Malaria and Rome: A History of Malaria in Ancient Italy Page 6
Evolution of malaria
29
trend towards anthropogenic global warming.⁸ The effects of these climate changes in Italy have recently attracted attention because of their relevance to the preservation of the famous ‘Iceman’ discovered in the Alps (as it turned out, just on the Italian side of the border with Austria). Fortunately for modern archaeologists, the Iceman died towards the end of the mid-Holocene climatic optimum, in the late fourth millennium , at a time when neoglacia-tion was commencing (i.e. the Alpine glaciers were starting to advance again as mean annual temperatures dropped). This covered his body with ice, preserving it until anthropogenic global warming in the last few years caused the glacier to begin to retreat again, exposing the frozen corpse.⁹
The development of P. falciparum is heavily dependent on the temperature. Since it requires a minimum temperature of about 20°C for the completion of sporogony inside the mosquito, climatic conditions during the Neolithic period were in fact substantially more favourable for the spread of P. falciparum and its vector mosquitoes into southern Europe than they were in the first millennium
or any other period after the Neolithic. What are now the Saharan and the Arabian deserts also received substantially more rainfall during the mid-Holocene climatic optimum than they do today, creating more breeding sites for mosquitoes.¹⁰ This would have assisted the spread of malaria from tropical Africa towards the southern shores of the Mediterranean. It is even conceivable that the geographic range of members of the Anopheles gambiae complex, the most important vector of malaria in tropical Africa today, may have extended further north in Africa than it has done in recent times. Mosquitoes can evolve very rapidly. For example, populations of Culex pipiens confined to London Underground tunnels and separated from above-ground populations have evolved new host preferences (mice, rats, and humans instead of birds), reproductive isolation from above-ground populations, new mating patterns (stenogamy instead of eurygamy), loss of winter diapause, and the possibility of oviposition without prior ingestion of a blood meal, all in no more than about one hundred years. Similarly the most ⁸ Zulueta (1973) and (1987); Sallares (1995) on the mid-Holocene optimum in western Eurasia. Recent research in China, reported in Nature, 390 (1997: 209), confirms that it was a worldwide phenomenon. It is estimated that the mean annual temperature was 2–4˚C
warmer, with 20–40% more rainfall, in China in the period 6000–2000 .
⁹ Baroni and Orombelli (1996).
¹⁰ Claussen and Gayler (1997).
30
Evolution of malaria
recent research in molecular evolution suggests that speciation in the A. gambiae complex in Africa is an active, ongoing process.
Mario Coluzzi has cogently argued that the modern strains of A. gambiae, which are exceedingly efficient at transmitting malaria, are of recent (Neolithic period onwards) origin.¹¹ The evolution of these extremely efficient vector strains would have permitted an increase in the transmission rate of P. falciparum malaria in tropical Africa. Coluzzi has suggested that an increase in the pathogenicity of P. falciparum might have accompanied the increased transmission rate. Recent research in molecular diversity indicates that the populations of the A. gambiae complex have been increasing recently but have also long had a large effective population size (the size of the breeding population). This suggests that its potential as a vector of malaria in Africa does go back to prehistory and is not a product of human population growth in modern times in Africa, even if it was not quite as efficient a vector then as it is today.¹² One population of Pygmies (hunter-gatherers until very recently) in central Africa has a very high frequency of haemoglobin S, a mutation which confers resistance to P. falciparum malaria on heterozygotes for the sickle-cell trait. Cavalli-Sforza argued that this implies that P. falciparum malaria was already widespread in central Africa before the invention or spread of agriculture in Africa.¹³ Because of the heat, the Neolithic period was the most likely period for the spread of P. falciparum malaria into southern Europe. Its previous evolutionary history, as has just been argued, suggests that the small human population sizes of that period would not have prevented it from becoming endemic in southern Europe then. Malaria is a disease that tends to exist in small foci because mosquitoes generally do not fly far from their breeding grounds, not more than five or six kilometres under Mediterranean environmental conditions, although occasional much longer migrations have been recorded.
In 1959 a migration of about 280 kilometres by the Egyptian mosquito species Anopheles pharoensis reintroduced malaria to Gaza and the coast of Israel, from which it had previously been eradicated.
¹¹ Byrne and Nichols (1999) on the London mosquitoes; Coluzzi et al. (1979); Coluzzi (1999).
