Paracoccin overexpression in Paracoccidioides brasiliensis reveals the influence of chitin hydrolysis on fungal virulence and host immune response

Paracoccidioides brasiliensis and P. lutzii, etiological agents of paracoccidioidomycosis (PCM), develop as mycelia at 25-30 °C and as yeast at 35-37 °C. Only a few Paracoccidioides spp. proteins are well characterized. Thus, we studied paracoccin (PCN) from P. brasiliensis, its role in the fungus biology, and its relationship with the host innate immune cells. Cloning and heterologous expression analysis revealed its lectin, enzymatic, and immunomodulatory properties. Recently, we employed a system based on Agrobacterium tumefaciens-mediated transformation to manipulate P. brasiliensis yeast genes to obtain clones knocked-down for PCN, which after all, are unable to transit from yeast to mycelium forms, causing a mild pulmonary disease. Herein, we generate P. brasiliensis overexpressing PCN (ov-PCN). To date, it was not explored the overexpressing of endogenous components in Paracoccidioides spp. Therefore, we investigate the role of PCN in fungal biology and pathogenesis. Augmented levels of PCN mRNA and protein, and N-acetylglucosaminidase activity confirmed PCN overexpression in ov-PCN of P. brasiliensis yeasts. Interestingly, PCN overexpression did not affect the yeasts’ growth or viability and favored cell separation. The ov-PCN clones transitioned faster to the mycelium form than the wt-PCN yeasts. Concerning infection, while most of mice infected with the wt-yeasts (90%) survive at least until the 70th day, all mice infected with ov-PCN yeasts were already died at the 35th day post-infection. In vitro assays showed that ov-PCN were more susceptible to phagocytosis by macrophages. Finally, it was verified that the chitin particles isolated from the ov-PCN cells were smaller than those obtained from the wt-PCN yeasts. Macrophages stimulated with the chitin isolated from ov-PCN produce IL-10, whereas the particles with a wider size range harvested from wt-PCN yeasts induced TNF-α and IL-1β secretion. The anti-inflammatory microenvironment from macrophage stimulation with small chitin particles hampers the development of a protective immune response against the fungus. We postulated that the high grade of chitin cleavage, as the results of augmented PCN expression, favors pathogenesis following P. brasiliensis infection. Thus, PCN is a relevant virulence fungal factor. AUTHOR SUMMARY Paracoccidioides spp. are pathogenic fungi that cause paracoccidioidomycosis (PCM) in humans, the main deep mycosis of Latin America. Recently, by knocking down the paracoccin gene, our group showed that this lectin is necessary for the morphological transition from yeast to hyphae, and that this decrease results in low P. brasiliensis virulence. Here, after overexpress PCN, we revealed the importance of the yeast chitin hydrolysis to the host response. Infection of mice with ov-PCN yeasts causes severe lung disease compared to moderate disease caused by wt-PCN yeasts. The release of smaller chitin particles was as a result of an accelerated chitin hydrolysis provided by ov-PCN yeasts. Interestingly, these smallest chitin particles are able to modulate host response by increasing IL-10 in the meantime that decrease TNF-α secretion, thus hampering Th1 immune response that is crucial in the fight against this fungi. These findings represent a significant advance in the knowledge about the role of PCN chitinase in P. brasiliensis.


