Enhanced virulence of Histoplasma capsulatum through transfer and surface incorporation of glycans from Cryptococcus neoformans during co-infection

Cryptococcus neoformans (Cn) and Histoplasma capsulatum (Hc) co-exist in the environment and occasionally co-infect individuals, which can lead to severe disease/lethal outcomes. We investigated specific interactions between Cn-Hc to determine the impact of synchronous infection in virulence and disease. Co-infected mice had significantly higher mortality than infection with either species or acapsular Cn-Hc. Coating of Hc with cryptococcal glycans (Cn-gly) resulted in higher pulmonary fungal burden in co-infected animals relative to control. Co-cultivation or addition of Cn-gly resulted in enhanced pellicle formation with a hybrid polysaccharide matrix with higher reactivity to GXM mAbs. Transfer and incorporation of Cn polysaccharide onto Hc surface was time and temperature dependent. Cn-gly transfer altered the zeta potential of Hc and was associated with increased resistance to phagocytosis and killing by macrophages. Mice infected with Hc and subsequently injected with purified Cn-gly died significantly more rapidly than Hc alone infected, establishing the precedent that virulence factors from one fungus can enhance the virulence of unrelated species. These findings suggest a new mechanism of microbial interaction involving the transfer of virulence traits that translates into enhanced lethality during mixed fungal infections and highlights the importance of studying heterogeneous microbial populations in the setting of infection.


Results
Co-infection resulted in enhanced mouse mortality. We explored the possibility that Cn and Hc co-infection could worsen disease prognosis in mice. Co-infection was assessed with an equal mixed inoculum of Hc and encapsulated Cn H99 or unencapsulated Cn cap59 (5 × 10 6 of each fungus) and compared with monospecies infected animals (10 7 yeasts). Co-infection with Hc and Cn H99 resulted in higher mortality rates, with 100% death after 12 days, relative to mice infected with Hc and acapsular Cn cap59 (p = 0.0038) or monospecies inoculum of either Hc (**p = 0.0007), Cn H99 (**p = 0.007) or Cn cap59 (**p = 0.0014; Fig. 1a).
To confirm that animals were indeed co-infected and to determine fungal burdens in the scope of Cn PS importance, the colony forming units (CFU) were determined in lungs of deceased animals in the course of the survival experiments for Hc (Fig. 1b) and Cn (Fig. 1c). Both fungal species were recovered in similar proportions, indicating they colonize the lungs with similar efficacy and that they could interact in vivo.
Scientific RepoRts | 6:21765 | DOI: 10.1038/srep21765 However, animals from the Hc + Cn H99 co-infection group had higher burdens of Hc and Cn during earlier time-points than those in the Hc + Cn cap59 group (Fig. 1b,c, respectively), which correlated with the increased lethality of the Hc co-infection with the encapsulated Cn. Hc fungal burdens from Hc + Cn H99 group varied from 2.9 × 10 6 to 7.0 × 10 7 (median 2.6 × 10 7 ), while the Hc + Cn cap59 group ranged from 3.5 × 10 5 to 3.0 × 10 7 (median 1.35 × 10 7 ). For Cn fungal burdens, group Hc + Cn H99 ranged from 1.2 × 10 7 to 3.6 × 10 7 (median 1.7 × 10 7 ), while Hc + Cn cap59 group ranged from 1.0 × 10 5 to 2.8 × 10 7 (median 4.5 × 10 5 ). These results suggest that co-infection of Hc with Cn that efficiently releases PS leads to an increase in the virulence of Hc in vivo. Additionally, the monoinfection with Hc yeast cell alone resulted in fungal burdens that were ~25% lower than the average observed for the Hc + Cn H99 co-infected animals, further suggesting that virulence of Hc is enhanced in the presence of Cn. This conclusion is further justified based on the fact that half as many Hc yeast cells (5 × 10 6 ) were introduced to the co-infected animals compared to those receiving Hc alone (10 7 ).
Hc incorporates Cn glycans in vivo. The possibility that Hc could interact in vivo with Cn-glycans was explored initially by the evaluation of the transfer of Cn PS to the surface of Hc. Recovered yeasts from co-infected lungs were incubated with 2D10 mAbs to cryptococcal GXM 33 and an anti-mouse IgM Alexa 546 conjugate and evaluated by fluorescence microscopy (Fig. 2a). Hc GFP yeasts recovered from Hc + Cn H99 co-infected lungs displayed strong labelling for GXM in comparison to the absence of labelling on fungal cells obtained from mice infected with Hc + Cn cap59 or Hc GFP control alone (Fig. 2a). To obtain more quantitative information, Hc yeasts recovered from infected mice were incubated with mAb 2D10 and anti-mouse IgM allophycocyanin (APC)-conjugate and evaluated by flow cytometry (Fig. 2b). As an additional control, GXM was exogenously added to Hc samples. As expected, Hc from Hc + Cn cap59 co-infected animals displayed low levels (background) of APC + , similar to Hc incubated with 2D10 mAb (p = 0.44) or Hc alone (p = 0.52) and suggested no labelling by 2D10 mAbs. Hc isolated from lungs of Hc + Cn H99 co-infected animals had higher fluorescence labelling by mAbs to GXM than to Hc+ 2D10 (*p = 0.016) or Hc + Cn cap59 ( # p = 0.019), indicating incorporation in vivo. Notably, the APC intensity values were similar for the Hc yeast isolated from animals co-infected with the PS-producing Cn H99 and isolated Hc spiked with GXM (p = 0.23). These results suggest that during co-infections the surface of Hc can be modified by the incorporation of cryptococcal PS material.