¹² Lehmann et al. (1998); Powell et al. (1999).
¹³ Cavalli-Sforza (1986: 153–5, 416–19). Pygmy populations in forest environments, which A. gambiae is reluctant to enter, do not have high frequencies of haemoglobin S. Phylogenies based on mitochondrial DNA sequences suggest that some of the Pygmies are among the most ancient human populations.
Evolution of malaria
31
A propensity towards such migrations by this important malaria vector-species in the lower Nile valley might have helped P. falciparum to break out of Africa in the distant past.¹⁴
J. L. Angel used to invoke the high frequency of porotic hyperostosis in crania belonging to skeletons recovered from Mesolithic–Neolithic archaeological sites in Greece as evidence for a high frequency of P. falciparum malaria in Greece at that time.
There is now a general consensus among those interested in this problem that porotic hyperostosis, whose proximate cause is an iron-deficiency anaemia, has several other possible ultimate causes besides malaria. Consequently porotic hyperostosis cannot be used as evidence for the existence or frequency of P. falciparum malaria in Europe in the Neolithic period.¹⁵ However, absence of evidence is not equivalent to evidence of absence. It still remains quite possible that it was spreading in that period. Recently the application of the techniques of molecular biology to ancient biomolecules has opened up new avenues of research. Immunological tests have been used by two different research groups to identify the histidine-rich protein-2 antigen of P. falciparum in the mummies of several predynastic individuals from Egypt, dating to c.3200 .¹⁶ This constitutes some direct evidence for the existence and activity of P. falciparum on the periphery of the Mediterranean world already in the fourth millennium . Further research into ancient biomolecules (especially DNA) from human skeletal remains excavated on archaeological sites in southern Europe offers the best prospect of obtaining direct evidence for P. falciparum malaria in Europe in prehistory.¹⁷ There was clearly some contact between Egypt and ¹⁴ M. T. Gillies in Wernsdorfer and McGregor (1988: i. 455) expressed the view that mosquito flight range is a property of the environment, not the species, depending on the availability of breeding sites and food, but it is clear that they generally do not fly far; Garrett-Jones (1962); Halawani and Shawarby (1957) on malaria in Egypt in recent times.
¹⁵ Angel (1966); Borza (1979), Zulueta (1987: 200), Sallares (1991: 275–7), Stuart-Macadam (1992), Grmek (1994), Corvisier (1994: 299–303), and Larsen (1997: 30–40) all agree that porotic hyperostosis does not necessarily indicate malaria.
¹⁶ R. L. Miller et al. (1994); Cerutti et al. (1999). cf. Marin et al. (1999).
¹⁷ The problem with trying to detect ancient proteins is that antibody reactions depend on the conformation of proteins. Since protein conformation would be expected to degenerate over time, it is not clear what degree of specificity could be expected in any particular antibody reaction with degraded proteins. G. M. Taylor et al. (1997) unsuccessfully tried to amplify ancient DNA from one of the same individuals studied by R. L. Miller et al. (1994), namely the Gurna mummy dating to c. 700 . There are many possible explanations for this failure. Similarly C. Plowe, reported in Parasitology Today,
14 (1998: 9) expressed scepticism about the results obtained by R. L. Miller et al. (1994). Consequently further research is need-32
Evolution of malaria
Greece at least as early as the Early Bronze Age in the third millennium . The most striking illustration of this contact was the construction of the small-scale imitations of Egyptian pyramids at Hellenikon and Ligourio in the Argolid in Greece, which Pausanias passed on his travels much later. Consequently it is quite possible that P. falciparum could have been transmitted directly from Egypt to Greece at that time.¹⁸
However, there is another, even earlier, possibility. The Neolithic period commenced in Europe with the introduction of agriculture by human populations from the Near East, according to the generally convincing arguments presented in the monumental book by Cavalli-Sforza and his colleagues, a very important contribution to knowledge.¹⁹ Agriculture—specifically the cultivation of cereals and legumes—first developed in the general vicinity of modern Israel, Jordan, and Syria. These regions in antiquity certainly included some significant areas of wetlands, along the Mediterranean coast and in the Jordan valley, which harboured amphibious animals as large as the hippopotamus and permitted the cultivation of aquatic plants like papyrus. In more recent times, until they were drained, these wetlands were intensely malarious.