76
Paracoccidioidomycosis (PCM) is a severe mycosis widespread in Latin America [1]. 77 The fungi that cause PCM belong to the genus Paracoccidioides, which include two species of thermo-dimorphic fungi developing as filaments at 25-30 °C, mainly in the soil, and assuming contributes for the fungal attachment to ECM/pneumocytes and fungal virulence [28], whereas Because the P. brasiliensis PCN-knocked down yeasts are less virulent than the wt-PCN yeasts [18], we hypothesized that overexpression of PCN could promote different 177 pathogenic profile of yeast in mice. The lungs of BALB/c mice, inoculated through the 178 intranasal route with 2×10 6 cells of ov-PCN or wt-PCN yeasts, were microscopically examined, 179 thirty days after infection, regarding the extension of the granulomatous lesions. In mice that 180 were infected with wt-PCN yeasts, small and circumscribed granulomas were focally 181 distributed in the pulmonary tissue, whereas in mice infected with ov-PCN yeasts, the 182 granulomas were large and coalescent, occupying an extended area of the lungs (Fig 3A). 183 Methenamine/silver-staining of the pulmonary sections revealed few and focally distributed 184 viable yeasts, more centrally localized in the small and compact granuloma of wt-PCN yeast-185 infected mice, while abundant viable yeasts were dispersed in all the area of coalescent 186 granulomatous lesions of the mice infected with ov-PCN yeasts ( Fig 3B). Morphometric 187 analysis of pulmonary tissue injury showed that the area occupied by lesions was 60% larger in 188 mice that were infected with ov-PCN yeasts than in wt-PCN yeast-infected mice ( Fig 3C). 189 Pulmonary CFU counting has quantitatively validated the results obtained by optical 190 microscopy; the number of colonies provided by the ov-PCN yeast-infected mice was at least 191 one order of magnitude higher than the one obtained from wt-PCN yeast-infected mice (Fig 192 3D). In our first analysis of the mechanisms accounting for the different profile of infection we 193 considered that the ability of P. brasiliensis to cause disease largely depends on the yeasts' 194 resistance to the defense mechanisms of the host phagocytes [33]. 195 As such, we assayed in vitro the sensitivity of ov-PCN and wt-PCN yeasts to RAW 196 264.7 murine macrophage effector functions. Phagocytosis was examined 4 h after incubating 197 macrophages with ov-PCN or wt-PCN yeasts. Our data indicate that phagocytosis of  yeasts is more effective than that of wt-PCN yeasts (Fig 3E). The fungal killing by macrophages 199 was investigated by obtaining the lysate of macrophages that were incubated for 48 h with yeasts. We recovered a significantly higher CFU number from the cells infected with ov-PCN 201 yeasts than with wt-PCN yeasts because a higher number of yeasts were internalized by 202 macrophages infected with ov-PCN yeasts than wt-PCN ( Fig 3E). 203 Concerning the mice survival to the fungal infection, we observed that all animals  Fig 4A). With the aid of the ImageJ software, we could verify 215 that chitin detection was reduced by 40% in ov-PCN yeasts, in comparison to wt-PCN yeasts 216 ( Fig 4B). We also detected the chitin content of yeast cells by using a TexasRed conjugate to 217 Wheat Germ Agglutinin (WGA), a highly specific chitin-binding lectin [34]. The analysis by 218 flow cytometry also showed a significant reduction (by 30%) of the chitin content in ov-PCN 219 yeasts, compared to that detected in wt-PCN yeasts ( Fig 4C). Consistently the cell wall of wt-220 PCN yeasts, as examined by electron microscopy, was about 6-fold thicker than that of ov-PCN 221 yeasts (Fig 4D and 4E).

222
Then, the size of the chitin particles present in the supernatant of yeast cultures and 223 captured by immobilized WGA was analyzed by electron microscopy. The isolated chitin 224 particles had variable sizes ( Fig 4F). We detected only particles with less than 60 nm 2 in the material derived from the supernatants of ov-PCN yeasts. This low size range was also 226 prominent in the material obtained from the wt-PCN yeasts; nevertheless, this preparation 227 included a wider size distribution, which included particles with areas as large as 240 nm 2 (Fig   228   4F).

229
Finally, the obtained results consolidate the notion that PCN hydrolyzes chitin of the 230 P. brasiliensis cell wall, reducing its chitin content and cell wall thickness. In addition, the 231 process triggered by PCN promotes the release of very small chitin particles to the extracellular 232 milieu.

234
It was previously reported that the size of chitin fragments correlates with the particles' 235 property of stimulating macrophages to produce inflammatory or anti-inflammatory cytokines 236 [35][36][37][38][39][40][41]. Based on these studies, we examined whether the chitin fractions we captured from 237 the supernatants of ov-PCN (containing only small chitin fragments) or wt-PCN yeasts 238 (containing a large spectrum of small and larger chitin fragments) ( Fig 4F) could result in 239 distinct responses from murine bone marrow-derived macrophages (BMDMs). Isolated chitin 240 particles of wt-PCN and ov-PCN yeasts were assayed for the ability of inducing cytokine 241 production by BMDMs. Dose-response and time-course curves (supplementary data, S1A-F) 242 were drawn for TNF-α, IL-1β, and IL-10 cytokines, whose production was tested toward five 243 different chitin concentrations (from 2.5 to 100 µg/mL) during the periods of 24, 48, and 72 h.

244
The dose-response curves for the levels of TNF-α, IL-1β and IL-10, representing inflammatory 245 and anti-inflammatory cytokines, respectively, which shows that the two curves are inverted: 246 lower chitin concentrations correlate with higher IL-10 levels, whereas higher chitin 247 concentrations determine higher TNF-α production (Fig 5). The production of the pro-248 inflammatory cytokines TNF-α ( Fig 5A) and IL-1β (Fig 5B), measured in the supernatant of particles derived from wt-PCN than ov-PCN yeasts. On the other hand, the IL-10 levels 251 measured in the supernatant of BMDMs (Fig 5C), also harvested 48 h after stimulation, were 252 significantly higher when stimulated with chitin particles derived from ov-PCN than wt-PCN 253 yeasts.

