**3. Results**

Although the cross-kingdom infective capacity of *Fusarium* has been hypothesized, there is no evidence of any sequential human–plant–human tissue reinfection. In order to better understand the broad infectious properties of *Fusarium* spp., the present study assessed the ability of *Fusarium* isolates from human keratomycosis to infect monocot (maize) and dicot (*Arabidopsis*) plants, as well as human tissue in the form of nails. First, the *Fusarium* genotypes were isolated from 15 fungal keratitis samples, initially taken from patients at the "El Hospital para Prevenir la Ceguera en México, Luis Sánchez Bulnes" between January 2013 and August 2016. Each sample was taken from the eye of a non-immunologically compromised patient. Patients came from nine Mexican states and had diverse occupations. Patient age was not restricted to any specific range, but 87.5% were middle-aged adults or older (Figure S1).

In the first stage, we evaluated the infective capacity of *Fusarium* isolates from patients with keratitis to infect plants using two assays. (1) The first assay consisted of the inoculated tissue of detached leaves and root maize seedlings, placed in wet chamber Petri dishes. (2) In the second assay, whole maize and *Arabidopsis* seedlings were inoculated in vitro while still alive. *Fusarium verticillioides*, which infects corn plants was used as a positive control [23,24]. In both experiments, *Fusarium* conidia germinated, colonized the detached tissues, and deterred the growth of in-vivo seedlings (Figure 3). These results demonstrate that *Fusarium* spp. from keratitis patients conserve their infective capacity during the cross-kingdom reinfection jump from humans to plants.

In the second stage, we evaluated the capacity of *Fusarium* strains to back-infect human tissue. In this experiment, the capacity of the isolates to cause onychomycosis was assessed, using human nail samples (Figure 4). Observations suggested that every strain that infected maize and *Arabidopsis* was also able to back-infect human nails. After 4 days of inoculation, *Fusarium* was well established (Figure 4A) and exudates were observed (arrowheads, Figure 4A). Due to the rough surface of the nails, *Fusarium* was able to colonize and use the keratin in them as a growth substrate (arrow, Figure 4B). These results, together with the literature [14], indicate that *Fusarium* might cause onychomycosis in humans. Even though *Fusarium* has been sampled and well-studied in immune-compromised patients [12], the species found in this study were obtained from non-immune-compromised patients, suggesting that the pathogen reached the patients' eyes by ocular injury.

**Figure 3.** *Fusarium* spp. isolated from human keratitis conserve their plant infective capacity. Detached leaf assay: (**A**) control non-infected leaf; (**B**–**D**) leaves infected by isolated *Fusarium* strains *Fv, Fs*3, and *Fo*14, respectively. Detached root assay: (**E**) control non-infected root; (**F**–**H**) roots infected by isolated *Fusarium* strains *Fv*, *Fs*3, and *Fo*14, respectively. In both detached tissues, leaves and roots were imaged 12 days after infection. Seedling assay: (**I**) control non-infected seedling; (**J**–**L**) seedlings from free-infected seeds inoculated with isolated *Fusarium* strains *Fv*, *Fs*3, and *Fo*14, respectively. Seedlings were imaged 5 days after infection. Samples in the detached tissue and seedling growth assays were under sterile in vitro conditions. Arrows show the sites with abundance fungi growth or damaged tissue. *Fv*: *Fusarium verticillioides*; *Fs*3: *Fusarium solani* isolate no. 3; *Fo*14: *Fusarium oxysporum* isolate no. 14.

**Figure 4.** *Fusarium* isolates are a potential causative agen<sup>t</sup> of onychomycosis in human tissue. (**A**) Macroscopic (upper panels) and microscopic (lower panels) imaging of human nails inoculated with *Fv*, *Fs*3, or *Fo*14. (**B**) Colonization of a human nail (which emits endogenous red fluorescence) by *Fusarium Fs*3 isolate (green fluorescence). WGA Alexa-488 was used to detect fungal hyphae in the colonized nail (see 'Merge' image). All photographs (**A**) and scans (**B**) were taken 4 days after inoculation in a wet chamber. *Fv*: *Fusarium verticillioides*; *Fs*3: *Fusarium solani* isolate no. 3; *Fo*14: *Fusarium oxysporum* isolate no. 14.

This study's findings show that *Fusarium*-infective agents may alter human health, and that their infective capacity to colonize human tissue and plants and to back-infect human tissue (nails) is intact.
