Abstract Candida auris is among the most important emerging fungal pathogens, yet mechanistic insights into its immune recognition and control are lacking. Here, we integrate transcriptional and functional immune-cell profiling to uncover innate defence mechanisms against C. auris. C. auris induces a specific transcriptome in human mononuclear cells, a stronger cytokine response compared with Candida albicans, but a lower macrophage lysis capacity. C. auris-induced innate immune activation is mediated through the recognition of C-type lectin receptors, mainly elicited by structurally unique C. auris mannoproteins. In in vivo experimental models of disseminated candidiasis, C. auris was less virulent than C. albicans. Collectively, these results demonstrate that C. auris is a strong inducer of innate host defence, and identify possible targets for adjuvant immunotherapy. __________________________________________________________________ The important emerging fungal pathogen Candida auris was first described in 2009 and has, since then, spread across six continents as a causative microorganism of hospital-acquired infections^[67]1. For several reasons, C. auris is among the most challenging of emerging human pathogens that have been identified in the last decade. It is highly resistant to many of the commonly used antifungal drugs^[68]1 and, within a few years, it has rapidly spread worldwide^[69]2,[70]3 through the nearly simultaneous (but independent) emergence of four distinct phylogeographical clades^[71]4. Recently, a potential fifth clade has been described in Iran^[72]5. Every major clade except for clades II and V has been linked to outbreaks with invasive infection^[73]6. By contrast, clade II generally shows antifungal susceptibility and has a propensity for ear infections. Similar to clade II, clade-V isolate was recovered from an ear sample. Clade III is associated with bloodstream infections and, together with clade II, tends to form large cell aggregates^[74]7. This has been linked to a reduction in virulence in a Galleria mellonella infection model^[75]8. C. auris poses difficulties in routine microbiological identification^[76]9,[77]10 and it is challenging to eradicate in healthcare settings^[78]11–[79]14. This is due to its strong ability to colonize skin, its transmittance through the patient-to-patient route or contaminated fomites, and its high survival ability on plastic surfaces and in the hospital environment^[80]15. The risk factors for C. auris infections are generally similar to those for other types of Candida infections, such as prolonged hospitalization, use of central venous catheters, abdominal surgery and exposure to broad-spectrum antibiotics or antifungal agents^[81]16. However, owing to its acquired resistance to many antifungal drugs, the overall crude mortality rate of C. auris candidemia is high, ranging from 30% to 60%, with infections typically occurring several weeks (10–50 d) after admission^[82]4,[83]12,[84]17,[85]18. Echinocandins are presently recommended by the CDC as empirical treatment of C. auris infections, although resistance has been reported. Several new therapeutic alternatives, such as fosmanogepix^[86]19,[87]20, ibrexafungerp^[88]21–[89]23 and rezafungin^[90]24,[91]25 are presently under clinical investigation. Considering the importance of C. auris as an emerging human pathogen, it is imperative to understand the host defence mechanisms. This is particularly true given the high resistance of this fungus to anti-mycotic drugs, making it a prime candidate for the development of host-directed therapy (that is, immunotherapy). However, almost nothing is known regarding the host immune response against C. auris. Host defence against Candida species is dependent on a finely tuned interplay of innate and adaptive immune responses. A first, physical barrier consists of the skin and mucosa. The second barrier, represented by the innate immune system, is largely dependent on the recognition of evolutionarily conserved fungal cell wall components (pathogen-associated molecular patterns) by innate immune cells such as monocytes, macrophages and neutrophils. In turn, the release of proinflammatory cytokines, combined with antigen-presentation activity of myeloid cells, is crucial for shaping the adaptive immunity, which represents a third, long-term barrier against fungal infection^[92]26. The Candida cell wall is divided into an outer layer of highly mannosylated proteins (mannoproteins) and an inner layer, mainly comprised of β(1 → 3) and β(1 → 6)-glucans and chitin^[93]27. These pathogen-associated molecular patterns are recognized by various pattern-recognition receptors (PRRs) on the surface of immune cells, including C-type lectin receptors (CLRs)—such as dectin-1, dectin-2, macrophage mannose receptor (MMR) macrophage-inducible C-type lectin (mincle) and dendritic-cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN)—and Toll-like receptors (TLRs), especially TLR2 and TLR4 (ref. ^[94]28). Coordinated engagement of PRRs results in the activation of innate immune effector mechanisms, such as phagocytosis, reactive oxygen species (ROS) release and the production of pro- and anti-inflammatory cytokines. In turn, together with the antigen-presentation activity of myeloid cells, the release of pro-inflammatory cytokines shapes the adaptive immune response^[95]26. Although the antifungal host defence mechanisms have been extensively studied for Candida albicans, little is known about the host immune response against C. auris. Almost all multi-drug-resistant C. auris strains are susceptible to killing by the salivary antimicrobial peptide histatin 5 (Hst-5)^[96]29, whereas Johnson et al.^[97]30 showed that neutrophil recruitment as well as the formation of neutrophil extracellular traps (NETs) were lower for C. auris than for C. albicans. It was recently reported that C. albicans, Candida tropicalis, Candida guilliermondii, Candida krusei and C. auris differentially stimulate cytokine production in peripheral blood mononuclear cells (PBMCs)^[98]31, but little is known regarding the particularities of these responses and the mechanisms that mediate them. Considering the knowledge gap in our understanding of anti-C. auris host defence mechanisms, we set out to comprehensively assess the mechanisms through which innate immune cells recognize C. auris, initiate innate antifungal immune responses and protect the host against C. auris infection. This mechanistic insight into C. auris host interactions is instrumental for the development of novel host-directed approaches to treat severe C. auris infections and, thereby, improve patient outcomes. Results Common and specific transcriptome signatures induced by C. albicans and C. auris in human immune cells. To gain a broad overview of the host immune response against C. auris, we performed RNA sequencing (RNA-seq) analysis of PBMCs obtained from three healthy donors who were exposed to either live C. albicans or C. auris for 4 h or 24 h. Owing to the high genome-wide nucleotide identity across C. auris clades I–IV (98.7%)^[99]32, we expected that the analysis of the C. auris reference strain KCTC171810 (clade II) would