254
In order to demonstrate the regulation of cytokines mediated by the chitin fragments 255 isolated from the supernatants of the ov-PCN and wt-PCN yeast cultures by BMDMs at 256 different time points, we used the concentration of 7.5 μg/mL of chitin (Fig 5D-F). Cytokine 257 induction by BMDMs exhibited increased levels of TNF-α and IL-1β from 24 h post-infection 258 by chitin particles isolated from wt-PCN yeasts than macrophages stimulated with ov-PCN 259 yeast-derived particles (Fig 5D and 5E). On the other hand, modulation by IL-10 was increased 260 from 48 h and was maintained over time in BMDMs stimulated with chitins derived from ov-261 PCN yeasts; macrophages stimulated with wt-PCN yeast-derived particles had basal levels 262 similar to the negative control ( Fig 5F). When we performed correlation analyses using the data 263 obtained in the dose-response experiment, we found that levels of IL-10 and TNF-α correlated 264 negatively with concentrations of 2.5 μg/mL and 50 μg/mL of chitin particles isolated from the 265 supernatant of ov-PCN and wt-PCN, respectively. From this, these results suggest that isolated 266 particles of the ov-PCN supernatant (small chitins) induce higher levels of IL-10 and lower 267 TNF-α than wt-PCN yeasts (Fig 5G), whereas particles isolated from the supernatant of wt-268 PCN yeasts (large chitins) induce lower levels of IL-10 and higher TNF than ov-PCN ( Fig 5H).

269
Taken together, these results demonstrate that all the chitin captured from the supernatant of    Table 4, further 315 supports the need of a fine tuning PCN expression.
The analyzed parameters were: detection of the PCN mRNA or protein; effects on the in vitro 319 yeast growth; effects on the transition from yeast to mycelium; sensitivity to killing by 320 macrophages; effects on the infection pathogenesis, and grade of virulence. The virulence was 321 inferred from the set of analyzed parameters.

322
The observation that fungal growth was not affected by overexpression of PCN shows 323 that the activity of this enzyme is needed for normal growth [22], but its increase has no surplus 324 effect. As such, our data further suggests that the PCN chitinase activity is required for chitin 325 hydrolyses allowing the separation of the budding of the daughter cells.

326
We found that overexpression of PCN had an acceleration effect on the transition from     Technologies, Camarillo, CA, USA) and gentamicin at 96 µg/mL (Gibco, Grand Island, USA).

556
The plates were incubated at 36 ºC for 7 days, after which the colonies were counted. Aldrich), supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine and 100 µg/mL 569 streptomycin/ampicillin. The fungal inoculum was prepared as described for the previous item.

570
To assess their phagocytosis, the macrophages were distributed (5×10 5 cells/well) in a 571 24-well microplate (Costar, Corning Inc., Corning, NY, USA) and incubated with yeast cells 572 (5×10 4 yeasts/well) for 4 h at 37 °C (yeast-to-macrophage ratio of 1:10) in a 5% CO 2 573 atmosphere, and the co-culture supernatant was discarded. The macrophage monolayers were lysed with ice-cold water, and the cell lysate was plated on BHI agar broth supplemented with 575 1% glucose, for 7 days at 37 °C.

576
To estimate the macrophage fungicidal activity, viable yeast cells were co-cultured 577 with macrophages at 37 °C in a 5% CO 2 atmosphere for 4 h. The culture supernatant was 578 discarded, and the cells were gently rinsed with PBS and incubated in RPMI (Sigma-Aldrich) 579 supplemented with 10% FBS (HyClone) at 37 °C in a 5% CO 2 atmosphere. After 48 h of 580 incubation, the culture supernatant was discarded, macrophages were lysed using ice-cold 581 water, and the cell lysate was serially diluted and cultured, for 7 days at 37 °C, in solid BHI  Ganiko, L., et al., Paracoccin, an N-acetyl-glucosamine-binding   integrated density (pixels) was analyzed by the ImageJ software. Bars depict the mean ± SEM 877 and were compared by Mann-Whitney´s test, ****p ˂ 0.0001, and ** p ˂0.0022. variance, followed by One-Way ANOVA test. ****р<0.0001, ***р<0.001, **р<0.01 and