As stated previously, natural hybridization has played an important role in the evolutionary history of numerous taxa of plants and animals (e.g. Arnold 1992, 1997; Grant & Grant 1992; Dowling & DeMarais 1993; Rieseberg 1997). However, I am unaware of a review that has focused on how this process has affected the origin and evolution of organisms that negatively impact human populations. In the following sections I discuss several examples for which natural hybridization has been shown to be so involved. In particular, I address the following questions:
The origin of plant pathogens and pests through natural hybridization
Brasier et al. (1999) observed that ‘Plant disease epidemics resulting from introductions of exotic fungal pathogens are a well known phenomenon.’ However, they also concluded that ‘An associated risk — that accelerated pathogen evolution may be occurring as a consequence of genetic exchange between introduced, or introduced and resident, fungal pathogens — is largely unrecognized.’ Though this quote is concerned with one of our current topics, i.e. plant pathogen evolution, it summarizes well the conclusion provided at the outset of this review, that certain roles played by natural hybridization in the evolution of species complexes remain under appreciated (Arnold et al. 1999; Burke & Arnold 2001; Arnold & Burke 2004). Brasier et al. (1999) reflected on the importance of testing for the action of natural hybridization in the evolution of plant pathogens because the outcome of such crosses can include ‘… the acquisition of new host specificities …’ and the ‘… emergence of entirely new pathogen taxa.’ Both types of outcomes have been postulated for nonpathogenic species. The evolution of ecological diversity and the origin of new species via hybridization and introgression have thus been detected in numerous species complexes (e.g. Arnold 1997; Rieseberg 1997; Schweitzer et al. 2002; Rieseberg et al. 2003).
The origin of virulent, fungal species is of the utmost concern for both cultivated and natural ecosystems. The formation of plant, fungal pathogens through natural hybridization is exemplified well by its occurrence in Phytophthora. This fungal genus includes species that attack cultivars (e.g. potato blight) and native species (e.g. forest dieback; Brasier et al. 1999). Significantly, pathogenic variants from the Phytophthora complex were found to be (i) hybrid derivatives and (ii) genetically very similar to the nonpathogenic species, Phytophthora cambivora (Brasier et al. 1999). However, three years after their discovery the hybrids were already responsible for the deaths of more than 10 000 trees (Brasier et al. 1999). Though the tree species affected (i.e. riparian alders) were not a human food source, they were a significant member of an important ecosystem and as such were critical for conservation efforts. This example illustrates a negative effect of plant pathogens on human populations, other than in attacking food sources.
The origin of the pathogenic Phytophthora via natural hybridization between nonpathogens is of particular importance in understanding evolutionary processes and disease control. On the one hand, this occurrence demonstrates the potential for natural hybridization to produce new fungal lineages that possess novel phenotypes and ecological amplitudes (e.g. Tsai et al. 1994; O’Donnell & Cigelnik 1997; Brasier et al. 1999). On the other hand, though the origin of some fungal lineages via hybridization may lead to beneficial interactions (e.g. plant endosymbionts; Tsai et al. 1994), the origin of pathogens obviously poses a threat to artificial and natural ecosystems (Brasier et al. 1999). Both the positive and negative outcomes (for human populations) reflect a recurring hypothesis in the natural hybridization literature; introgression may produce hybrid genotypes possessing novel ecological adaptations allowing their occupation of niches different from their progenitors (Anderson 1948, 1949; Anderson & Stebbins 1954).
Like plant pathogens, some plant pests also show the effect of natural hybridization and introgression. For example, Lewontin & Birch (1966) concluded that the evolution of the Australian fruit fly pests, Bactrocera tryoni and B. neohumeralis (formerly Dacus tryoni and D. neohumeralis) included reticulation. These authors proposed that introgression from B. neohumeralis into B. tryoni allowed the latter species to expand its range and impact additional areas of cultivated fruit species. According to Lewontin & Birch (1966) the footprints of this process included (i) the introgression of a morphological marker and (ii) outbreaks of B. tryoni on both temperate crops such as apples, peaches and various citrus taxa as well as tropical fruit species such as guavas. It was subsequently argued that this hypothesis was likely incorrect (Gibbs 1968; Birch & Vogt 1970). However, recent applications of molecular genetic markers have supported the original conclusion that B. neohumeralis genes have introgressed into B. tryoni (Morrow et al. 2000). Indeed, Pike et al. (2003) demonstrated that introgression between these two species is likely to be asymmetric, from B. neohumeralis into B. tryoni, because there is a difference in mating time for the two species and the F1 and backcross 1 hybrids mate when B. tryoni is active (Pike et al. 2003). Thus, the hypothesis that introgression can expand greatly a pest species’ niche (Lewontin & Birch 1966) is supported by findings from the Bactrocera complex.