provide valuable insights into generic C. auris-induced host responses. This C. auris reference strain was compared to C. albicans 10061110, which, to date, remains the most common cause of mucosal and systemic candidiasis^[100]33. Principle component analysis (PCA) of the normalized PBMC RNA-seq dataset revealed that the majority of the variance in the experiment as a whole was time dependent, as demonstrated by a clear separation of the 4 h and 24 h stimulation time points ([101]Extended Data Fig. 1a). A comparison of the stimulated and non-stimulated samples at the 4 h time point indicates that a limited response was induced early on. Moreover, stimulus clustering at 4 h suggests that this short-term response was similar in C. auris and C. albicans ([102]Extended Data Fig. 1a). At the 4 h time point, clustering of the donors, irrespective of stimulus, indicates that interindividual differences underpin the observed variance ([103]Extended Data Fig. 1b, left). As the PBMC donors were considered to be biological replicates, comparison of the average PBMC response to their control condition revealed a considerable overlap in the 4 h host response between C. albicans and C. auris. With 71 differentially expressed genes (DEGs; fold change (FC) ≥ 2, adjusted P < 0.01) upregulated by both Candida species, the respective overlap ranges from 67% of the total number of DEGs for C. albicans (71 out of 109) to 95% of the total number of DEGs for C. auris (71 out of 79). By contrast, at 24 h, the response is primarily stimulus driven ([104]Extended Data Fig. 1b, right), as indicated by the scattering of donor responses dependent on pathogen exposure. After 24 h of stimulation, the common C. auris- and C. albicans-induced host responses increased to 243 DEGs ([105]Fig. 1a), in turn accounting for 55% of the total number of DEGs for C. albicans (243 out of 442) and 50% of the total number of DEGs for C. auris (243 out of 484). This late shared response between both Candida species was consistent with the observation that the 24-h-induced PBMC transcriptomes were more stimulus specific ([106]Extended Data Fig. 1b). Pathway enrichment analysis revealed that the 4 h Candida intrinsic response was delineated by a common activation of the CC and CXC chemokines ([107]Supplementary Table 1). By contrast, the 24 h PBMC transcriptomic response was characterized by a broader upregulation of chemokines, interleukins (IL), tumour necrosis factor and their receptors ([108]Fig. 1a and [109]Supplementary Table 2). Fig. 1 |. Comparative analysis between the general and clade-specific C. auris- and C. albicans-induced host response at 24 h. Fig. 1 | [110]Open in a new tab a, The number of DEGs of both Candida species and their relative overlap revealed a substantial overlap between the C. albicans (10061110) and C. auris (KCTC17810, clade II) live induced host response at 24 h. DEGs were processed for pathway enrichment analysis, which in turn revealed the top-15 Candida intrinsic (overlapping DEGs, middle) and species-specific (DEGs unique to C. albicans (left); DEGs unique to C. auris (right)) pathways. Enrichment was determined using Consensus PathDB, including pathways as defined by KEGG (red) and Reactome (pink). Adjusted P < 0.01 (q value) was considered to be significant. The size of the geometric points indicates the amount of DEGs in relation to the size of the pathway. The exact q values and DEGs in pathways are provided in [111]Supplementary Table 2. b,c, C. auris is a more potent inducer of the immune system in comparison to C. albicans. b, TNF-α, IL-6, IL-1β and IL-1RA levels in the supernatants of PBMCs after stimulation without (RPMI; negative control) or with live C. albicans and C. auris for 24 h. n = 12. c, TNF-α, IL-6, IL-1β and IL-1RA levels in the supernatants of PBMCs after stimulation without (RPMI; negative control) or with live C. albicans and C. auris from all five geographical clades for 24 h. n = 8. For b and c, data are mean ± s.e.m., pooled from at least two independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001; statistical analysis was performed using two-sided Wilcoxon matched pairs signed-rank tests comparing respective C. auris strains with C. albicans as control or reference species. The substantial activation of glucose, fructose and mannose metabolism was unique to the C. albicans-induced transcriptional response of PBMCs at 24 h. By contrast, the DEGs that were more strongly induced after PBMC exposure to C. auris appeared to be linked to type-I and -II interferons, as well as antiviral mechanisms triggered by interferon- (IFN)-stimulated genes, including the ISG15 immune mechanisms ([112]Fig. 1a and [113]Supplementary Table 2). Collectively, these data show that C. albicans and C. auris are potent activators of the host immune system, and they are not only able to activate common transcriptional responses but also induce pathways specific to each pathogen. C. auris is a more potent inducer of host immune response compared with C. albicans. We also found that one-third of the top-15 enriched pathways, on the basis of those DEGs that were unique for C. auris, were enriched in the common Candida response ([114]Fig. 1a). This was not the case for C. albicans, indicating that C. auris has the ability to upregulate more genes in these pathways compared with C. albicans. Most pronounced within these unique DEGs were distinct interleukins, such as IL1RN (encoding IL-1RA), IL10, IL19, IL26 and IL27, as well as IFN-associated genes, such as STAT2, DDX58, EIF2AK2, OAS2, OAS3, IFIT2, IFIT3, IFIT35 and IFITM1 ([115]Supplementary Table 2). Furthermore, DEGs were more potently induced in response to C. auris than in response to C. albicans (Student’s t-test, P = 0.003). An additional pathway enrichment analysis of all of the upregulated DEGs confirmed that the total number of DEGs for mutually enriched pathways was higher when the PBMCs were stimulated with C. auris rather than with C. albicans ([116]Extended Data Fig. 1c). Collectively, the broader and stronger induction of DEGs by C. auris resulted in higher enrichment scores (q values) for corresponding pathways in comparison to C. albicans, suggesting that C. auris is a more potent trigger of the host response. Given that the transcriptomic analysis suggests that cytokine signalling is at the core of the host response, we aimed to verify these observations at the protein level. For this, we assessed the cytokine-production ability of PBMCs isolated from healthy volunteers after 24 h exposure to three different live clinical isolates for each Candida species, all cultured under similar conditions. As a measure of Candida-induced cytotoxicity, the detection of lactate dehydrogenase (LDH) release revealed that PBMC viability after 24 h was not impacted ([117]Extended Data Fig. 2a). With the exception of the anti-inflammatory cytokine IL-1RA, PBMCs exposed for 24 h to both clinical isolates of C. auris produced significantly higher amounts of the pro-inflammatory cytokines TNF-α, IL-6 and IL-1β compared with the C. albicans-stimulated PBMCs ([118]Fig. 1b). Moreover, to test whether this observation is clade dependent or reflects the general