The little caterpillar that could – Tobacco Hornworm (Manduca sexta) caterpillars as a novel host model for the study of fungal virulence and drug efficacy

Pathogenic yeast species can cause life-threatening infections in humans. The two leading yeast pathogens, Candida albicans and Cryptococcus neoformans, cause systemic infections in >1.4 million patients world-wide with mortality rates approaching 75%. It is thus imperative to study fungal virulence mechanisms, stress response pathways, and the efficacy of antifungal drugs. This is commonly done using mammalian models. To address ethical and practical concerns, invertebrate models, such as wax moth larvae, nematodes, or flies, have been introduced over the last two decades. To address short-comings in existing invertebrate host models, we developed fifth instar caterpillars of the Tobacco Hornworm moth Manduca sexta as a novel host model for the study of fungal virulence and drug efficacy. These caterpillars can be raised at standardised conditions, maintained at 37°C, can be injected with defined amounts of yeast cells, and are susceptible to the most threatening yeast pathogens, including C. albicans, C. neoformans, C. auris, and C. glabrata. Infected caterpillars can be rescued by treatment with commonly deployed antifungal drugs and importantly, fungal burden can be assessed daily throughout the course of infection in a single caterpillar’s faeces and hemolymph. Notably, these animals are large enough so that weight provides a reliable and reproducible measure of fungal virulence. This model combines a suite of parameters that recommend it for the study of fungal virulence.


85
In an effort to reduce the usage of mammals as model hosts, alternative invertebrate 86 models have been developed and used in fungal virulence research over the past two decades. 87 The most commonly employed invertebrate species include the nematode Caenorhabditis 88 elegans, the fly Drosophila melanogaster, and larvae of the Greater Wax moth Galleria 89 mellonella. All three species can be easily maintained in the laboratory at a much lower cost 90 than mice or rabbits and have been successfully used for the study of diverse yeast pathogens, 91 such as C. neoformans 15,16 , C. albicans [16][17][18] , C. parapsilosis [18][19][20] , C. glabrata 19,21 . Of note, 92 invertebrate models differ in their applicability and the best suitable model should be 93 5 mice virulence studies. Indeed, the caterpillars grow at 37˚C while maintaining susceptibility, 126 specific C. albicans mutants are just as attenuated in their virulence in M. sexta as they are in 127 mice, and notably, M. sexta are susceptible to the leading yeast pathogens C. albicans, C. 128 neoformans, as well as the emerging C. auris. To expand M. sexta's applicability as a host 129 model, we developed an infection protocol that permits screening of fungal burden 130 throughout the course of infection in a single animal and uses weight as a proxy measure for 131 virulence in addition to survival. M. sexta can furthermore be used to test efficacy of common 132 antifungal drugs. Our results define M. sexta characteristics that recommend the caterpillars 133 as a non-mammalian host model for the study of fungal virulence.

135
Origin of the Bath colony of Manduca sexta 136 The colony has been in continuous culture since 1978 without the addition of animals 137 from elsewhere. Bath's genetic stock was derived from animals from the Truman-Riddiford were maintained in 125 ml disposable cups (Sarstedt Ltd., Cat. No. 75.1335), on a wheat 144 germ-based diet (Appendix 1), at a constant temperature of 25˚C with 50% humidity, and 12 145 hours of light and dark cycles. Three days prior to infections with fungi, animals were shifted 146 to a formaldehyde-free diet as the compound is toxic to non-methylotrophic yeast.

147
For infection assays, yeasts were grown overnight in 50 ml YPD (1% yeast extract, 148 2% peptone, 2% dextrose) and cells harvested by centrifugation for 3 minutes at 3,000 rpm. 149 The cell pellet was washed twice with 1x phosphate buffered saline (PBS) and suspended in 5 150 ml 1x PBS. Cells were counted and numbers adjusted as indicated. C. albicans YSD85 151 (Table 1) cells were heat-inactivated by incubation at 65˚C for 20 minutes. For long-term 152 storage, yeast isolates were maintained at -80˚C in 25% glycerol.

153
Yeast infections and measurements of fungal burden and drug efficacy 154 6 100 µl of washed and number-adjusted yeast suspension were injected into each 155 caterpillar with a 30G1/2" needle (BD Microlance) and a 1 ml NORM-JECT syringe.