Introgressive hybridization and the evolution of the human disease vector Anopheles gambiae
Though each of the above sections illustrates the widespread impact of natural hybridization, the effect of this process on the evolution of organisms that impact human populations may be most readily apparent when we consider vectors of human diseases and human diseases themselves. The role of this process in the evolution of such organisms is obvious due to the ensuing human suffering. However, I argue that the unique outcome of this process for vectors and pathogens, as measured in disease and death tolls, does not reflect different evolutionary mechanisms. The generation of novel genotypes and phenotypes is the basis of geographical spread, ecological shifts and the origin of new taxa of disease vectors and diseases, just as it is the starting point for the same evolutionary and ecological outcomes in other natural and domesticated populations.
The Anopheles gambiae species complex includes the major malarial vectors, An. gambiae and An. arabiensis. It also encompasses taxa, such as An. melas, An. merus, An. bwambae and An. quadriannulatus, which are not considered to be major routes for malaria transmission (Besansky et al. 2003). Significant genetic discontinuities, indicating a level of subspecific reproductive isolation, have been detected within the widespread An. gambiae (Gentile et al. 2002). Further support for the presence of reproductive barriers within this species comes from the co-occurrence of different genotypes at the rDNA cistron. The variation in rDNA nucleotide sequence was discovered within both the internal transcribed spacer (ITS) and intergenic spacer (IGS) regions (Gentile et al. 2002). The detection of (i) no heterozygotes for the different variants and (ii) nearly total linkage disequilibrium for the pairs of ITS and IGS genotypes defined as ‘S/type I’ and ‘M/type II’ indicated that these races were partially, reproductively isolated (Gentile et al. 2002). In contrast, other regions of the genome demonstrated no such genetic discontinuities (Wang et al. 2001; Gentile et al. 2002). These contrasting patterns indicate that the S and M genomes are ‘semipermeable’ (Key 1968), with some regions recombining at a high frequency and others, like the rDNA cistron, recombining rarely.
The concept of semipermeability can also be applied across the An. gambiae species complex. For example, two earlier studies of the genetic architecture of the sibling species, An. gambiae and An. arabiensis, reported limited (Lanzaro et al. 1998) and extensive (Besansky et al. 1997) introgression of nuclear and mitochondrial genetic markers, respectively (Fig. 4). More recent studies have confirmed the observation of ongoing genetic exchange between these two species (Wang et al. 2001; Besansky et al. 2003). With regard to the current topic, the most important hypothesis was stated in the following manner by Besansky et al. (2003): ‘The proposed acquisition by An. gambiae of sequences from the more arid-adapted An. arabiensis may have contributed to the spread and ecological dominance of this malaria vector.’ Like, for example, landraces of maize and Australian fruit flies, the increase in ecological amplitude and the subsequent spread of a major disease vector may have been caused by introgressive hybridization.
Figure 4. Relationships among 59 mtDNA haplotypes identified from the ND5 genes of Anopheles gambiae and An. arabiensis. The white and black circles indicate haplotypes that are found only in An. gambiae or An. arabiensis, respectively. Hatched circles indicate haplotypes that are found in both species. The size of each circle is proportional to its frequency, and the connections between circles reflect one nucleotide difference. A ‘0’ along a branch indicates a missing variant (from Besansky et al. 1997).
Download figure to PowerPoint
Natural hybridization and the origin of human pathogens
Trypanosoma cruzi is the protozoan parasite that causes Chagas’ disease, otherwise known as American trypanosomiasis. The Centers for Disease Control and Prevention (CDC website 2003) estimate that 16–18 million people are infected with T. cruzi and that each year 50 000 will die from the infection. It has long been recognized that this protozoan species is both genetically and phenotypically highly variable (Machado & Ayala 2001). Using an analysis of sequence variation at unlinked genes, Machado & Ayala (2001) were able to test whether this variability was due to natural hybridization. They concluded that natural hybridization had affected this species complex. Indeed, the important evolutionary role played by natural hybridization in the T. cruzi lineage is summarized by the following quote: ‘The results provide evidence of hybridization between strains from two divergent groups of T. cruzi, demonstrate mitochondrial introgression across distantly related lineages, and reveal genetic exchange among closely related strains’ (Machado & Ayala 2001). The basis of the genetic and phenotypic (e.g. growth rate, pathogenicity, infectivity, drug susceptibility) variability within T. cruzi appears to be recombination following hybridization between divergent lineages.