C. auris response, we assessed cytokine production after 24 h stimulation with different C. auris strains originating from the five different clades. Although variation in cytokine production was found between C. auris strains ([119]Fig. 1c), a similar pattern between pro-inflammatory cytokines was observed. Notably, C. auris clade V induced similar levels of pro-inflammatory cytokines and significantly lower IL-1RA levels compared with C. albicans. C. auris clades II and III induced cytokine production moderately, but this was still significantly higher compared with C. albicans. C. auris clades I and IV are extremely potent inducers of pro-inflammatory cytokines. Regardless of this clade-dependent variability, we observed that all C. auris clades except for clade V drive a significantly enhanced pro-inflammatory cytokine response compared with C. albicans ([120]Fig. 1c). However, at the transcriptional level, PBMCs stimulated with C. auris showed a trend of stronger induction of only IL6 (log[2] [FC] ± s.e.m. = 8.41 ± 1.4) and IL1B (log[2][FC] ± s.e.m. = 6.45 ± 1.29) expression levels compared with after exposure to C. albicans (log[2][FC] ± s.e.m. = 7.58 ± 1.37 for IL6; log[2][FC] ± s.e.m. = 5.59 ± 1.31 for IL1B) (not significant; [121]Extended Data Fig. 2b). C. auris replicates faster than C. albicans in vivo, leading to an altered m.o.i., but does not cause macrophage lysis. The killing of Candida by professional phagocytes of the innate immune system, such as monocytes, macrophages and dendritic cells, is an important line of defence at the site of infection. To study the differences in the phagocytosis dynamics of professional phagocytes between C. auris and C. albicans, we used live-cell video microscopy coupled with dynamic image analysis of bone-marrow-derived macrophages (BMDMs). We assessed phagocytosis ability by combining BMDMs with live and thimerosal-killed C. auris and C. albicans strains at an intended multiplicity of infection (m.o.i.) of 3:1 yeast cells per macrophage. By including fixed yeasts, we were able to assay phagocytosis in the absence of rapid adaptive changes in the composition of the Candida cell wall. Results are expressed as the percentage of phagocytic BMDM (percentage uptake), indicating the percentage of macrophages that have phagocytosed at least one fungal cell. Furthermore, we assayed the phagocytic index, which is defined as the number of fungal cells engulfed and fully inside the phagosome per 100 macrophages (by excluding the fungal cells adhering but not internalized). No significant differences in C. auris and C. albicans phagocytosis (percentage uptake) were observed for fixed Candida ([122]Extended Data Fig. 2c). There was a trend towards a higher phagocytic index in BMDMs at the later time points (after the second hour) for both of the fixed C. auris strains compared with the C. albicans strains ([123]Extended Data Fig. 2c), possibly because fungal cells tend to be phagocytosed in clusters. However, C. auris strain 10051893 had a lower phagocytosis efficiency after 1 h compared with C. albicans SC5314 ([124]Extended Data Fig. 2d). Focusing on live strains, we observed that the phagocytic index of both C. auris strains gradually increased over time, to a greater extent than the C. albicans strains ([125]Fig. 2a). To evaluate the phagocytosis dynamics using human cells, we incubated human PBMCs stimulated for 15 min, 30 min and 120 min (2 h) with thimerosal-fixed fluorescein-isothiocyanate-(FITC)-labelled C. albicans 10061110 and C. auris KCTC17810, and then calculated the percentage of FITC-positive cells as well as the mean fluorescence intensity (MFI) in the CD14^+ population. C. auris showed a significantly higher rate of phagocytosis and MFI at all the time points measured compared with C. albicans ([126]Fig. 2b). Fig. 2 |. Evaluating the dynamics of C. auris phagocytosis by human and murine host immune cells, and the heat sensitivity of the cell wall component that is responsible for C. auris-induced cytokine production. Fig. 2 | [127]Open in a new tab a, The capacity of BMDMs to phagocytose live C. albicans or C. auris strains in a 3 h period. BMDM engulfment is shown as the percentage of macrophages having phagocytosed at least one fungal cell (left). The phagocytic index was considered to be the number of fungal cells engulfed per 100 macrophages (right). Data are mean (n = 9), pooled from at least two independent experiments. b, Phagocytosis assay in human PBMCs showing the percentage of FITC-positive cells in the CD14^+ population. Phagocytosis efficiency was assessed as the percentage of CD14^+ cells that engulfed FITC-labelled Candida (left) and corresponding MFI of the total CD14^+ population (right). Data are mean ± s.e.m. (n = 6), pooled from two independent experiments. *P < 0.05; statistical analysis was performed using two-sided Wilcoxon matched-pair signed-rank tests, comparing respective C. auris strains with C. albicans as control or reference species. c, The distribution of phagocytosed live fungal cells per macrophage in a period of 3 h. n ≥ 100 observations per condition. d, The killing capacity of live C. albicans and C. auris is shown as the percentage of lysed macrophages (BMDMs) after 3 h of exposure. The ratio of yeast to macrophages (m.o.i.) was 3:1. Data are mean ± s.e.m.; n = 9 (n = 10 for C. auris 10051895), pooled from at least two independent experiments. *P < 0.05; statistical analysis was performed using Kruskal–Wallis tests with two-sided Dunn’s multiple comparison between the two C. auris strains and the two C. albicans strains. e, TNF-α (n = 10), IL-6 (n = 13), IL-1β (n = 13) and IL-1RA (n = 9) levels in supernatants of PBMCs after stimulation without (RPMI; negative control) or with heat-killed C. albicans or C. auris for 24 h. Data are mean ± s.e.m., pooled from at least two independent experiments. **P < 0.01, ***P < 0.001; statistical analysis was performed using two-sided Wilcoxon matched pairs signed-rank tests, comparing respective C. auris strains with C. albicans as control or reference species Using real-time live-cell microscopy, we observed live C. auris cells budding repeatedly outside the macrophages, with a doubling time of ~1 h. The C. auris budding rate decreased following phagocytic engulfment, although cells continued to multiply within the phagosomes ([128]Supplementary Video 1). Interestingly, C. auris 10051895 accumulated in high numbers within macrophages, indicating that the starting m.o.i. had exceeded the intended initial 3:1 ratio ([129]Supplementary Video 2). We therefore estimated the actual m.o.i. ratio at the start of image acquisition. Owing to the time elapsed between the counting of Candida in each sample and the image acquisition, the starting 3:1 ratio for C. auris increased to 7:1, presumably due to ongoing budding, despite the fact that samples were stored at 4 °C in PBS until the live imaging commenced. By contrast, the m.o.i. values for live C. albicans as well as for the fixed strains remained around the desired target of 3:1 yeast:macrophage. The increased m.o.i. for C. auris may be a contributing factor to the trend for higher phagocytic index achieved at 3 h for live strain 