Malaria and Rome: A History of Malaria in Ancient Italy Page 5
³⁶ Tibullus, 3.5.1–4: Vos tenet, Etruscis manat quae fontibus unda, | unda sub aestivum non adeunda Canem, | nunc autem sacris Baiarum proxima lymphis, | cum se purpureo vere remittit humus.
22
Types of malaria
observed from March onwards under Mediterranean environmental conditions and was not especially associated with the harvest as such, although it certainly was frequently transmitted during the harvest in the past (see Ch. 5. 4 below). Grmek rightly noted that acute enteric diseases are also prevalent in summer and autumn, as seen, for example, in the population of Florence at the time of the Catasto (census) of 1427 (see Ch. 5. 2 below).³⁷ However, the association with the harvest is very significant. Peasants often slept out in the fields during the harvest, where they were very vulnerable to mosquito bites but far from stagnant sources of water in urban areas, the commonest source of enteric diseases. A very important practical reason for farmers in Lazio to stay in their fields at night was to prevent their crops being stolen. The vulnerability of farm workers to malaria infection during the harvest has frequently been noted in Italy and other Mediterranean countries, and indeed all over the world, wherever malaria occurs. In the village of Lodé in northeastern Sardinia, where in the 1930s the morbidity rate from malaria was 90%, 50% of the population slept outside the village during the summer.³⁸
Besides this Homeric text, there is another, completely different, but much stronger line of argument for the presence of P. falciparum malaria in Greece by the eighth century . The B+ IVS nt 110
mutation for b-thalassaemia, which confers some resistance to P. falciparum malaria, reaches its highest frequencies in Greece and was spread by Greek colonization. It is also common in the modern populations of those parts of southern Italy which were colonized by the Greeks in the seventh and eighth centuries . This indicates that the mutation was already present in human populations in Greece by then. The implication is that P. falciparum was already active in Greece by the eighth century (see Ch. 5. 3 below for further discussion).
³⁷ Marchiafava and Bignami (1894); Grmek (1983: 65–6).
³⁸ For infection of harvesters see M. E. Danubio in Greene and Danubio (1997: 328) on Lodé; Bercé (1989: 242); Celli (1900: 176); North (1896: 243–4) on Lazio; Desowitz (1992: 111–12) on south-east Asia; M. T. Gillies in Wernsdorfer and McGregor (1988: i. 473) for Turkey; McNalty (1943) for infection of harvesters by P. vivax in Scotland in the eighteenth century. Plutarch, Moralia 137c, discussed by Jones (1908: 545), does not describe any symptoms of the ailments of harvesters. The Chronicle of Joshua the Stylite, ed. Wright (1882), ch. 85, p. 67, shows that in Syria at the end of the fifth century it was normal for farmworkers to sleep outside on the threshing floor at the time of threshing.
3
Evolution and prehistory of malaria
Several quite different arguments have been proposed in favour of the theory of a late introduction (or reintroduction) of P. falciparum malaria to Greece and Italy. W. H. S. Jones was looking for an explanation for the decadence, as he saw it, of ancient Hellenic civilization. This approach can be safely dismissed now without further discussion. Zulueta and Grmek and other authors exploited a series of much more scientific arguments which certainly merit extensive discussion, one by one. These ideas are briefly summarized here, before discussion: (1) the hypothesis that P. falciparum must be a human pathogen of recent origin because its extreme virulence suggests that it has not had time to adapt to humans as a host;
(2) the hypothesis that large human population sizes would have been needed before malaria could become endemic in Greece and Italy;
(3) the hypothesis that the species of mosquito which were the most effective vectors of malaria in Europe would have taken several thousand years to spread into southern Europe after the end of the last Ice Age;
(4) the hypothesis that because these species of mosquito are refractory to tropical strains of P. falciparum a long period of evolution would have been required before they could become efficient vectors of the parasites.
It was once widely believed, as a general principle of parasite ecology and epidemiology, that parasites tend to evolve towards symbiosis or commensalism with their hosts, because in the long run it would not be in the interests of a parasite to kill its host.