Cn and Hc interacted during co-cultivation. Given that both fungi can co-exist in nature and in tissues, we evaluated the interactions between Cn and Hc during in vitro cultivation. Fungal growth was examined semi-quantitatively on microtiter plates by measuring total metabolic activity of adherent cells and pellicle formation using an XTT assay (Fig. 3a) 34 . When cultivated separately, Cn H99 grows more robustly under biofilm conditions compared to Hc (*p = 0.016, 2 h), which is consistent with the differences in replication rate between the two fungi (approximately 2 and 6 h, respectively) and the well-described capacity of Cn to form a biofilm/PS matrix 23,34,35 . However, co-incubation of both fungi in a 1:1 ratio to create the same total initial inoculum, resulted in the formation of a hybrid pellicle, with similar metabolic activity relative to monospecies Cn biofilms (p = 0.46). The capacity of forming pellicles was nearly absent in Hc + Cn cap59 co-cultivation, where the metabolic activity was 27% lower relative to pellicles containing Hc and Cn H99 (*p = 0.029, 2h). 3-D image reconstruction . Immunofluorescence punctuate surface labelling of Hc GFP recovered from lungs of Hc GFP+ Cn H99 groups with GXM-binding mAb 2D10 (red) and Uvitex2B (chitin in the cell wall) after isolation from lungs of co-infected animals. In comparison, Hc recovered from lungs of Hc GFP+ Cn cap59 co-infected or monospecies (Hc GFP) infected mice are not labelled by the mAb. Scale bar = 5 μ m. (b) FACS demonstrates labelling of Hc cells by GXM 2D10 mAbs upon co-infection with Cn H99 or GXM added controls ( in comparison to unlabelled Hc GFP (*p = 0.013 and **p = 0.0069, respectively) and Hc yeast from co-infection with Cn cap59 ( # p = 0.037 and # p = 0.019).
Scientific RepoRts | 6:21765 | DOI: 10.1038/srep21765 of pellicles displayed the complex architecture formed when Hc was co-cultured in static conditions with Cn H99, in comparison with monospecies control and Hc + acapsular Cn cap59 and ( Supplementary Fig. S1a-c) along with the detection of fluorescence intensity ( Supplementary Fig. S1f), and correlated with the above described results.
To examine the PS matrix of these fungal pellicles, we performed an indirect ELISA using mAb 2D10 to cryptococcal GXM. Mixed Hc + Cn H99 pellicles displayed an average reactivity increase of 10% relative to Cn biofilms (Fig. 3b, *p = 0.029). No difference on the reactivity was observed when comparing mixed pellicles of Hc + Cn cap59 with pellicles from Hc yeasts alone (p > 0.99). These results also suggest that PS material from Cn is transferred to Hc and that these fungi form a hybrid pellicle matrix with increased serological reactivity. Altogether, these findings might suggest an interplay between Cn and Hc when grown together. In fact, Cn also has been reported to produce quorum sensing molecules that affect the growth of other fungi 36 . (a) Pellicle formation in HAM's F-12 media was determined by measuring fungal metabolic activity using XTT colorimetric analysis. The initial inoculum for each well was 10 6 yeast cells, either all of one species or a 1:1 mix of 5 × 10 5 of Hc co-cultured with either Cn H99 or Cn cap59. Heat-killed Hc cells were used as background control and discounted from the readings. Co-cultures of Hc GFP + Cn H99 formed pellicles that were similar to biofilms produced by monospecies Cn H99. In contrast, Hc GFP + Cn cap59 and monospecies Hc GFP were extremely poor pellicle producers. (b) The reactivities of PS matrix of fungal pellicles were examined by ELISA using 2D10 IgM mAb to GXM. The pellicles from Hc GFP + Cn H99 displayed a slight, but significant increase in matrix reactivity to GXM-binding mAb compared to biofilms formed by Cn H99 alone. For both A and B, bars represent mean ± standard error of quadruplicates. (c) Cn glycan transfer to Hc surface is temperature dependent. More glycan transfer occurs at 37 °C compared to 30 °C during co-cultivation of Hc (GFP strain, FL1-H + ) + Cn H99 in HAM's F-12 medium as determined by flow cytometry using GXM-binding mAb 2D10 (IgGM) and a goat anti-mouse IgM-APC (FL4-H + ), in comparison to controls or monospecies mixed Hc GFP + Cn H99 yeasts just before incubations with mAb, which displayed no PS transfer. Temperature dependency of cryptococcal glycan incorporation by Hc. Hc cells were individually evaluated by FACS upon incubation with mAb 2D10 and APC-labelled conjugate anti-mouse IgM after sonication of grown Hc + Cn H99 co-cultures at 30 °C and 37 °C. The optimal growth temperature for Cn is 30 °C and Hc yeast cells grown best at 37 °C. Co-cultivation of Hc (GFP) and Cn H99 resulted in transfer and incorporation of Cn-glycan fractions on the Hc surface (Fig. 3c), related to mixed monospecies control (Hc + Cn H99), which barely had increase in Hc fluorescence in comparison to Hc control. Incubation of cells at 37 °C resulted in a 5-fold increase in average APC + fluorescence intensity of GXM-positively labeled Hc compared to Hc grown at 37 °C, while cells co-cultured at 30 °C displayed a 3-fold increase in average fluorescence compared to Hc grown at the same temperature.