10051895 ([130]Fig. 2a). However, an increased starting m.o.i. for C. auris 10051893 did not enhance phagocytosis, as the live form of this strain was poorly recognized by BMDMs, showing a significantly lower percentage of uptake compared with C. albicans 10061110 ([131]Extended Data Fig. 2e). From the representative videos recorded, we quantified the distribution of yeast per individual macrophage after 3 h and found that, for both C. auris strains, there was a tendency for some macrophages to engulf many fungal cells, yet for other macrophages to engulf none. This phenomenon was less surprising for the Candida experiments using live microorganisms ([132]Fig. 2c), because C. auris continues to divide before and during the phagocytosis experiment. However, the fixed strains also displayed this varied distribution; both strains of C. auris were phagocytosed in large numbers by some macrophages ([133]Extended Data Fig. 2f). [134]Supplementary Video 3 shows that C. auris cells are taken up extensively into a subpopulation of macrophages but, despite the vast burden, these phagocytes continue to move around in pursuit of further fungal targets. Finally, macrophage lysis was determined following the engulfment of live Candida and we found that the C. auris strains examined had a significantly decreased ability to kill macrophages after 3 h compared with C. albicans 10061110, but not compared to C. albicans SC5314 ([135]Fig. 2d and [136]Supplementary Video 4), despite having a comparable (or greater, in the case of C. auris 10051895) phagocytic index. These findings demonstrate that C. auris is differentially recognized by phagocytic BMDMs and internalized with a higher phagocytic index compared with C. albicans but does not have the ability to induce the lysis of the phagocytic cells. Host immune response after C. auris exposure is mediated by heat-sensitive cell wall components. Variability in the cell surface structures, such as differentially expressed mannoproteins or altered β-glucan exposure between the C. auris and C. albicans cell walls, could account for the differential cytokine responses triggered by these pathogens. To elucidate whether this might account for the observed differences in cytokine induction, the C. auris and C. albicans strains were killed using heat. This approach facilitates disruption of the outer layer of the Candida cell wall^[137]34, in turn enhancing β-glucan exposure^[138]35,[139]36. The heat-killed strains were used to stimulate PBMCs for 24 h and 7 d. As the production of ROS can positively contribute to immune responses^[140]37, we assessed—in addition to PBMCs cytokine production—ROS release in both neutrophils and PBMCs during 1 h stimulation with serum-opsonized heat-killed C. auris and C. albicans strains. The area under the curve (AUC) was calculated to examine quantitative differences in ROS release. In neutrophils, both the ROS release over time and the total amount of production (AUC) were significantly lower following C. auris stimulation compared with C. albicans ([141]Extended Data Fig. 3a). In PBMCs, although the time-course luminescence was significantly decreased in C. auris-stimulated cells, only C. auris 10051893 induced a significantly lower total ROS production than C. albicans ([142]Extended Data Fig. 3b). Unexpectedly, compared with C. albicans, the cytokine response was almost completely abrogated after PBMC stimulation with heat-killed C. auris for 24 h ([143]Fig. 2e). This phenomenon was observed for all of the C. auris clades ([144]Extended Data Fig. 3c), we therefore annotated this characteristic for the general C. auris-induced host immune response. Furthermore, after 7 d, the production of IFN-γ and IL-17 by PBMCs stimulated with heat-killed C. auris was significantly lower compared with the production of IFN-γ and IL-17 by PBMCs stimulated with heat-killed C. albicans ([145]Extended Data Fig. 3d). We therefore reasoned that a heat-sensitive component of the cell wall might be responsible for most of the increased cytokine induction by C. auris. Mannans drive the host response to C. auris. We attempted to unravel the contribution of the different fungal cell wall components to the activation of host responses by C. auris, compared to C. albicans. PBMCs were exposed to the purified cell wall components, β-glucans and mannans from both Candida species, and the transcriptional responses of the stimulated immune cells were assessed using RNA-seq. The species-specific cell wall contribution was assessed by comparing the number of shared DEGs after exposure to the different cell wall components, and expressed as a proportion of the respective live stimulus. The early 4 h host response was predominantly induced by β-glucan, which was sufficient to explain around 82% and 57% of the live responses of C. albicans (89 out of 109) and C. auris (45 out of 79; [146]Fig. 3a), respectively. The β-glucans from each species resulted in similar PBMC gene expression profiles ([147]Extended Data Fig. 4a). Although the relative contribution of C. albicans β-glucan decreased to approximately 13% (55 out of 442) in the late phase, 24 h after stimulation, C. albicans β-glucan was able to upregulate several of the top-50 DEGs of C. albicans, indicating that they are a main contributor of the evoked response in the live setting ([148]Fig. 3a and [149]Extended Data Fig. 4a). By contrast, β-glucans from C. auris failed to elicit a response analogous to the live C. auris exposure, explaining only 2% (10 out of 484) of the live C. auris-induced response. By contrast, mannans from C. auris stimulated 28% (136 out of 484) of the evoked transcriptional response to live C. auris cells. Moreover, C. auris mannan seemed to outperform C. albicans β-glucan in relation to the top-50 DEGs of C. albicans, displaying an induction pattern that was similar to its live setting ([150]Extended Data Fig. 4a). Overall, these results indicate that host recognition and subsequent initiation of downstream responses against C. albicans is mainly dependent on β-glucan. For C. auris, early 4 h stimulation of PBMCs is mainly mediated by β-glucan, whereas mannans are fundamental for orchestrating the C. auris-specific host response at later time points (24 h). Fig. 3 |. Mannans are fundamental for orchestrating the C. auris-induced late host response. Fig. 3 | [151]Open in a new tab a, Split Venn diagrams indicating the number of DEGs after C. albicans 10061110 (left) and C. auris KCTC17810 (clade II; right) live stimulation with its respective overlap between exposure to the purified cell wall components β-glucan and mannan. The left split Venn diagram shows the early (4 h) response and the right split Venn diagram shows the late (24 h) response. b, PBMC production of the cytokines TNF-α, IL-6, IL-1β and IL-1RA after 24 h stimulation without (RPMI; negative control) or with purified β-glucans from C. albicans and C. auris strains in the presence of 10% human serum. n = 6 (n = 3 for C. auris β-glucan 10051244). c, PBMC production of TNF-α, IL-6, IL-1β and IL-1RA after 24 h stimulation without (RPMI; negative control) or with Pam3cys and/or purified β-glucans from different C. albicans and C. auris