This is frequently given as a reason why P. falciparum malaria, an extremely virulent disease, must be a human disease of recent origin, which might only have evolved during the last few thousand years. This theory bolstered the conclusions of A. P. Waters and co-workers, who argued on the basis of phylogenetic analyses of 24
Evolution of malaria
0.02
0.044
0.018
Babesia
0.042
0.050
9,998
Theileria
0.069
0.022
Eimeria
0.056
9,998
Toxoplasma
0.048
P. malariae
0.042
7,782
P. ovale
0.012
10,000
9,963
P. vivax
8,802
0.026
P. cynomolgi
0.014
6,120
10,000
P. knowlesi
6,123 0.015
P. fragile
0.029
0.024
P. lophurae
0.064
0.026
0.013
10,000
P. gallinaceum
9,993
0.009
0.034
P. reichenowi
10,000
P. falciparum
0.056
P. berghei
Figure 1. Evolutionary relationships of selected Plasmodium species indicated by neighbour-joining analysis of 18S ribosomal gene A sequences in a ClustalX alignment (Thompson et al. (1997) ), modified manually in the BioEdit sequence editing program (Hall (1999) ). The apicomplexan species Toxoplasma gondii, Eimeria mitis, Babesia bigemina, and Theileria annulata were used as outgroups, with 10,000 bootstrap replications. Bootstrap values are shown below the branches and branch lengths above the branches.
DNA sequences that P. falciparum is closely related to avian malarias, but not to other primate malarias, with the implication that P. falciparum malaria is the result of a lateral transfer of a malaria parasite from birds.¹ It remains controversial whether or not P. falciparum is more closely related to avian malaria species than it is to other primate malaria species. Settling this phylogenetic question is not essential for the purposes of this book, since the evolutionary relationships in question date back to over a hundred million years ago. However, the ecological problem is important for current purposes. Research in ecology in recent years has reached the conclusion that it is not inevitable that a parasite or other pathogen will evolve towards avirulence. It all depends on the precise circumstances, in particular it depends on the factors favouring transmission of the parasite to new hosts, which is what will determine its evolutionary fitness.
¹ Waters et al. (1991), discussed by Brooks and McLennan (1992) and McCutchan et al.
(1996).
Evolution of malaria
25
Since all the species of human malaria depend on Anopheles mosquitoes to convey them from human to human, it is not in the interests of the parasite to harm its vector. That might adversely affect transmission of the parasite to new hosts. Consequently none of the species of human malaria causes any great harm to mosquitoes, or at least mosquito immune defence systems appear to be adequate to limit damage. However, the situation with regard to the human host is much more complicated. In an environment where transmission from host to host is possible all the year round, as in the tropics, the type of parasite which will achieve the greatest evolutionary fitness is the one which achieves the highest rate of reproduct
ion in the host, irrespective of what that does to the host.
P. falciparum achieves a very high rate of reproduction in a number of ways, for example by having the ability to invade erythrocytes of all ages, whereas P. vivax only invades reticulocytes (immature erythrocytes) with the Duffy antigen. P. falciparum may infect up to 80% of all erythrocytes, whereas P. vivax does not infect more than about 2%, and P. malariae more than about 1% of all red blood cells.
In cold environments, on the other hand, where transmission by mosquito is not possible all the year round, the parasite requires the host to survive during the winter in order to have an opportunity for transmission to new hosts the following year. These ecological considerations explain why extreme virulence is adaptive for P.
falciparum in its home in tropical Africa, while avirulence is adaptive for P. vivax and P. malariae in colder environments. Consequently the extreme virulence of P. falciparum does not constitute evidence for a recent evolutionary origin.²
The exponential expansion of DNA sequencing in recent years has yielded the important result that the human parasite P. falciparum forms a monophyletic clade with P. reichenowi, a malaria species which infects chimpanzees in Africa. This clade is not closely related to any of the other three species of human malaria.