Cellular glycan cross-incorporation between Hc and Cn. Based on the incorporation of Cn PS by Hc,
we investigated also whether both thermodimorphic phases of Hc would incorporate Cn-glycan, since these fungi co-exist in nature in a wide variety of temperatures. When evaluated by fluorescence microscopy, the majority of Hc yeast cells in pure culture were not labelled by mAb 18B7 37 although few cells displayed a discrete punctuated pattern of labelling ( Supplementary Fig. S2a). The filamentous phase of Hc also displayed only few cells labelled by mAb 18B7 some concentrated at the septae ( Supplementary Fig. S2d). To evaluate the Cn-glycan incorporation, as previously described for GXM and Hc 25 , Hc yeasts were incubated with Cn-glycan resulting in a radial labelling surrounding the Hc surface or "pseudoencapsulation" of Hc by the Cn-glycan ( Supplementary Fig. S2b). The surface of filamentous forms was also able to incorporate Cn-glycan onto it's the surface, with the most intense fluorescence surrounding micro and macroconidia ( Supplementary Fig. S2e). Pre-treatment of yeast or hyphal cells with the cell wall degrading cocktail Novozyme completely abrogated the binding of Cn-glycan ( Supplementary  Fig. S2c,f), suggesting the requirement of cell surface molecules in this process.
Incorporation of distinct cryptococcal glycan fractions onto the Hc surface. The attachment or anchoring of the Cn capsule involves PS-PS interactions between GXM and other cell wall gly (i.e. glucans and chitin) 25,26 . Given that the Hc surface is richly composed of glucans and N-acetylglucosamine polymers 32 , we examined the possibility of a transference and/or incorporation of distinct cryptococcal PS fractions onto the Hc surface. This was assessed by incubation of Hc cells with isolated capsular C-gly-Cn (DMSO extracted) and secreted E-gly-Cn (filtered supernatant) fractions, isolated from Cn. The C-gly-Cn fraction was readily incorporated by Hc yeasts based on a 7-fold increase in mAb 2D10 labelling relative to control (Fig. 4a). The E-gly-Cn fraction was less well incorporated; nevertheless, incubation with this fraction led in a 4-fold increase in antibody labelling.
C-gly-Cn incorporation modified the charge of Hc as demonstrated by the change in the surface electrostatic potential of Hc cells. The association of C-gly-Cn with Hc cells resulted in a significant increase in the negative magnitude of the zeta potential (− 46.56 ± 10.25 mV) relative to uncoated Hc yeasts (− 34.10 ± 7.10 mV, *p = 0.0066) (Fig. 4b), most likely due to the addition of glucuronic acid residues, which are absent on Hc surface gly. Incubation with E-gly-Cn not significantly alter Hc surface charge (− 35.63 ± 4.18 mV), consistent with the lower incorporation and the lower relative levels of glucuronic acid in this fraction compared to C-gly-Cn 38 . Together, these results demonstrate an interaction between both fungi involving the transfer and incorporation of Cn PS material to the Hc surface gly via PS-PS interactions, which leads to significant alterations in Hc cell surface charge.
Growth of Hc in the presence of C-gly-Cn or E-gly-Cn enabled these yeasts to more effectively form a pellicle structure (Fig. 4c) equally in the presence of either glycan fraction (p = 0.77, 2 h), relative to Hc alone (*p = 0.037 and p = 0.045, respectively). Structurally, these pellicles were characterized by dense aggregates of yeasts, which could anchor each other through interactions with Cn-gly working as a extracellular polymeric scaffold substance ( Supplementary Fig. S1d,e).
The post-incorporation ultrastructure of Hc was evaluated by SEM (Fig. 5). As a control, Cn cap59 yeasts were incubated with C-gly-Cn or E-gly-Cn and uniform attachment of capsule was observed, with C-gly-Cn producing the most robust capsules, in comparison to E-gly-Cn, and agreement with the previously described size of PS fibers from these distinct fractions 38 . Similarly, Hc yeasts incubated with C-gly-Cn displayed significantly larger PS fibers incorporated onto their surface compared to the smoother surface by E-gly-Cn, which had a more sparsely coated surface, but more wrinkled than Hc control. Together, these resulted were consistent with the FACS and zeta-potential determinations.

Kinetics of Cn glycan incorporation by
Hc and α-glucan requirement. The requirement for α -1, 3-glucans in the incorporation of C-gly-Cn or E-gly-Cn was evaluated by comparing Hc strains expressing variable amounts of these surface glucans. Decreasing concentrations of C-gly-Cn or E-gly-Cn were incubated up to 1 h with low (G217B) or high α -1, 3-glucan content (G186A) Hc strains 39 . C-gly-Cn was incorporated by both Hc strains ( Supplementary Fig. 3Sa,b). Despite higher incorporation of C-gly-Cn by G186A, in agreement with Reese et al. 25 , this process was effective only at 1 h incubation, in contrast with G217B strain, which displayed a statistically significant incorporation of C-gly-Cn after a 30 min incubation, when compared to controls. Similar behaviour was observed for both strains with E-gly-Cn incubation; however, as expected, absorbance values were lower than those obtained for C-gly-Cn, due to the lower incorporation of this fraction by Hc strains.