strains in the presence of 10% human serum, n = 6. d, PBMC production of the cytokines TNF-α, IL-6, IL-1β and IL-1RA after 24 h stimulation without (RPMI; negative control) or with purified mannans from C. albicans and C. auris strains in the presence of 10% human serum. n = 10. Data are mean ± s.e.m., pooled from at least two independent experiments. *P < 0.05, **P < 0.01. Statistical analysis was performed using two-tailed Wilcoxon matched pairs signed-rank test comparing both cell wall components extracted from C. auris strains with C. albicans as control or reference species (SC5314). Our data suggest that C. auris induces a stronger host immune response than C. albicans, and that the cell wall components of C. albicans and C. auris have a differential role in inducing gene expression. Therefore, to investigate the differences in cell wall structure between C. auris and C. albicans, we first compared forward light scatter (FSC) and side light scatter (SSC) of fungal cells using flow cytometry. Consistent with previous research^[152]23, we found that the C. auris strains have a smaller average cell size compared with C. albicans. Of the C. auris strains, strain 10051893 shows more complexity/granularity (higher SSC) than strain 10051895 ([153]Extended Data Fig. 4b). Next, we measured β-glucan exposure on the fungal cell surface using flow cytometry analysis of thimerosal-fixed Candida cells stained with Fc–dectin-1. C. auris 10051893 displayed significantly reduced exposure of β-glucan compared with C. albicans SC5314 ([154]Extended Data Fig. 4c). At the gene expression level, C. auris β-glucan (log[2][FC] ± s.e.m. = 2.9 ± 0.6) tended to induce IL1RN less effectively compared with β-glucan isolated from C. albicans (log[2][FC] ± s.e.m. = 4.5 ± 0.6; not significant; [155]Extended Data Fig. 2b). At the cytokine level, although large variation between different strains was observed, no significant differences were found in the cytokine production of PBMCs stimulated with purified β-glucans from C. auris compared with C. albicans for TNF-α and IL-1β. By contrast there was significantly lower IL-6 production in response to β-glucan from C. auris 10031160, C. auris 10051256, C. auris 10051263, C. auris 10051522 and C. auris 10051252 compared with C. albicans β-glucan. Interestingly, β-glucan from C. auris strains, except for C. auris 10051522, C. auris 10051244 and C. auris 10051252, induced a significantly lower IL-1RA production compared with C. albicans β-glucan ([156]Fig. 3b). Moreover, similar to C. albicans β-glucan, C. auris β-glucan synergistically boosted Pam3Cys (TLR2 agonist)-induced IL-1β production in PBMCs, as well as TNF-α and IL-6 production ([157]Fig. 3c). Having ruled out a major role for β-glucans in explaining the difference in cytokine stimulation induced by C. auris and C. albicans, we assessed the role of glycosylated mannoproteins from the fungal cell wall^[158]34. Examination of mannan exposure, by staining thimerosalfixed Candida cells with concanavalin A (ConA), revealed that there is a relatively low level of exposure of surface mannans in C. auris strains. This difference was significant for C. auris 10051893 as compared with C. albicans 10061110 ([159]Extended Data Fig. 4c). However, we observed that C. auris mannans significantly induce the gene expression of pro-inflammatory cytokines IL6 (adjusted P = 0.0001) and IL1B (adjusted P = 0.0003) compared with those of C. albicans ([160]Extended Data Fig. 2b). In line with these observations, mannans from all eight C. auris strains induced significantly higher cytokine production than C. albicans mannans 24 h after stimulation of PBMCs—both pro-inflammatory and anti-inflammatory ([161]Fig. 3d). Except for IL-1RA production, opsonization by human serum was necessary for mannan-induced production of cytokines ([162]Extended Data Fig. 5a). After 7 d of stimulation, no significant differences between mannan from C. auris and C. albicans as well with the unstimulated cells was observed ([163]Extended Data Fig. 5b). Unique structure of C. auris mannans. Nuclear magnetic resonance (NMR) spectroscopy analyses of the respective cell wall components for both Candida species revealed no structurally unique features in β-glucans isolated from C. auris. However, the distance between side-chain branching points (the average number of β-linked glucosyl repeat units) was larger for C. auris β-glucans than for C. albicans β-glucans ([164]Supplementary Table 3). By contrast, these side chains were much shorter for C. auris than for C. albicans β-glucans. For C. auris mannans, the acid stable portion was similar across all clinical isolates, revealing long side chains with varying lengths containing linked α−1,2-mannose, α−1,3-mannose and β−1,2-mannose in varying amounts ([165]Fig. 4). Strikingly, the acid labile portion of C. auris mannans revealed two distinct α−1,2-mannose-phosphate (M-α−1-phosphate) side chains ([166]Fig. 4), marking a unique structural feature. Subsequent analysis using multidetector gel-permeation chromatography highlighted that C. auris mannans are extremely small biopolymers with a molecular mass (MM) ranging from 6.1 × 10^3 Da to 16.1 × 10^3 Da among the clinical isolates. This represents a major difference compared with C. albicans mannan, which has a MM of 500 × 10^3 Da ([167]Supplementary Table 3). Fig. 4 |. NMR analysis of C. auris mannans revealed unique structural features. Fig. 4 | [168]Open in a new tab Mannans derived from eight clinical strains of C. auris show two distinct M-α−1-phosphate side chains as determined by two-dimensional (2D) COSY NMR spectroscopy analysis. The mannans were isolated from C. auris clades I, II and IV, and C. albicans. Although NMR revealed varying side-chain lengths containing α−1,2-mannose, α−1,3-mannose and β−1,2-mannose across the clinical isolates, the two distinct M-α−1-phosphate side chains were found in all eight strains and are characteristic of these C. auris mannans. We assessed the effects of these distinct and unique structural features of C. auris mannans on their ability to bind to the recombinant human (rh)-dectin-2 and rh-mannose receptors, compared with C. albicans mannans. Variability in binding affinities was observed among mannans from the different clinical C. auris isolates, ranging from an equilibrium dissociation constant (K[D]) of 1.0–6.0 μM for rh-dectin-2 and 2.1–6.3 μM for rh-mannose. Although differentially recognized by both receptors, their overall binding affinities were an order of magnitude lower than those that were observed for C. albicans mannans ([169]Supplementary Table 3). Moreover, of the examined structural features, the affinity of C. auris mannans for the rh-dectin-2 receptor was solely associated with a higher MM (r^2 = 0.4488, P = 0.034). A slightly lower association was observed for rh-mannose (P = 0.096), albeit displaying a similar trend. Collectively, our data suggest that the double M-α−1-phosphate side chains and small molecular size represent highly unique physicochemical properties of C. auris mannans that probably contribute to the decreased recognition efficiency by two important anti-fungal recognition receptors. The CLR complement