Cryptococcus neoformans is a human fungal pathogen that causes meningitis in both immunocompetent and immunocompromised individuals (Cogliati et al. 2001). Between 0.2 and 0.9 infections per 100 000 persons occur in the general population and between 2 and 4 infections per 1000 occur among persons with AIDS. The overall mortality rate is ~12% (CDC website 2003). As with T. cruzi, C. neoformans demonstrates a high level of genetic variability (Cogliati et al. 2001). Though normally haploid, diploid isolates have been identified from the recognized varieties of C. neoformans (Cogliati et al. 2001). The origin of these diploid recombinants from crosses between the various haploid serotypes was thought to be rare. However, an analysis of sequence variation (Cogliati et al. 2001) demonstrated that these derivatives of hybridization were actually quite common. Forty-nine diploid strains were identified from a sample of 133 isolates gathered from Italian patients. These findings indicated that, like T. cruzi, evolution of the C. neoformans complex had been impacted greatly by the transfer of genetic material via natural hybridization. Such is also the case for the evolution of viral pathogens.
Viral lineages play an enormous role in human diseases. For example, viruses have caused some of the most catastrophic pandemics. Many diseases that emaciate, and thus make humans susceptible to death by secondary causes, are also of viral origin. The human rotavirus is an example of this latter class of pathogen. Human rotaviruses have the distinction of being the most common cause of severe diarrhoea in infants and young children, causing an estimated 600 000 deaths per year (Laird et al. 2003; CDC website 2003). These viruses are widely distributed among animal species, however, prior to a DNA sequence analysis by Laird et al. (2003) the strains were thought to be restricted to a single species, with no transmission from other animal species to humans. In contrast to this paradigm, Laird et al. (2003) detected not one, but two apparent recombination events. These cases of hybridization resulted in the origin of two diverse rotavirus lineages found in infected humans. The two strains have combinations of human and canine and human and porcine DNA segments, respectively (Laird et al. 2003). It is apparent that not only is transmission possible between the different animal species, but variants of this highly pathogenic viral assemblage have arisen through subsequent recombination.
The 1918 ‘Spanish flu’ is one of the best known, and most lethal, human viral pandemics. Significantly, Gibbs et al. (2001) have proposed a hybrid origin for the influenza variant that caused this major outbreak. Like some of the rotavirus strains, the ‘Spanish flu’ virus was hypothesized to have arisen from hybridization between human and swine viral genomes. In contrast, Basler et al. (2001) proposed that the virus causing the 1918 pandemic was not of hybrid origin, but was instead introduced from an avian source into both swine and humans shortly before 1918. The uncertainty concerning the hybrid origin of the 1918 influenza outbreak does not extend to those epidemics occurring in 1957 and 1968. These two pandemics also killed millions, but were definitively found to be of hybrid origin. In both cases, recombination between avian and human influenza strains gave rise to these lethal pathogens (see Webster 2001 for a discussion). These findings suggest that the 1957 and 1968 outbreaks were due to hybridization that brought together genes from divergent lineages thus causing their increased pathogenicity to humans.
Our final example of human pathogens formed from hybridization and recombination comes from the bacterial complex, Haemophilus influenzae. This bacterial species is one of the most common microorganisms in the upper respiratory tract of humans (Kroll et al. 1998). It is also normally benign, but sometimes invades leading to serious illnesses most often in children. One of the diseases for which this species has been implicated is Brazilian purpuric fever. The name of this disease reflects its original detection within a Brazilian population of infected children, 70% of who died from its effects. However, the description of the clinical features of this novel disease, i.e. meningococcal sepsis, suggested initially the involvement of a meningococcus species (Kroll et al. 1998). Surprisingly, the disease was found instead to be the result of infection by H. influenzae biogroup aegyptius, an organism that had previously been known to cause nothing more serious than conjunctivitis (Kroll et al. 1998). This novel H. influenzae had thus somehow acquired the meningococcal phenotype. Kroll et al. (1998) hypothesized that the evolution of this novel phenotype was due to the transfer of virulence genes from Neisseria meningitidis into H. influenzae (Fig. 5). Smoot et al. (2002) subsequently found evidence for such hybridization and recombination between these different genera. The increase in pathogenicity of H. influenzae is conceptually similar to the increase in ecological amplitude of, for example, maize landraces; both increases are thought to reflect the acquisition of traits from related organisms allowing the invasion of additional adaptive space. However, unlike the positive benefits derived from the maize introgression events, the derivation of pathogenic H. influenzae biogroup aegyptius illustrate the origin through hybridization of a ‘… clone with the phenotype of rapidly fatal invasive infection …’ (Kroll et al. 1998).
Figure 5. Gene trees derived from (A) the protein and (B) 16S RNA sequences of the bacterial species Haemophilus influenzae (H), Neisseria meningitidis (N) and Escherichia coli (E). Numbers along the branches indicate the number of mutational steps between the terminal taxa (from Kroll et al. 1998).
Download figure to PowerPoint