Cn glycans-coated Hc yeasts are more resistant to phagocytosis and antifungal activity by peritoneal macrophages. Given the antiphagocytic and immunomodulatory properties of cryptococcal PS, we examined if these virulence traits could occur with pseudoencapsulated Hc cells. Hc yeasts coated with C-gly-Cn were more resistant to phagocytosis by peritoneal macrophages compared to untreated Hc (38.6% vs 59.8% phagocytosed, ***p = 0.0001; Fig. 6a). Similar results were achieved with E-gly-Cn incorporation onto Hc (39.5%, **p = 0.0008). When the phagocytosis index was evaluated, i.e., the average number of yeast by macrophages, only C-gly-Cn reduced this number effectively (2, 67; **p = 0.0013), in comparison to E-gly-Cn (3,07; p = 0.072) and Hc control (3,71). Moreover, resistance to killing by macrophages was also increased for Hc coated with C-gly-Cn, as the CFUs were 4.3 times higher for these cells compared to uncoated Hc (1.2 × 10 5 vs colonies 2.6 × 10 4 , **p = 0.0038; Fig. 6b, left axis). E-gly-Cn coated-Hc similarly displayed a 3 times higher resistance to intracellular killing (7.8 × 10 4 , *p = 0.047) compared to untreated Hc (2.6 × 10 4 ). Resistance to killing (CFU) was normalized by the total yeast number inside the macrophages and yeast viability under each condition evaluated (Fig. 6b, right axis). This reduced macrophage antimicrobial efficacy was in part associated with the decreased levels of nitric oxide produced by these effector phagocytic cells when infected with C-gly-Cn or E-gly-Cn coated Hc (p < 0.05; Fig. 6c, left axis). Nitric oxide levels were normalized to the number of yeast inside macrophages Hc virulence is enhanced in vivo via Cn glycan transfer. In vivo mouse models were used to determine the importance of the transfer of individual pools of Cn-gly during co-infection in vivo. After infection with Hc and administration of C-gly-Cn or E-gly-Cn intratracheally, survival rates and lung CFUs were compared. Animals challenged with Hc and treated with E-gly-Cn had the highest mortality index, with all mice dying by day 11 (**p = 0.0011 compared to Hc infection alone, Fig. 7a), followed by mice in the C-gly-Cn treatment group, with animals dying by day 19 (*p = 0.035). Notably, some animals infected with Hc that received PBS instead of PS survived until the termination of the experiment at day 30. The CFUs recovered from Hc-infected and C-gly-Cn treated animals ranged from 1.98 × 10 7 to 6.28 × 10 7 (median 4.64 × 10 7 ) and the CFUs from infected, E-gly-Cn treated mice ranged from 1.71 × 10 7 to 6.23 × 10 7 (median 2.71 × 10 7 ), both of which were significantly higher than CFUs recovered from mice infected with Hc alone (1.14 × 10 7 to 3.19 × 10 7 ; median 1.95 × 10 7 ; p < 0.05; Fig. 7b). Since CFU numbers were higher in animals treated with either C-gly-Cn or E-gly-Cn than control, we wanted to determine if the higher virulence was correlated with the presence of Hc-coated yeasts with the administered Cn-gly. Significantly, organ homogenates from each of the Cn-gly-treated Hc displayed Hc with intense fluorescence staining by mAb 2D10, indicating the presence of Cn-gly coated Hc yeast (Fig. 7c). Hc recovered from animals challenged with Hc alone did not react with the GXM-binding mAb.
We also tested the impact of PS-coating of Hc using the in vivo invertebrate model Galleria mellonella (Fig. 8); however, this approach was limited by the use of only Hc yeasts pre-incubated with purified Cn-gly. The results demonstrated a dose-dependent increase in virulence of Hc yeasts when coated with C-gly-Cn as Hc coated with 100 μ g displayed higher virulence relative to untreated Hc (**p = 0.004). Treatment with 10 μ g C-gly-Cn did not reach statistical significance compared to Hc alone (p = 0.062). In contrast, co-incubation with E-gly-Cn prior to infection resulted in similar mortality rates as that observed for the untreated Hc in our Galleria model. Nevertheless, our finding that coating of Hc with C-gly-Cn enhanced virulence in this second model strengthens our thesis that gly transfer between Cn and Hc during co-infection can enhance virulence and exacerbate disease.

Glycan transfer occurs inside phagocytic cells. GXM can easily be ingested by macrophages through
pinocytosis and phagocytosis 40 . Cn and Hc can also be phagocytosed by these cells and localized within phagosomes. We therefore evaluated intracellular glycan transfer within macrophage phagolysosomes. Macrophages were infected with Hc and then exposed to Cn H99, Cn cap59, C-gly-Cn or E-gly-Cn. Hc GFP was detected as a green fluorescent cells inside phogosomes (Fig. 9). Glycans reacting with mAb 2D10 were detected as punctuated patterns inside the macrophages in the presence of Cn H99, or upon incubation with C-gly-Cn or E-gly-Cn as described previosuly 40 . Hc GFP and Cn H99 co-localized within the same phagosome, and a punctuated labelling for GXM was observed around Hc GFP yeast (Fig. 9). When C-gly-Cn or E-gly-Cn were administered upon incubation of macrophages with Hc, higher distribution of GXM and labelling of Hc GFP yeasts by GXM antibody was observed, with a predominance of GXM staining on the surface of Hc yeasts (Fig. 9). Systems where Hc GFP was used only or where infection with Hc GFP was followed by Cn cap59 produced no labelling for GXM.

Discussion
Histoplasmosis and cryptococcosis are the most prevalent pulmonary mycoses in HIV-infected patients 2,3,20 . Hc var. capsulatum infection has emerged as one of the most common systemic mycosis in the setting of HIV-infected patients in developing countries 41 , where disseminated histoplasmosis continues to cause severe morbidity and   mellonella model. The incorporation of C-gly-Cn onto the surface of Hc increased mortality in a dose dependent manner, with the addition of 100 μ g C-gly-Cn producing a statistically more rapid time to death compared to untreated Hc (100 μ g,p = 0.004; 10 μ g p = 0.062). There were no significant differences between Hc exposed to E-gly-Cn coated Hc and Hc alone (100 μ g, p = 0.23; 10 μ g, p = 0.22). Figure 9. Hc co-localize with Cn-gly and is able to incorporate these glycans on its surface within the macrophage environment. Macrophages were infected with Hc GFP (green) and incubated with either PBS, Cn H99 (Uvitex labeled -blue), Cn cap59 (Uvitex labeled -blue), C-gly-Cn or E-gly-Cn. Fluorescence was performed using 2D10 mAb and anti-IgM Alexa 568 conjugated (red). In the presence of Hc GFP yeasts and either Cn H99, C-gly-Cn or E-gly-Cn, Hc surface was labeled with 2D10 antibody as indicated in several instances by the black arrow heads. For Hc and PBS or Cn cap59 groups, no labelling for GXM was observed. Left column (Hc GFP-green; Cn Uvitex -blue; mAb 2D10 -red). Right column (Hc GFP-green; Cn Uvitexblue; mAb 2D10 -red) merged with light microscopy. Scale bar = 10 μ m.
