Vacuoles and peroxisomes are involved in Aspergillus fumigatus gliotoxin production and self-protection

Aspergillus fumigatus is a saprophytic fungus that can cause a variety of human diseases known as aspergillosis. Mycotoxin gliotoxin (GT) production is important for its virulence and must be tightly regulated to avoid excess production and toxicity to the fungus. GT self-protection by GliT oxidoreductase and GtmA methyltransferase activities is related to the subcellular localization of these enzymes and how GT can be sequestered from the cytoplasm to avoid increased cell damage. Here, we show that GliT:GFP and GtmA:GFP are localized in the cytoplasm and in vacuoles during GT production. Peroxisomes are also required for proper GT production and self-defense. The Mitogen-Activated Protein (MAP) kinase MpkA is essential for GT production and self-protection, interacts physically with GliT and GtmA and it is necessary for their regulation and subsequent presence in the vacuoles. Our work emphasizes the importance of dynamic compartmentalization of cellular events for GT production and self-defense.


Introduction
An important aspect of the A. fumigatus GT production and self-protection 112 by GliT and GtmA is related to the cellular sub-location of these enzymes and 113 how GT can be temporarily sequestered from the cytoplasm to avoid increased 114 cell damage. It has been speculated that reduced dithiol gliotoxin (dtGT) may be 115 sequestered into intracellular vesicles where it is converted to the oxidized form, 116 by an unidentified activity, prior to release from the cell by an exocytotic 117 mechanism complementary to GliA-mediated efflux 20 . Therefore, the aim of this 118 work is to elucidate GliT and GtmA subcellular localization under GT production 119 and self-protection. We demonstrate that both proteins are localized in the  partial biotin deficiency phenotype ( Figure 4D), but this partial auxotrophy is not 237 observed in A. fumigatus ΔpexE mutant ( Figure 4A).

238
A. fumigatus ΔpexE mutants have a reduced GT (about 7.5-to 15-fold 239 reduction; Figure 4G, left graph) and bmGT production (about 15-to 3-fold 240 reduction; Figure 4G, right graph) while surprisingly ΔpexG mutants show a 241 dramatic increase in GT production (about 25-to 50-fold compared to the wild-242 strain, Figure 4H, left graph) and a bmGT production comparable with wild-type 243 ( Figure 4H, right panel). Not only the SMs production is altered in the pex mutants 244 but also several other SMs, such as brevianamide F, fumigaclavine C, pseurotin 245 A, and fumagillin have increased or decreased production in these mutants when 246 compared to the wild-type strain ( Table 1)  These results strongly indicate that PexE and PexG are important for A. 258 fumigatus GT production and self-protection, and virulence, and also affect the 259 production of other SMs.  (Table 2).

291
Taken together these results suggest that GliT and GtmA are interacting 292 with several protein kinases and phosphatases and enzymes involved in the 293 glutathione metabolism and oxidative stress response during GT-production.

295
MAP kinase MpkA is important for GT production and self-defense. 296 Considering that A. fumigatus MpkA has previously been described as being functionality of these putative phosphorylation sites remains to be investigated.

323
These data strongly support the view that GliT and MpkA associate in vivo during

324
GT production and GT self-protection while GtmA associates only during GT self-325 protection.

326
The ΔmpkA has a strong growth defect and its complementation by mpkA + 327 restores this phenotype to the wild-type growth ( Figure 6A). The ΔmpkA mutant 328 does not produce either GT or bmGT ( Figure 6B). Since other MAP kinases (like 329 SakA and MpkC, Table 2) were shown to be interacting either with GliT and/or 330 GtmA, we also investigated the GT production in the corresponding null mutants, 331 including another MAPK mutant, ΔmpkB, and the double mutant ΔsakA ΔmpkC.

332
All the mutants showed comparable GT and bmGT production with the wild-type, decreased mRNA accumulation at 72 hours GT production but no differences to 355 the wild-type during GT self-protection ( Figure 6G).

356
Taken together these results indicate that MpkA is essential for GT  Table S2) were grown on solid MM 367 and MM+30 µg/mL of GT. In the primary screening we identified 11 PK mutants 368 (5 and 6 with less growth and more growth, respectively) ( Figure 7A). Further  wild-type during GT production ( Figure 7D) but increased levels during GT 388 protection (about 3-fold at 1 h exposure to 5 µg/mL GT, Figure 7E). The ΔslnA 389 mutant has higher gliT and gtmA mRNA accumulation upon GT production than 390 the wild-type ( Figure 7F). Upon GT self-protection, gliT mRNA accumulation is 391 lower and higher at 0.5 and 1 h exposure to 5 µ/mL GT than the wild-type strain, 392 respectively ( Figure 7F); gtmA mRNA accumulation is also lower in the ΔslnA 393 mutant upon GT self-protection conditions than the wild-type ( Figure 7F). slnA null mutant has lower GT production than the wild-type. These results

517
indicate that SlnA is one of the sensors that activate GT self-protection via MpkA 518 phosphorylation. Since GT production is intimately related to GT self-production 519 it is possible the reduction of GT production in the slnA null mutant is due to a 520 reduction of GliT and GtmA activities, and consequently low GT accumulation.