receptor and MMR contribute to the C. auris-induced cytokine production. Next, we investigated the downstream effects of the reduced binding affinity of C. auris mannans on intracellular signalling pathways and the subsequent activation of the immune system. The spleen tyrosine kinase Syk is an important mediator downstream of several CLRs^[170]38. CLR signalling from dectin-1 and other lectins also involves the serine–threonine-protein kinase Raf-1 (ref. ^[171]39). Inhibition of Syk and Raf-1 decreased TNF-α, IL-6 and IL-1β production in response to C. auris stimulation ([172]Fig. 5a), indicating that these two signalling pathways are involved in cytokine production. Fig. 5 |. Examination of PRR and signalling pathways that are involved in the C. auris-induced host cytokine production. Fig. 5 | [173]Open in a new tab a, PBMC production of the cytokines TNF-α, IL-6, IL-1β and IL-1RA after 24 h stimulation without (RPMI; negative control) or with PFA-fixed C. albicans and C. auris strains, treated with vehicle (DMSO; n = 9) or a 1 h preincubation with Syk (n = 6) and Raf-1 (n = 9) inhibitors (inh.). b, PBMC production of the cytokines TNF-α, IL-6, IL-1β and IL-1RA after 24 h stimulation without (RPMI; negative control) or with live C. albicans and C. auris strains, treated with a 1 h preincubation with the isotype antibodies IgG2b, goat IgG and IgG1, or anti-DC-SIGN, anti-dectin-1, anti-mincle, anti-MMR, anti-CR3 and anti-dectin-2 blocking antibodies. n = 6. Data are mean ± s.e.m., pooled from at least two independent experiments. For a and b, *P < 0.05, **P < 0.01. For a, statistical analysis was performed using two-sided Wilcoxon matched-pair signed-rank tests, comparing (within each Candida strain) the respective inhibitor with its vehicle; for b, two-sided Wilcoxon matched-pair signed-rank tests, comparing (within each Candida strain) the neutralizing antibodies with the correspondent isotype controls Owing to the importance of the Syk and Raf-1 pathways in CLR pathway mediation, we subsequently hypothesized a role for these receptors in C. auris recognition. We therefore preincubated PBMCs with neutralizing antibodies against important Candida CLRs (such as dectin-1, dectin-2, mincle, DC-SIGN, MMR, CR3 and their control isotypes) 1 h before stimulation with live C. albicans or C. auris. As expected, blocking dectin-1 significantly decreased TNF-α production after C. albicans stimulation but, surprisingly, increased TNF-α after C. auris stimulation. ([174]Fig. 5b). We observed a significant reduction in C. auris-induced IL-6 and IL-1RA production after blocking MMR ([175]Fig. 5b). Interestingly, neutralization of dectin-1 and DC-SIGN led to a significant increase in IL-6 induced by C. auris compared with the IgG2b isotype control. Moreover, blockade of CR3 led to a significant decrease in IL-1β production and an increase in IL-1RA production ([176]Fig. 5b). We conclude that CR3 and MMR signalling promotes cytokine production in response to C. auris, whereas blocking dectin-1 functionality perturbs this cytokine production. C. auris is less virulent than C. albicans in an experimental model of murine disseminated candidiasis. To evaluate the virulence of C. auris in vivo, immunocompetent C57BL/6J mice were injected intravenously (i.v.) with either 10^7 colony-forming units (c.f.u.) of C. auris 10051895 (n = 10) or C. albicans 10061110 (n = 11). Their survival was monitored over the course of 14 d. Significantly more immunocompetent mice survived infection with C. auris than with C. albicans (3 out of 11 deaths for C. auris and 11 out of 11 deaths for C. albicans over 14 d; χ^2 = 21.42; P < 0.0001, Mantel–Cox test; [177]Fig. 6a). To evaluate the differential organ invasion ability between C. auris and C. albicans infection, we i.v. injected mice with either 10^6 c.f.u. of C. auris or C. albicans and euthanized the mice at day 3 (n = 5, C. albicans; n = 5, C. auris) and day 7 (n = 4, C. albicans; n = 5, C. auris) for c.f.u. counting both in liver and kidneys. Although there were not any significant differences after day 3, we found a significantly lower fungal burden at day 7 in the kidneys of C. auris-infected mice compared with C. albicans-infected mice ([178]Fig. 6b). To confirm whether the ex vivo greater pro-inflammatory response towards C. auris found in human PBMCs holds in vivo, we measured myeloperoxidase (MPO) in organs and several cytokines both in the plasma and organs of mice infected with 10^6 c.f.u. of either C. albicans or C. auris. Except for a significantly lower keratinocyte chemoattractant (KC) at day 7 ([179]Fig. 6c), no significant differences in cytokine levels were found between C. auris- and C. albicans-infected mice ([180]Fig. 6d and [181]Extended Data Fig. 6a–[182]e). To understand the inflammatory cytokine induction after the same load of Candida c.f.u., we normalized the inflammatory cytokine production to the actual remaining c.f.u. in organs by calculating the ratio of the mean MPO/c.f.u. and KC/c.f.u.. MPO and KC production per remaining C. auris c.f.u. were higher than the remaining C. albicans c.f.u. count ([183]Fig. 6f), supporting our in vitro findings in human PBMCs. In conclusion, improved survival and better clearance of invasive C. auris infection compared with C. albicans is ensured by an adequate immune response in immunocompetent mice. Fig. 6 |. C. auris is less virulent than C. albicans in an experimental model of mouse disseminated candidiasis. Fig. 6 | [184]Open in a new tab a, Survival curve of immunocompetent mice that were i.v. challenged with C. albicans (n = 11) or C. auris (n = 10). Mice were i.v. injected with 1 × 10^7 c.f.u. of the respective Candida strain and monitored daily. ***P = 0.001. b, The fungal burden of immunocompetent mice that were i.v. challenged with 1 × 10^6 c.f.u. of C. albicans (day 3, n = 5; day 7, n = 4) or C. auris (day 3, n = 5; day 7, n = 5) in the liver and kidney at 3 d and 7 d after injection. **P = 0.01. c,d, KC (c) and MPO (d) production in supernatants from liver, kidney and spleen homogenates. n = 6 per group per time point. e, KC production in the plasma of mice (n = 6 per group per time point) that were infected i.v. with 1 × 10^6 c.f.u. of C. albicans or C. auris. Data are mean ± s.e.m., pooled from at least two independent experiments. For a, statistical analysis was performed using two-sided Mantel–Cox log-rank tests; for b–e, two-sided Mann–Whitney U-tests. f, Ratio of the mean MPO or KC production (log scale, data from c and d) to the mean of fungal burden (log scale, data from b) in kidney (blue bars) and liver (white bars) of mice infected with C. albicans or C. auris. Data are the ratio of the mean log-transformed MPO and KC values (n = 6 per group per time point) to the mean log-transformed c.f.u. (C. albicans: day 3, n = 5; day 7, n = 4; C. auris: day 3, n = 5; day 7, n = 5). Discussion Here we investigated the transcriptional and functional responses of human PBMCs and mouse BMDMs to the rapidly emerging fungal pathogen C. auris. C. auris-induced host responses were compared to the responses elicited by C. albicans, as this species remains the most frequent cause of nosocomial fungal infections in