Analysis of DNA sequences from ribosomal RNA genes (see Fig. 1) and from the circumsporozoite protein gene suggests that the ² e.g. Fiennes (1978: 105–12) regarded P. falciparum as a recent pathogen of man because of its virulence, but Garnham (1966: 279) was sceptical of such theories. Ewald (1994: 42–6), Anderson and May (1991: 648–52, cf. 392–419), and Coluzzi (1999) give various views on the significance of its virulence; Mackinnon and Read (1999 a) and (1999 b). Marchiafava and Bignami (1894: 103) observed that ‘malignancy coincides with an exceptionally abundant quantity of parasitic forms, a quantity much more abundant—where the cases terminate fatally—in the blood of the viscera than in the blood of the finger’.
26
Evolution of malaria
common ancestor of the P. falciparum/P. reichenowi clade diverged from the common ancestor of the P. vivax/P. malariae clade about 165 million years ago. Anopheles mosquitoes, which transmit human malaria, do not appear in the fossil record until the Oligo-cene period (26–38 million years ago), but studies of molecular evolution suggest that the Anopheles family is very ancient. The protein and DNA sequences of the 35kb circular DNA molecule and the enolase gene of P. falciparum manifest very ancient kinship, or at least very extensive horizontal transfers of DNA, embracing not only organellar but also nuclear DNA, with a plant-related lineage.³ It remains controversial whether the various species of malaria were originally parasites of vertebrates or parasites of mosquitoes. It is possible that they were originally parasites of vertebrates because of the similarity of their developmental cycles to those of coccidian intestinal parasites of the suborder Eimeriina.⁴
However, the most interesting result of this research in molecular biology for current purposes is that statistical analysis of the degree of divergence between the DNA sequences of the human parasite P. falciparum and the chimpanzee parasite P. reichenowi puts their date of divergence in the time range of 5–11 million years ago.
Given the inevitable margin of error in these statistical calculations, this date approximates the date of divergence between humans and chimpanzees given by palaeoanthropologists. Consequently it is likely that P. falciparum has been attacking humans and their hominid ancestors since the dawn of human evolution, the split from the chimpanzee lineage.⁵ Similarly recent research suggests ³ Escalante et al. (1995) and Qari et al. (1996) on the molecular evolution of Plasmodium from rRNA gene sequences, cf. Escalante et al. (1998 a) for data from the cytochrome b gene and Rathore et al. (2001) for data from plastid sequences; Capasso (1991), Besansky et al. (1992), and Coluzzi (1999) on mosquito evolution; Hyde et al. (1994), Read et al. (1994) (nuclear DNA), and Köhler et al. (1997) (plastid DNA) on the links of P. falciparum to plant-related lineages; Felger et al. (1997) illustrate the sort of genetic variation which is now being discovered.
⁴ Missiroli (1934: 10–11) was one prominent Italian malariologist who advocated the theory of the close evolutionary relationship of malaria parasites to coccidian intestinal parasites. Capasso (1985: 301) supported the alternative theory that malaria parasites were originally parasites of the salivary glands of mosquitoes. This theory leaves unresolved the transmission question, namely how did the parasites get from mosquito to mosquito, since mosquitoes don’t bite each other. Going back even further in time, Halevy (1998) suggested that plasmodial parasites owe their similarities to plant genomes to descent from toxic algae which infected fish.
⁵ Rich et al. (1998 a) and Ayala and Rich (2000) found a very low rate of synonymous sub-stitutions in housekeeping genes of P. falciparum. They drew the inference from the apparent lack of genetic variation in housekeeping genes of modern strains that all currently existing P. falciparum populations are derived from a single ancestor that lived a few thousand years Evolution of malaria
27
that other major parasitic diseases such as visceral leishmaniasis and trypanosomiasis also co-evolved with humans in Africa. P.
falciparum is one of mankind’s oldest, deadliest, and most persistent foes. This conclusion has considerable implications for the question of the size of host population required by P. falciparum. Evidently it was able to survive for very long periods during which all humans and their hominid ancestors were hunter-gatherers, long before the invention of agriculture, periods when human population sizes were very small. One thinks for example of the figure of 10,000 frequently given by molecular biologists as the effective population size (i.e. the size of the adult breeding population) of the populations (not necessarily the same population) to which belonged ‘mitochondrial Eve’, the last common female ancestor of all currently existing human mitochondrial DNA genotypes (assuming a rarity of recombination), and her male counterpart, the ‘Adam’
currently being revealed by studies of DNA sequences from the Y chromosome. P. falciparum is an extremely ancient human pathogen which was able to survive in small human populations in Africa, the cradle of human evolution.