Pulmonary infections by both Hc and Cn frequently display overlapping symptoms 42,43 . In addition, their clinical, pathologic and imaging findings can be similar 44 . Both fungi can be isolated from bronchoalveolar aspirates, but Cn is able to overgrow Hc in culture and even inhibit its growth 22 , which may be a reflection of its simpler nutritional requirements and faster replication rate 42 . Besides culture, standard microscopic examination does not uniformly distinguish between these species, due to morphological similarity of these fungi in clinical samples, particularly when hypocapsular strains of Cn are involved 14 . Cn and Hc also share the ability to proliferate within macrophages and both species are considered to be facultative intracellular pathogens 10 .
A PubMed search for the words Hc, Cn and co-infection renders many hits 9-21 , including multi-center reports of several patients 9,15,20 , with the first co-infection observation reported by Mider et al. in 1947 16 . The diagnoses of co-infection was made by either histological examination and/or cultures of various tissues and body fluids. The majority of the reports date from the last decade, and are frequently associated with disseminated infection by both fungi in the setting of advanced HIV disease 9,12,13,15,17,18,20,21 . However, as an example of co-infection in a non-HIV infected patient, Hc and Cn were found in samples of respiratory secretions in an individual on chronic steroid therapy who presented with a cavitary pulmonary lesion 19 . In a study to validate an ELISA for the diagnosis of histoplasmosis, 12% of the histoplasmosis patients also had positive results for the presence of Cn by detection of GXM in the cerebrospinal fluid (unpublished and 45 ). In this context, we speculate that the total number of co-infection cases is generally underestimated, primarily due to the lack of sensitivity of the methods currently in use to diagnose histoplasmosis. Additionally, the diagnosis of non-meningeal cryptococcosis is difficult 1,46 . However, the advent of more sensitive molecular diagnostic techniques has increased the ability for detecting Hc in the setting of co-infections. It should be noted that other co-infections with dimorphic fungi also occur, such as the recently reported lethal human dual infection with Blastomyces and Coccidioides spp 47 .
We postulate that interactions of Hc with cryptococcal GXM may contribute to the pathogenesis of a significant number of histoplasmosis cases. Hc and Cn are frequently found in the same natural sites 48 , as they are highly associated with soils enriched with organic nitrogen sources, such as animal excrements. For instance, Cermeno et al. 49 co-isolated Cn and Hc from many sites in Venezuela, reinforcing the possibility of environmental interactions and an enhanced risk of co-infection with both pathogens.
In both Hc and Cn, surface PSs are key molecules of the fungal cells since they are directly mediating interactions with the immune system. Cn GXM is recognized by Toll-like receptors 2 and 4 and/or CD14 on phagocytes, resulting in an incomplete activation of pathways necessary for TNF-α production and activation of inflammatory responses 40 . GXM is also recognized by CD18 resulting in the blockage of the receptor, which subsequently inhibits leukocyte infiltration into inflammatory sites 23,40 . In Hc, α -and β -glucans form the outer cell wall layers of both yeast cells and mycelia, playing different biological roles 31 . The β -1, 3-glucan, which predominates in the mycelial phase, is antigenic and modulates the host immune response 31 . In most Hc isolates, α -1, 3-glucan surrounds the β -1, 3-glucan layer, blocking its innate recognition by dectin-1 on host phagocytes 50 , and thereby suppressing the production of TNF-α 31 .
The interaction between Cn and Hc can result in hybrid pellicle formation. We found that pellicle formation was increased when GXM producing Cn yeast cells were co-incubated with Hc. This observation suggests that cell wall components of Hc could interact with Cn-gly to promote adhesion of matrix components resulting in effective pellicle formation. In fact, Hc can incorporate exogenous Cn GXM but the mechanism by which PSs are attached to Hc cell surfaces remains obscure. Previous reports have demonstrated that only an α -1, 3-glucan-producing Hc strain could anchor soluble GXM based on immunofluorescence staining 25 . However, no direct labelling control of mAbs to Hc was performed. In our system, strain G217B, which displays no α -1, 3-glucans, had a slightly lower efficacy in incorporating C-gly-Cn in comparison with the strain G186A, a well-recognized α -1, 3-glucan-producing strains 39 , possibly indicating that Hc α -1, 3-glucans are not specific determinants for interaction with cryptococcal gly. The PS adsorption was more efficient when C-gly-Cn were used in comparison with E-gly-Cn. Coating Hc with C-gly-Cn also resulted in an increase in the magnitude of the fungal cell's negative charges, most likely due to the fact that this cellular fraction was better incorporated onto the cell surface and that it has higher amounts of glucuronic acid residues than the extracellular soluble fraction, E-gly-Cn 38 .
In addition, the incorporation of Cn-gly by Hc in both environmental and infection-related conditions, may have the potential to modify the outcome of the interaction between yeasts and phagocytes and/or environmental predators. Such an altered outcome was observed with the environmental in vivo model G. mellonela, which likely favors the survival of both microorganisms under environmental stress conditions and/or during interactions with the innate immune system. Coating of Hc with crypotoccocal PS might also inhibit the interaction with phagocytes, including macrophages, dendritic cells, neutrophils in mammalian models and haemocytes in G. mellonella invertebrate model. Within phagocytic cells, as shown with macrophages, GXM is extensively released by Cn in the phagosome 51 . In the case of co-infection of a single magrophage, as shown, the Cn-gly could be incorporate and associate with Hc yeast cell surface.
