Date Published: November 17, 2016
Publisher: Public Library of Science
Author(s): David Triest, Marijke Hendrickx, Deborah A. Hogan.
Banana is one of the most important tropical crops and is affected by several fungal diseases, such as crown rot postharvest disease . Crown rot is responsible for significant losses in banana fruits [1, 2]. Predominantly, Colletotrichum musae and Fusarium spp. are its causative agents [1, 2]. Inoculum sources include mainly infected flowers but also decaying leaves, and fungal transfer can occur from banana stalks onto the crown surface during the cutting of banana bunches (knife-induced) as well as when the bunches are cleaned in contaminated water (Fig 1) . Fungal infection starts at harvest, and the first symptoms of crown rot appear only after packaging and shipping from producing countries to consuming countries [1, 2]. Crown rot begins with a mycelium development on the crown surface, followed by an internal development . This internal development can, subsequently, affect the peduncle and the whole fruit, leading to softening and blackening of the fruit tissue . Postharvest fungicidal treatments are applied to control crown rot disease, though severely affected banana fruits are still found in consumer markets . Moreover, onset and spreading of the disease is unpredictable and can also induce early ripening of banana fruits during transport .
Recent studies showed that the fungal species F. musae is frequently found associated with banana crown rot [1, 3]. The species was installed in 2011 (according to multilocus phylogeny and mating experiments) as a separate, sister species from F. verticillioides sensu stricto, which is also frequently found associated with banana crown rot [1, 3, 4]. Originally, F. musae was known as a distinct, banana-infecting population within F. verticillioides, and both species are practically impossible to distinguish morphologically . Whereas F. verticillioides has a broad plant host specificity (maize, rice, banana, etc.), F. musae seems restricted in its plant pathogenicity and has, until now, only been recovered from banana fruits [4, 5]. To date, F. musae has been isolated from banana fruits coming from several producing countries in Latin America (Mexico, Panama, Ecuador, etc.), the Canary Islands, and the Philippines, but not from banana-producing countries in Africa [1, 4, 5]. Moreover, it has been observed that F. musae strains have a significantly greater ability to cause infection on banana fruits than F. verticillioides strains . Another difference with typical F. verticillioides strains (i.e., isolated from maize) is that F. musae strains are unable to produce the mycotoxin fumonisin because of the absence of the fumonisin biosynthesis gene cluster .
Because of the recent installment of F. musae as a separate species (in 2011) and, consequently, the fact that several strains previously identified as F. verticillioides were actually shown to be F. musae, its epidemiology is not yet fully elucidated. The latter justifies the need for retrospective studies that reidentify F. verticillioides strains. One such retrospective study showed that five strains collected in the period 2001–2008 and morphologically identified as F. verticillioides appeared to be F. musae, according to multilocus phylogeny . Of interest is that all five strains were clinical isolates, and four were isolated from immune-suppressed patients .
At least 70 species of the soil-borne fungal genus Fusarium are known to be associated with opportunistic human disease, and most of them are members of a species complex . The species complex most frequently found associated with human infection is the F. solani species complex (between 40% and 60% of the reported cases), followed by the F. oxysporum species complex and the F. fujikuroi species complex (each accounting for approximately 20% of the cases) . Both F. musae and F. verticillioides are members of the F. fujikuroi species complex. Fusarium spp. are also well-known plant pathogens that are able to infect a diverse range of plant hosts, and several formae speciales have been defined. Increasing evidence indicates the presence of host-specific virulence determinants . Despite the availability of whole genome sequences of some important plant pathogenic Fusarium spp., knowledge about Fusarium virulence determinants remains fragmentary and largely undefined, except for the secreted in xylem (SIX) effectors in F. oxysporum (i.e., virulence factors secreted by F. oxysporum in the plant vascular system) and several Fusarium mycotoxin encoding regions . Gene expression data from F. oxysporum f. sp. cubense causing banana wilt disease has highlighted an important role for proteins putatively involved in root attachment, cell degradation, detoxification, transport, secondary metabolite biosynthesis, signal transduction, and conidial germination, which is crucial for spreading of the disease in plants as well as for transfer to humans [15, 16]. In addition, comparative genome analyses have revealed that F. solani f. sp. pisi and F. oxysporum f. sp. lycopersici have, in addition to their core genome, an adaptive genome with dispensable chromosomes that are enriched in host-specific genes towards pathogenicity of pea and tomato, respectively. [14, 17, 18]. Also, F. verticillioides appears to exhibit chromosomes with rapidly evolving genes encoding potential virulence determinants . Moreover, Ma et al.  showed that horizontal gene transfer of the pathogenicity chromosomes can convert a non–plant pathogenic F. oxysporum strain (initially having only a core genome) into a tomato pathogen. Whether Fusarium spp. require an adaptive genome to cause a human infection or whether a core genome is already sufficient remains to be elucidated.
Since F. musae has only been found on banana fruits, unlike F. verticillioides, its mode of transfer to cause an opportunistic human infection seems clear and involves an important role for banana crown rot postharvest disease. Although a direct link between a human pathogenic F. musae infection and a F. musae-infected banana fruit as the source has not yet been established, two important observations are made: (i) The only known environmental habitat of F. musae is the banana fruit; (ii) All currently known cases of human infection with F. musae as the causative agent involve patients hospitalized in non–banana-producing countries (the US and European countries). As such, it is hypothesized that marketed banana fruits contaminated with F. musae but not yet visibly affected by crown rot postharvest disease most likely lead to an opportunistic human pathogenic F. musae infection after the susceptible human host is brought into contact with it (Fig 1). However, two alternative hypotheses can be proposed. A first is that the currently known cases of F. musae-infected patients acquired their infection after travelling to a banana-producing country, where they came into contact with F. musae-contaminated banana material or cleaning water. The second alternative hypothesis assumes that the habitat and distribution of F. musae is not as limited as currently described, and F. musae is also present on currently unknown plant or environmental substrates other than banana.
The occurrence of F. musae as a human pathogen in non–banana-producing countries and its plant host specificity for banana suggests that banana crown rot postharvest disease may be a potential public health concern of importance. The establishment of a direct link between a human infection and a F. musae-infected banana fruit as the source is one of the major future perspectives. Difficulties for future studies, however, will be the fact that transfer to humans can occur at multiple points (the travel history of the patient will need to be evaluated) and that the potential banana fruit source will often no longer be available for analysis at the time a patient is diagnosed with a F. musae infection. However, when a potential banana fruit source is still available, F. musae may be isolated and compared with the clinical isolate by molecular typing methods or whole genome sequencing. Also, it needs to be investigated whether other Fusarium spp. associated with banana crown rot postharvest disease, such as F. verticillioides, are transferred in the same way to cause a human pathogenic infection. Further (retrospective) analyses and clinical surveys, using multilocus sequencing for identification, will need to be performed to fully elucidate the epidemiology of F. musae infections in banana fruits as well as in humans. Moreover, it needs to be investigated whether F. musae can be isolated from other plant or environmental substrates, such as hospital water–related systems or air samples. Nevertheless, improvement of postharvest fungicidal treatments to prevent banana crown rot postharvest disease, better control measurements, and better process hygiene criteria for the processing of banana fruits are recommended. In addition, resistant banana cultivars may need to be searched.