humans to date^[185]33. A broad assessment of various clinical strains, as well as further verification among the five C. auris clades, revealed that, with the exception of clade V, C. auris induces a stronger immune response than C. albicans in vitro. Functional and structural assessment of β-glucans and mannans highlighted the presence of small and structurally unique C. auris mannans, which were crucial for immune recognition. Compared with C. albicans, the C. auris isolates examined in this study were more efficiently phagocytosed by immune cells, induced lower levels of macrophage lysis and displayed lower virulence in a mouse model of disseminated infection. C. auris induced robust transcriptional changes in human PBMCs. These included not only both of the common pathways that were also induced by C. albicans, but also more robust specific IFN-dependent transcriptional programs and explicit cytokine responses. This conclusion is supported by a recent study by Mora-Montes and colleagues^[186]31. Second, C. auris seems to induce stimulation of immune cells by sequential engagement of different components of the cell wall. The early (4 h) responses are mainly induced by β-glucans, and this initial phase of the response is largely similar to the response induced by C. albicans. This is probably explained by the similar structure of C. auris and C. albicans β-glucans. By contrast, the late transcriptomic responses (24 h) induced in PBMCs by C. auris showed significant differences and broader upregulation of immune genes compared with those induced by C. albicans. These late responses are mainly mediated by C. auris mannoproteins with a specific structure that includes a unique M-α-1-phosphate side chain in the acid labile portion of C. auris mannans. In line with the results of Yan et al. reporting that C. auris mannans bind strongly to serum IgG and mannose-binding lectin^[187]40, we showed that opsonization by human serum was necessary for C. auris-mannan-induced cytokine production. By comparing C. albicans-induced cytokine production with the different C. auris clades, we observed variability that is probably linked to their distinct phenotypic, epidemiological and drug-resistance properties. In particular, clade V was the least immunogenic, while clades I and IV were the strongest inducers of cytokine production. The amount of cytokines induced by the various strains might be correlated with clade-specific characteristics that, in turn, might influence the level of colonization/persistence in the host. When a healthy host encounters C. auris from clade I or IV, a large and fast protective pro-inflammatory cytokine response is induced. Although clade-III isolates are also linked with invasive infections, they show a lower in vitro induction of cytokine production compared with isolates from other clades. This may be linked to their tendency to form aggregates, which might make innate immune recognition challenging. Finally, clade II induces a relatively lower cytokine production, despite a higher phagocytosis rate compared with C. albicans. This might be related to its simpler mannan structure compared with isolates from clade I and clade IV ([188]Fig. 4). An important question concerns the PRRs that are responsible for recognizing C. auris. Our experiments using neutralizing antibodies revealed a substantial role for the CLRs, especially MMR and CR3, in the induction of cytokines by C. auris. The role of these receptors in the recognition of mannans is well known^[189]34. However, our binding-affinity data show that r-MMR binds to C. auris mannans with a low affinity, and the in vitro neutralization of MMR led to only a partial loss of cytokine production, suggesting that additional mannan-recognizing receptors contribute to anti-C. auris host defence. By contrast, blocking dectin-1 significantly increased TNF-α and IL-6 after C. auris stimulation for 24 h. This interesting observation could be explained by differences in relative β-glucan abundance. However, Navarro-Arias et al.^[190]31 recently quantified the abundance of different cell wall components and found that total β-glucan and mannans in C. auris were comparable to C. albicans. We suggest that this phenomenon might be due to a combination of two factors: (1) the different exposure of C. auris β-glucan compared with C. albicans ([191]Extended data Fig. 4c); (2) and the differential and variable cell wall adaptation of C. auris strains during the interaction with the host that determines the dectin-1 dependence of the C. auris host response. Such a phenomenon occurs in vivo during C. albicans infection in a strain-specific manner, and the differences in the levels of cell wall chitin influence the role of dectin-1 (ref. ^[192]41). Interestingly, Navarro-Arias et al.^[193]31 reported significantly more cell wall chitin in C. auris compared with C. albicans. As high chitin levels reduce the dependence on dectin-1 recognition, we speculate that differences in C. auris cell wall adaptation to the host, variations in chitin content (higher chitin in C. auris) and differences in cell wall structure (less exposure of β-glucan in C. auris and structurally unique mannoproteins) might provide an explanation, at least in part, for the C. auris dectin-1 (in)dependency. Cytokine induction is important after pathogen recognition, but the induction of phagocytosis is also crucial^[194]42. We observed that C. auris has a higher phagocytic index compared with C. albicans. This is probably due to better recognition of C. auris mannans by immune cells, as cell wall glycosylation is critically important for the recognition and ingestion of C. albicans by macrophages^[195]43. Therefore, to shed more light on these processes, future investigations might examine the phagocytosis dynamics of mutant strains of C. auris that are defective in their cell wall architecture. Interestingly, when the fate of the fungus was assessed through video time-lapse microscopy, it was also clear that the continued cell division of C. auris leads to altered m.o.i. values that are greater than C. albicans and this may also contribute to the stronger stimulation of inflammation. However, this did not result in the death of the phagocytes, most likely due to the lack of hyphae formation by engulfed C. auris cells. In line with previous studies pointing to glucose competition as the main effector mechanism through which C. albicans induces macrophage death^[196]44, a clear enrichment of glucose, fructose and mannose metabolism pathways in the host was observed that was unique for 24 h C. albicans exposure ([197]Fig. 1a and [198]Supplementary Table 2). By contrast, the respective absence of the induction of such a metabolic shift by C. auris suggests that this is a probable mechanism to explain the lower macrophage lysis induced by phagocytosed C. auris, and may subsequently explain the lower virulence. Furthermore, this could represent an important difference in the paths followed by different Candida species as survival mechanisms. The stronger induction of cytokines and lower macrophage lysis after 3 h of phagocytosis might have been expected to lead to decrease the virulence of