156
Animals were injected through their distal left proleg. Following injection, each animal's 157 weight was recorded. Animals were scored for survival and weight once daily for three to 158 four days post infection. During the course of the experiment, animals were kept on a 12 hour 159 light and dark cycle at the temperature indicated and on their regular diet.

160
To measure fungal burden in caterpillar faeces and hemolymph, six animals were 161 injected with either 1x PBS or 10 6 cells of the wild type YSD89 or the hog1 mutant strain 162 YSD883 and kept at 37˚C. On day 1, two animals were selected from each group. These 163 animals were weighted and their hemolymph and faeces collected daily throughout the course 164 of infection. To collect hemolymph, animals were first kept on ice for 15 minutes. The 'horn' 165 was then surface sterilised with 70% ethanol and its top 1-2 mm clipped with a pair of micro 166 scissors. Hemolymph was collected in a pre-chilled 1.5 ml Eppendorf tube and cooled 167 immediately to reduce polymerisation and melanisation. One faecal pellet was collected daily 168 with sterile forceps, weighted and suspended in 500 µl 1x PBS. Prior to diluting, the mixture 169 was thoroughly vortexed for 10 seconds, and centrifuged for 5 seconds using a table top 170 centrifuge to separate faecal matter. To quantify fungal burden, hemolymph and faecal 171 samples were plated either directly onto YPD-agar with Kanamycin 50 µg/ml or in ten-fold 172 serial dilutions. Agar plates were incubated at 30˚C for 48 hours and colonies counted.

173
To assess the efficacy of commonly used antifungal drugs, animals were infected with 174 10 7 cells of YSD85 or PBS and treated with increasing doses of fluconazole and caspofungin 175 (Sigma Aldrich, Inc.) as indicated. Drugs were injected with an ethanol-sterilized Hamilton 176 syringe in a total volume of 10 µl per animal, 30 minutes post-infection. Caterpillars were 177 weighted and scored for survival on the day of injection and the following three days.

Statistical analyses
179 Survival plots were made using the survminer R package (https://CRAN.R-180 project.org/package=survminer), and differences were evaluated using the Kaplan-Meier 181 method. Weight and fungal burden were plotted using ggplot2 41 and weight differences were 182 evaluated using linear models with day post-inoculation and the interaction between 183 treatment and dpi as fixed effects and individual as a random effect using nlme 184 7 (https://CRAN.R-project.org/package=nlme). All analyses were done using RStudio version 185 1.1.442.

187
We first aimed to determine if M. sexta fifth instar caterpillars, reared and maintained 188 at standard conditions (Fig. 1a), are susceptible to Candida albicans. To do so, groups of ten 189 animals were infected with increasing doses of the widely used C. albicans laboratory strains 190 SC5314 and SN95 42 . Animals were scored daily for survival for three consecutive days while 191 being maintained at 25˚C. Dead animals differ from live ones in that their bodies go limp and 192 turn grey-green in colour, which is in stark contrast to the vivid turquoise of live animals 193 (Fig. 1b). Indeed, caterpillars that were infected with C. albicans succumbed to the yeast in a 194 dose-dependent manner. Both C. albicans strains killed M. sexta caterpillars efficiently at 195 inocula of 10 6 or 10 7 cells per animal (Fig. 1c) is as essential for virulence in caterpillars as it is in mice. Cka2 is not required to establish 201 systemic infections in mammals but is in caterpillars. Ahr1, while required for virulence in 202 mammals, appears to be dispensable for virulence in caterpillars (Fig. 1d).

203
Given the importance of temperature for fungal virulence, we aimed to determine if 204 M. sexta retained their susceptibility to C. albicans at human body temperature of 37˚C.

205
Temperature itself does not affect caterpillar survival or development ( Fig. S1) but animals 206 are ten times more susceptible to infections with C. albicans at 37˚C than they are at 25˚C 207 ( Fig. 2a). At 37˚C, 10 6 C. albicans cells per animal lead to 100% mortality on day 4, while 208 10 7 cells are required for the same outcome at 25˚C (Fig. 1b). To exclude the possibility that 209 mortality is due to starvation rather than the outcome of a host-pathogen interaction, we 210 infected caterpillars with live and heat-killed C. albicans wild-type cells at 37˚. Only live 211 cells, but not heat-killed Candida cells, kill caterpillars suggesting that killing is not due to   (Table 2).   The authors report no interest of conflict.         the mutant strain gain significantly more weight than those infected with the wild type strain.

621
Infection with the complemented strain results in a comparable lack of weight gain.