases serves to remind us of this reality.