Similarly in antiquity Josephus described as pestilential in summer the air of the Great Plain around Lake Tiberias and the Dead Sea.²⁰ Consequently the earliest Neolithic farmers lived in a region that included some environments that were extremely favourable for malaria.²¹ This consideration supports Angel’s hypothesis that ed to confirm their results. Sallares and Gomzi (2001) discussed the problems in applying immunological tests to ancient materials. However, the results of the studies in molecular evolution cited earlier based on comparisons of modern DNA sequences make it extremely likely in any case that P. falciparum was present in Egypt by the fourth millennium , since such research does not suffer from the same technical problems as research on ancient biomolecules. Schiff et al. (1993) described the Para Sight test.
¹⁸ Theocharis et al. (1997) and Pausanias 2.25.6 on the Greek pyramids.
¹⁹ Cavalli-Sforza et al. (1994).
²⁰ Josephus, de bello Iudaico 4.8.2, ed. Bekker (1855–6): ƒkpuroıtai d† ¿r6 qvrouß tÏ ped≤on, ka≥ di’ Ëperbol¶n aÛcmoı perivcei nos*dh tÏn åvra (The plain is burnt up during the summer season, and extreme drought makes the air unhealthy).
²¹ Tacitus Histories 5.6–7 also described the Dead Sea region as pestilential with bad air: lacus immenso ambitu . . . gravitate odoris accolis pestifer (a lake with a huge circumference . . . whose oppressive smell brings pestilence to the local inhabitants). Hirsch (1883: 202) noted that malaria affected extensive regions near the Dead Sea in the nineteenth century, as well as the Bekaa valley in Lebanon at an altitude as high as 1200 metres (cf. Leeson et al. (1950) ). Fisher (1952) made the interesting observation that mosquitoes are carried up to the Bekaa valley by rising air currents which regularly occur in that region; Amadouny (1997); Filon et al. (1995) extracted ancient DNA showing the presence of b-thalassaemia from human skeletal Evolution of malaria
33
all the three species of human malaria under consideration, including the most dangerous, P. falciparum, were carried to Europe inside the bodies of the very first Neolithic farmers. The diagnosis of thalassaemia, a human genetic disease that confers some resistance to malaria, in the skeleton Homo 25 (a male sixteen or seventeen years old) from the PPNB (Pre-Pottery Neolithic B) village of Atlit Yam (now submerged off the coast of Israel) supports the idea that malaria was already active in the Levant at the dawn of agriculture.²² Given that the climate in the Neolithic period was actually exceedingly favourable to it, whether P. falciparum would have survived in new environments in southern Europe depended, as was noted earlier, on whether it encountered species of mosquito that were capable of acting as efficient vectors. Only a minority of the European species of Anopheles mosquito are good vectors for malaria.
This problem leads on to the fourth pillar of the late-introduction theories, namely the question of the possible refractoriness of mosquitoes to infection with P. falciparum. Experiments were performed using samples of A. labranchiae, originating from the coast of Tuscany near Tarquinia, and of A. atroparvus, from the Orcia river valley near Siena and from the upper Volturno valley north of Naples, to see if these Mediterranean populations of mosquitoes could ingest gametocytes from tropical strains of P. falciparum and successfully transmit sporozoites to new hosts.²³ The results were negative, in agreement with more extensive research of this kind subsequently performed in Russia which indicated that in general tropical strains of P. falciparum are not adapted to the mosquito species of Eurasia. Zulueta used these results to argue that a long period of adaptation would have been required to overcome refractoriness on the part of mosquito species in Greece and Italy.²⁴
However, even if this were the case, it would not prove that P. falciparum was a newcomer in classical times. Since the new data for its remains from Akhziv in Israel. For hippopotamus and papyrus in the region see Sallares (1991: 26, 370, 400–2). Theophrastus, HP 9.7.1–2 also mentioned the marshes.
²² Hershkovitz et al. (1991).
²³ Ramsdale and Coluzzi (1975); Zulueta, Ramsdale, and Coluzzi (1975). Earlier experiments at Horton Hospital in England had shown that the English malaria vector A. atroparvus is similarly unable to transmit tropical African strains of P. falciparum, although it can transmit Italian strains of P. falciparum. However, it could transmit all strains of P. vivax that were tested, although it is very inefficient at transmitting P. malariae (Shute (1940) and (1951)).