The results presented in this study suggest that Cn and Hc share a number of physiological steps required for gly formation and surface assembly. In addition, they also reveal a new pathogenic mechanism, resulting in increased virulence or synergism, with potential relevance for hosts co-infected with these fungi. Our in vivo observations suggest that these fungal pathogens can interact during infection, and Hc could modify its cell surfaces in a manner that alters recognition by the immune system. The explanation for the Cn-gly incorporation effect on Hc virulence may primarily be due to a subversion of the host immune recognition mechanisms of cell wall components with subsequent increase in yeast survival, which is an effect observed when comparing highly capsulated Cn strains to minimally capsulated ones 30 . Hence, direct PS transfer resulted in enhanced Hc virulence associated with the suppression of the antifungal functions of phagocytic cells.
Our findings also suggest that an increased understanding of the role of PS in fungal infections may lead to promising strategies for the design of new therapeutics 37,52 , as PSs constitute important targets for vaccines and passive immunization 53 . The mechanism used by fungal cells to incorporate exogenous molecules with consequent change of their surface architecture consists of a new avenue for cell biology studies and likely for the design of new therapeutic options. In summary, our findings show that Hc can co-opt GXM, the major virulence factor of Cn, during mixed infection in vivo and that this phenomenon was associated with increased virulence, both in vitro and in vivo. This observation establishes the precedent of one pathogenic microbe using a virulence factor from another to increase its virulence, suggesting that other such interactions may exist in host-microbe relationships. Although this is a new concept for synergistic dual fungal infection, the paradigm is well known in bacterial diseases and increasingly emerging in fungal-bacterial infections. For example, infection with mixed bacterial species can produce synergisms in virulence resulting in severe disease, such as Fournier gangrene. Bacterial-fungal interactions such as those described for Pseudomonas aeruginosa and Candida albicans can affect the expression of several fungal characteristics including some associated with virulence 54 . Our experiments extend the phenomenon of microbial synergy in virulence due to mixed infections within the fungal kingdom. Hc was cultured in HAM's F-12 (Invitrogen) medium as described 55 . Cn and Hc cells were grown at 30 °C and 37 °C, respectively, for 48 h with shaking at 150 rpm. For the co-cultivation of both fungi, Hc and Cn were centrifuged at 1100 × g for 10 min at room temperature (RT) and pellets were washed three times with PBS followed by centrifugation. The cells were then suspended in HAM's F-12 and enumerated using a hemocytometer. Hc and Cn yeasts were added to a final density of 5 × 10 5 yeasts/mL in HAM's F-12 and co-cultures were incubated at 30 °C and 37 °C. Monospecies controls of Hc (G217B of GFP) or Cn (H99) at 10 6 yeasts/mL were incubated separately in 50 mL of HAM's F-12 at 37 °C and 30 °C.

Methods
Mouse co-infection model. To evaluate survival rates during co-infection in vivo, C57BL/6 mice (6-8 weeks old) were challenged intranasally with 5 × 10 6 Hc GFP, followed 2 h later by an intratracheal infection with 5 × 10 6 of Hc GFP (monospecies control), Cn H99 or Cn cap59. The infected mice were checked four times daily by the scientific team and daily by the veterinary staff. All animal experiments were carried out in "accordance" with the approved guidelines and protocols of the Institutes for Animal Studies at the Albert Einstein College of Medicine and the Fluminense Federal University. To determine fungal burdens, immediately after they were detected, deceased animals had their lungs removed and the organs were then weighed and homogenized in PBS using 70 μ m cell strainers (BD Biosciences, NJ, USA). Organ homogenates were serially diluted and plated in duplicates on Sabouraud dextrose agar (Difco Laboratories) for Cn growth. After 2 d of incubation at 30 °C, Cn colony forming units (CFUs) were enumerated. For Hc growth determination, homogenates were simultaneously also plated on brain heart infusion (BHI) agar supplemented with 5% sheep blood and bleomycin at 10 μ g/mL (to suppress Cn growth in co-infection conditions). BHI plates were incubated in the dark for 10-15 d at 37 °C and Hc CFUs were then enumerated. The plates were also observed under UV light for expression of GFP proteins by Hc GFP strain and correlated with colony morphology.
To examine cryptococcal gly incorporation by Hc during co-infection, aliquots of lung homogenates were spun down and evaluated by immunofluorescence. Homogenates were pipetted into microcentrifuge tubes and quickly spun down to remove excess liquid. For detecting bound Cn PS, Hc yeasts were incubated with 10 μ g/mL of the IgM isotype GXM-binding mAb 2D10 or isotype-matched irrelevant antibody 33 and a 1:100 of a goat anti-mouse IgM Alexa 546 conjugate. After three washes, fungi were stained using 0.5 mg/mL of Uvitex 2B, fixed with 4% paraformaldehyde and analysed in an AX70 fluorescence microscope. Alternatively, we used goat anti-mouse IgM APC conjugate and performed analysis of FL1+ FL4+ cells (GFP and APC labelled, respectively) using a FACScalibur Flow Cytometer (BD Biosciences, Franklin Lakes, NJ) and Hc fluorescence intensity was determined under each condition.