In contrast, P. vivax and P. malariae are closely related to malaria parasites of monkeys in south-east Asia,outside the cradle of human evolution.⁶ P. vivax, for example, closely resembles P. cynomolgi, a parasite of Macaca monkeys in south-east Asia, in respect of both morphology and DNA sequences. P. vivax and P. malariae were not originally human diseases. They probably first encountered the evolving hominids when Homo erectus spread out from Africa across Asia, probably between one and two million years ago. The ago, even though they accept that the divergence between P. falciparum and P. reichenowi occurred several million years ago. Their controversial theory about P. falciparum cannot be discussed in detail here, but it is probably incorrect or, at best, an exaggeration (their views on the evolution of P. vivax and P. malariae are completely untenable). It does appear that different results are obtained from different parts of the genome, a problem frequently observed in research on molecular evolution (Gillespie (1991: 41) ). Other regions of the P. falciparum genome currently being studied by other scientists are yielding results incompatible with those obtained by Ayala and Rich (e.g. Verra and Hughes (2000) ). I hope that it will be possible within the next few years to obtain direct evidence from ancient DNA permitting an evaluation of the theory of Ayala and Rich concerning a recent cenancestor of P. falciparum.
⁶ Of course the distribution both of species of nonhuman malaria and of other primates might have been different in earlier geological epochs. Skinner et al. (1995) suggested that periodic episodes of linear enamel hypoplasia in fossil teeth of Dryopithecus apes from Can Llobateres in northeastern Spain, dating to the Miocene period about 9.5 million years ago, might have been caused by malaria.
28
Evolution of malaria
question of the origin of P. vivax malar
ia in humans is tied to the question of the FY*O allele in the Duffy blood group locus, which prevents P. vivax parasites from adhering to and entering erythrocytes in nearly all members of sub-Saharan African populations. It is not known whether this allele spread in response to an existing parasite burden and drove P. vivax out of sub-Saharan Africa (its niche being taken by P. ovale), or whether an already high prevalence of this allele (perhaps in response to another pathogen) prevented P. vivax from ever establishing itself in Africa in the first place. In the last few years the Duffy negative allele has also appeared in Papua New Guinea, where P. vivax is endemic. This is an example of evolution in action in human populations today in response to malaria.⁷
Since P. falciparum was present in the heartland of human evolution in East Africa, presumably it would have been carried out of Africa by every successive wave of hominids and humans, from Homo erectus onwards. Whether it would have prospered outside Africa would have depended on the climate and on whether in new environments it encountered species of mosquito able to transmit it. These two factors are the last two pillars of the theory of the late spread of malaria into Mediterranean countries. Zulueta has quite correctly argued that the climate of Ice Age Europe was too cold both for the completion of the developmental cycle of P. falciparum itself within the mosquito and for the principal mosquito vector species in Italy, Anopheles labranchiae and A. sacharovi (= elutus).
He then reckoned that it would have taken thousands of years for conditions to become favourable enough for P. falciparum and its vectors to spread into southern Europe. However, this argument was based on old and out-of-date literature about the Holocene climate. It ignores the mass of evidence which is now available for what climatologists call the mid-Holocene climatic optimum, a period after the end of the last Ice Age and encompassing the Neolithic period until c.3000 , when, owing to periodic shifts in the earth’s position relative to the sun, the northern hemisphere received considerably more insolation than it does today. This resulted in the climate of many parts of the northern hemisphere being up to 2°C hotter than in subsequent millennia. Such temperatures are only now being approached again with the recent ⁷ Livingstone (1984); Zimmerman et al. (1999); Hamblin and Di Rienzo (2000).