²⁴ Zulueta (1973), (1987), and (1994).
34
Evolution of malaria
presence in Egypt in the fourth millennium suggest that P. falciparum was already present in the Mediterranean world thousands of years before classical times, a long period of time was indeed available for the refractoriness of European species of mosquitoes to be overcome. Moreover the experiments yielded no information about the critical factor of the length of time required for refractoriness to be overcome. It is not clear at the moment whether in this particular case overcoming refractoriness required evolution in the mosquito, evolution in the malaria parasite, or coevolution.
Mosquitoes certainly have several defence mechanisms against intruding foreign bodies in general, and may have genes that specifically respond to invasion by malarial parasites. Melanotic encapsulation or melanization is the most well known of these processes. This process is employed by mosquitoes and other insects to surround and inactivate pathogens. A thick layer of melanin is deposited around the malaria parasite when it tries to cross the mid-gut epithelium of the mosquito, en route to the salivary glands for the formation of sporozoites for transmission to another human. Cross-breeding experiments suggest that it is under fairly simple genetic control (no more than about three major loci being involved), with the implication that the expression of the process of melanotic encapsulation can be increased or diminished rapidly in mosquitoes.²⁵
However, there is the complication in the Mediterranean case that the mosquitoes, which transmitted Mediterranean strains of P. falciparum in the past, are refractory to modern tropical strains, as has just been seen. This suggests that differences between various strains of P. falciparum were also important in some as yet un-defined way. Evolutionary processes tend to be very rapid among micro-organisms and, as will be seen shortly, P. falciparum has the capacity for very rapid genetic change. If it did not exist in the western hemisphere before Columbus, several species of mosquito indigenous to the western hemisphere quickly became effective vectors of P. falciparum, a pathogen which they had never encountered before 1492. The failure of Amerindian populations to develop high frequencies of any of the wide range of genetic mutations that confer degrees of resistance to P. falciparum malaria in ²⁵ Lombardi et al
. (1986); Collins et al. (1986); Richman and Kafatos (1996); Yan et al.
(1997); Zheng et al. (1997); Billker et al. (1998); Feldman et al. (1998); Paskewitz and Gorman (1999); Barillas-Mury et al. (2000); Oduol et al. (2000).
Evolution of malaria
35
many Old World human populations suggests that they have only recently been exposed to it.²⁶ Grmek rightly observed that the different responses of European mosquitoes to strains of parasite from different geographical areas in fact suggest a long period of separation between the tropical and the subtropical/temperate strains of P. falciparum.²⁷ The experiments at Horton Hospital in England suggested that Italian strains of P. falciparum from Sardinia and Salerno were more severe and more virulent and exhibited a faster growth rate than strains originating in tropical countries (with the proviso that the sample sizes in these experiments were small).²⁸
A faster progression of the infection on the part of European strains of P. falciparum might well have been an evolutionary adaptation to the shortness of the season each year that was suitable for its reproduction because of the climatic constraints in Europe.
Another crucial argument is that since the western Mediterranean mosquito vector-species A. labranchiae is common in North Africa, which in fact is its main area of distribution, as well as in Italy, while A. sacharovi, the second major Mediterranean vector with a more easterly distribution, also occurs in the Near East as well as in Greece, P. falciparum had every opportunity to evolve adaptation to its European vectors in Africa and in the Near East before even arriving in Europe. The nature of P. falciparum malaria in North Africa in the past is an important unresolved question.
The French colonists in North Africa in the nineteenth century had severe problems with malaria. Of course it was in North Africa that Alphonse Laveran discovered malarial parasites, at the hospital in Constantine in 1880. In 1832–3 two earlier French army doctors, Antonini and Maillot, working at the hospital of Bône in Algeria, found that giving high doses of quinine and a generous diet and ending the practice of bleeding reduced malaria mortality rates from 30% to 5%.²⁹ However, it is not clear whether the French colonists encountered the ‘European’ strains of P. falciparum, or the tropical strains, or both. Nevertheless, since A. labranchiae is abundant in North Africa as well as in Italy, and since the tropical strains ²⁶ Effertz (1909), Dunn (1965), and Cook ((1998: 48–9), wrongly identifying P. vivax with quartan fever) on the Americas. Li et al. (2001) published molecular evidence for two separate introductions of P. vivax to the western hemisphere, possibly both since Columbus.