Hc pellicle formation induced by Cn or their products. Monospecies cultures of Hc and Cn yeasts were obtained as described above, collected by centrifugation, washed with PBS (3X), and suspended at 10 7 cells/mL in HAM F-12 media. An aliquot of Hc yeast suspension was heat-killed at 56 °C for 1 h and used as negative control. Next, 100 μ L (10 6 total yeast) of each suspension (Hc or Cn) was added to individual wells of polystyrene 96-well plates (Fisher, MA). In co-incubations conditions, 50 μ L (5 × 10 5 ) of Hc GFP and 50 μ L (5 × 10 5 ) of Cn H99 or Cn cap59 were added to the same well (10 6 total yeast cells per well). Plates were incubated at 37 °C without shaking for 48 h. Following incubation, wells were washed (3X) with PBS 0.05% Tween 20 to remove planktonic cells. Pellicle formation, as agglutination of cells on a surface, was measured by XTT (2, 3-bis (2-methoxy-4-nitro-5sulfophenyl)-5-[(phenylamino) carbonyl]-2H-tetrazolium-hydroxide) reduction assay as previously described 34  To determine the pellicle formation and initial accumulation of PS extracellular matrix component, an ELISA with IgM mAb 2D10 was performed 33 . After 48 h incubation, plates were washed (3X) with TBS-T (10 mM Tris-HCl, 150 mM NaCl, 1 mM NaN 3 , 0.1% Tween 20, pH 7.4) and incubated with blocking solution (2% Bovine Serum Albumin in TBS-T) for 1 h at 37 °C. After washes, mAb 2D10 was diluted at 10 μ g/mL in blocking solution. Fifty microliters of mAb solution was added to separate wells containing yeast cells in quadruplicate and the plates were incubated at 37 °C for 1h. An irrelevant IgM antibody 5C11 was used as a control 56 . Plates were washed (3X) with TBS-T and incubated with a 1:1000 dilution of a goat anti-mouse Ig (Southern Biotech) in blocking solution, for 1h at 37 °C. After washes (3X), plates were incubated with 50 μ L/well of 1 mg/mL p-nitrophenyl phosphate diluted in substrate buffer (1 mM MgCl 2 × 6H 2 O, 0.05 M Na 2 CO 3 , pH 9.8) at 25 °C for 30 min. Absorbances were measured at 405 nm on a microplate reader (BioTek Instruments, Winooski, Vermont, USA). Results shown are the average of 3 independent experiments.
Analysis of cross-incorporation of Cn polysaccharides. Co-cultivated Hc GFP and Cn yeasts at different temperatures (30 and 37 °C) were washed (3X) with PBS and incubated with 10 μ g/mL of mAb 2D10 or irrelevant isotype-matched antibody, diluted in blocking solution for 1 h at RT. As a control, Hc and Cn were mixed right before incubation with mAb. The yeasts were washed and suspended in 100 μ L of a goat anti-mouse IgM APC-conjugate (Southern Biotech) diluted 1:100 in blocking solution. The suspension was incubated for 1h at RT and washed with PBS. Cells were sonicated with 1 min cycles to disrupt any possible aggregates or biofilm formed during growth/incubations, fixed for 20 min using formalin buffer (Fisher Scientifics) and washed with excess of PBS. Analysis of FL1+ FL4+ cells (GFP and APC double-labelled) was performed in a FACScan Flow Cytometer (BD Biosciences, Franklin Lakes, NJ) and fluorescence intensity was determined for each condition.
Alternatively, Hc grown in the filamentous phase on microslides at RT were incubated with 100 μ g/mL of total PS obtained from Cn culture supernatants 25,38 for 1h at RT. Similarly, Hc yeast were adhered to poly-L-lysine coated slides and incubated with Cn PS. Slides were then washed and incubated with 10 μ g/mL of mAb 18B7 37 to GXM or isotype-matched antibody and a 1:100 of a goat anti-mouse FITC-conjugated Ab. As a control for glycan incorporation through the requirement of cell surface carbohydrates or proteins, cells were treated with Novozyme 234 (Novoenzyme, Windsor, UK), a multi-enzyme preparation containing carbohydrate and peptide hydrolases 57 . After three washes, fungi were stained using 0.5 mg/mL of Uvitex 2B and fixed with 4% paraformaldehyde. Glycan incorporation was examined with an immunofluorescence Olympus AX70 fluorescence microscope, with a magnification of 40X.
Isolation of fungal glycans. Two-day old 1 L cultures of Cn H99 yeasts were centrifuged for 10 min at 1100 × g. Both cells and culture supernatants were collected for the extraction of cellular attached gly (C-gly) and isolation of secreted extracellular gly (E-gly), respectively. C-gly extraction was performed with DMSO as described 38 . E-gly were obtained by ultrafiltration of the supernatant using nitrocellulose membranes with a nominal molecular weight limit (NMWL) of 10 kDa (Millipore, MA, USA) as described 38 . Concentrated E-gly and C-gly were dialyzed against water for 24 h (with at least 8 water exchanges) and then lyophilized. The Cn-gly were quantitated by inhibition ELISA as described 58 .
Incorporation of cryptococcal cellular and extracellular glycan fractions by Hc. C-gly-Cn and E-gly-Cn (100 μ g) were incubated with 10 7 GFP Hc yeasts for 1 h at 37 °C in PBS. Hc yeasts incubated in PBS alone were used as a control. Following incubation, cells were washed (3X) with PBS to remove unbound gly and enumerated using a haemocytometer. Gly incorporation by Hc GFP yeast was determined by FACS analysis using mAb 2D10 as described above.
Hc were also suspended at 10 7 cells/mL in HAM F-12 medium. Next, 50 μ L (10 6 total yeast) was added to individual wells of polystyrene 96-well plates (Fisher, MA) and incubated with 10μ g of C-gly-Cn or E-gly-Cn in 50 μ L of HAM F-12 media). Plates were incubated at 37 °C without shaking for 48 h. Pellicle formation was assessed as described previously.
To compare the relative incorporation of Hc G217B with the high α -1, 3-glucan strain Hc G186A, yeasts were incubated with the Cn gly fractions for different time intervals and incorporation was detected by indirect ELISA as described 58 .
Zeta potential measurements. Hc zeta potential was examined before and after incubation with C-gly-Cn or E-gly-Cn, and included untreated Hc yeast as a control. Analysis was done using 10 6 cells/mL in pure distilled LPS free water (Thermo Scientific HyClone). Zeta potential (ζ ) and mobility values of intact cells were measured in a Zeta potential analyser (ZetaPlus, Brookhaven Instruments Corp., Holtsville, NY) as described 58 .
Scanning electron microscopy. Acapsular Cn cap59 mutant, Hc or the Cn-gly-coated yeasts were washed three times in PBS and fixed with 0.1 M sodium cacodylate buffer containing 2.5% glutaraldehyde for 1 h. Yeast were washed with 0.1 M sodium cacodylate, 0.2 M sucrose and 2 mM MgCl 2 and fixed on coverslips coated with poly-L-lysine for 20 min. Preparations were then gradually dehydrated in alcohol (30%, 50%, 70% and 100% for 5 min and 95% and twice in 100% for 10 min), and submitted to critical point drying and metallization. The cells were observed in a Quanta-FEI scanning electron microscope (FEI,USA).
Phagocytosis. Four-to-six weeks-old female BALB/c mice were used for the isolation of peritoneal macrophages 59 . Macrophages were plated onto a culture chamber at 2 × 10 5 cells/well. Hc yeasts were labelled with 40 μ g/mL of NHS Rhodamine (Thermo Scientific, Rockford, lL, USA) for 30 min at 25 °C and washed (3X) with excess of PBS. Cells were incubated with the distinct Cn-gly or PBS as described above. Following incubation, cells were washed, suspended in DMEM, enumerated, and added to the macrophages in a 5:1 (yeast:macrophage) ratio. Plates were incubated for 1 h in 5% CO 2 atmosphere. After three washes with PBS, yeasts were stained using 0.5 mg/mL of Uvitex 2B to distinguish internalized versus extracellular yeasts. Wells were washed (3X) with PBS and fixed with a 4% formaldehyde solution in PBS. The number of macrophages and yeasts were recorded for each field by microscopic enumeration and at least 200 macrophages were counted. The percentage of phagocytosis was determined as the ratio of macrophages with internalized yeast cells divided by total macrophages, and the phagocytic index as the average number of yeast inside macrophages 55 . Yeast killing assay. Cn-gly-coated Hc yeast cells were suspended in DMEM and added in a 5:1 (yeast:macrophage) ratio to 96-well culture plates containing 10 5 macrophages/well. Plates were incubated overnight at 37 °C under 5% CO 2 . The wells were washed with cold PBS and macrophages lysed by adding sterile water. Aliquots were plated onto BHI-blood agar plates (10 g/L glucose, 0.1 g/L cysteine, 1% Pen-Strep and 5% v/v sheep red blood cells) and incubated at 37 °C for 10-15 days. The numbers of CFUs were enumerated and compared among groups.
Nitric oxide synthase activity. Nitric oxide production by peritoneal macrophages following incubation with control or gly-coated Hc yeast cells was determined from culture supernatant using the Griess reagent (Promega, Madison, WI, USA) according to manufacturer's instructions. A nitrite standard reference curve was prepared for accurate quantization of NO 2 levels in experimental samples. Experimental conditions were performed in quadruplicates. Plates were read in a spectrophotometer at 540 nm.
Survival in mammalian and invertebrate host models against Cn-PS coated Hc. Mice were intranasally infected with 10 7 Hc GFP yeast followed 2h later by intratracheal injection with 10 μ g (in 50 μ L) of C-gly-Cn, E-gly-Cn or PBS. Mice were checked four times daily by the scientific team and daily by the veterinary staff. Evaluations of Hc virulence were performed by survival and CFU quantification as described previously.
To further assess the effects of the incorporation of distinct gly pools by Hc in pathogenesis, survival experiments were also conducted in Galleria mellonella according to our established methods 60 . Prior to infection, Hc yeast cells were treated with cellular C-gly-Cn, extracellular E-gly-Cn or PBS (control) as described above. Infections were performed by injecting the hemocoel of each caterpillar via the last left proleg with 10μ L aliquot containing 10 6 yeast using a 10-μ l Hamilton syringe. Groups consisted of 10 larvae per group and experiments was repeated 3 times with similar results achieved.

Model of glycan transfer during infection of macrophages.
Peritoneal macrophages (2 × 10 5 in 200 μ L) were plated on 8-chambers culture slides (Falcon) and cultivated overnight at 37 °C under 5% CO 2 . Hc GFP yeasts were washed and added to macrophages at a 2:1 ratio, and infection performed for 2 hours. Chambers were washed three times with DMEM to remove extracellular Hc GFP yeasts. Cn yeasts were incubated with Uvitex 2B as described above, and either Cn H99 or Cn cap59 were added to macrophages in a 5:1 ratio. For C-gly-Cn or E-gly-Cn, glycans were diluted at 10 μ g/mL in 200 μ L of DMEM and added to individual wells. Chambers were incubated overnight at at 37 °C under 5% CO 2 . After washing with PBS, chambers were fixed as described and immunofluorescence conducted as described above.
Statistical analysis. All analyses were performed using GraphPad Prism version 6.00 for Windows (GraphPad Software, San Diego California USA). One-way ANOVA statistics using a Kruskall-Wallis non-parametric test was used to compare the differences among groups with a 95% confidence interval in all experiments. Individual comparison between groups was performed using Bonferoni post-test. Survival results were analyzed by Kaplan-Meyer to determine the difference among groups.