C. auris in vivo compared with C. albicans. Consistent this hypothesis, experiments in a model of mouse disseminated candidiasis demonstrate that C. auris is less virulent compared with C. albicans, a conclusion that is supported by other recent studies^[199]45,[200]46. Neutrophils are considered to be one of the most important components of the host immune response to fungi through phagocytosis and intracellular killing, or by releasing NETs^[201]47. In a recent study^[202]30, human neutrophils were poorly recruited to sites of C. auris infection, were less able to kill C. auris compared with C. albicans and failed to form NETs in response to C. auris infection. However, neutrophils are important contributors to the host defence against Candida species^[203]48. Our in vivo results show that MPO production in C. auris-infected mice is similar to C. albicans, indicating that neutrophil activation is comparable. Furthermore, the comparable innate and adaptive cytokine production in mice infected with C. albicans and C. auris, as well as the similar or even better organ clearance, suggest that the immune response against C. auris is fully functional in an immunocompetent host. When the cytokine levels were linked to the remaining organ c.f.u. count by a cytokine/c.f.u. ratio, we observed that there was a higher cytokine production per singular C. auris c.f.u. cultured from the organ compared with C. albicans. This is consistent with the potent pro-inflammatory response that was observed in human PBMCs stimulated with C. auris. Future studies are warranted to investigate the relative importance of neutrophils and macrophages in the host defence against C. auris. In conclusion, we performed a comprehensive assessment of the innate host defence mechanisms against the rapidly emerging human pathogen C. auris. The overall conclusion is that the host defence mechanisms induced by C. auris are generally classical antifungal mechanisms, but important specific responses are also triggered by unique C. auris-specific mannoprotein structures. The ensuing immune responses are effective and lead to an effective elimination of the fungus. Our study argues that the intrinsic virulence of C. auris is not higher than other Candida species circulating in the patient population, but it is, rather, the infection-control problem of this pathogen and its high resistance to antifungal drugs that make it dangerous. The challenges that need to be pursued in the coming years include identifying in even more detail the most effective components of the anti-C. auris host defence, and designing and testing novel host-directed therapies to enhance these pathways and improve the outcome to the infection. In this respect, on the basis of our results highlighting the peculiarity of C. auris mannoprotein structures, one promising therapeutic possibility could be fosmanogepix (APX001A), which is a novel agent that targets the fungal protein glycosylphosphatidylinositol-anchored wall transfer protein 1 (Gwt1), thereby inhibiting the maturation and localization of GPI-anchored mannoproteins in the cell wall^[204]20. Several studies reported higher survival rates and lower fungal burden in C. auris-infected mice that were treated with this novel drug^[205]20,[206]49. One of the reasons for this higher efficacy could be the crucial role of mannans for C. auris pathogenicity in the host. In addition to being a comprehensive study of the host immune response to C. auris, our data provide further support from an immunological and microbiological perspective for the development of new and efficient anti-fungal drugs for C. auris that potentially target mannan synthesis. Methods Experimental model and participant details. Ethics statement for ex vivo and in vitro human PBMC stimulations. Inclusion of healthy controls was approved by the local institutional review board (CMO region Arnhem-Nijmegen, 2299 2010/104) and conducted according to the principles of the International Conference on Harmonization—Good Clinical Practice guidelines. Buffy coats from healthy donors were obtained after written informed consent (Sanquin, Nijmegen, the Netherlands). Ethics statement for in vivo mice studies. All of the animal experiments were conducted in the unit of animals for medical scientific purposes of University General Hospital ‘Attikon’ (Athens, Greece) according to EU Directive 2010/63/ EU for animal experiments and to the Greek law 2015/2001, which incorporates the Convention for the Protection of Vertebrate Animals used for Experimental and Other Scientific Purposes of the Council of Europe (code of the facility EL 25BIO014, approval number 1853/2015). All of the experiments were licensed from the Greek veterinary directorate under the protocol number 7467/24–12-2013. All of the animal experiments were reported using the ARRIVE guidelines. Isolation and stimulation of PBMCs. Venous blood from the antecubital vein of healthy volunteers was drawn in EDTA tubes after obtaining written informed consent. PBMC isolation was performed as previously described^[207]50. In brief, the PBMC fraction was obtained using density centrifugation in Ficoll-Paque (Pharmacia Biotech). Cells were then washed twice in PBS and resuspended in RPMI 1640+ medium (RPMI 1640 Dutch modification supplemented with 50 μg ml^−1 gentamicin, 2 mM l-glutamine and 1 mM pyruvate; Gibco, Invitrogen). PBMCs were then counted and resuspended at a concentration of 5 × 10^6 cells per ml. PBMCs (5 × 10^5) were added in 100 μl to round-bottom 96-well plates (Greiner) and incubated with 50 μl of stimulus (RPMI, live, 4% paraformaldehyde (PFA) or heat killed C. albicans yeast (1 × 10^6 per ml) or C. auris (1 × 10^6 per ml); 100 μg ml^−1 purified C. albicans or C. auris mannan; or 10 μg ml^−1 purified C. albicans or C. auris β-glucan) and 50 μl of eventual inhibitor or medium with or without 10% human serum. Serum was complement active unless otherwise indicated, or heat-inactivated by incubation for 30 min at 56 °C in a water bath according to a commonly used protocol. After 1 h of preincubation with inhibitor or medium, stimuli or medium was added. In detail, for receptor blockade experiments, before stimulation with C. albicans or C. auris, PBMCs were preincubated for 1 h with 5 μg ml^−1 anti-DC-SIGN, 10 μg ml^−1 anti-dectin-1 and 10 μg anti-mincle antibodies, and 10 μg ml^−1 control IgG2b; 10 μg ml^−1 anti-dectin-2 antibody and 10 μg ml^−1 of its control IgG1; 10 μg ml^−1 anti-CR3 antibodies and control IgG (R&D); and 10 μg ml^−1 MR-blocking antibodies and 10 μg ml^−1 goat IgG isotype control. After 1 h, cells were stimulated with 10^6 heat-killed C. albicans and C. auris. For the intracellular-pathway blockade experiment 50 nM Syk inhibitor, 1 μM Raf-1 inhibitor or the same concentration of vehicle (DMSO) was used. Concentrations of inhibitors were selected as being the highest non-cytotoxic concentrations. All of the supernatants were stored at −20 °C until analysed. A detailed list of specific antibodies is provided in [208]Supplementary Table 5. Cytokine and lactate measurements. All cytokine levels were measured in the cell culture supernatants using commercially available enzyme-linked immunosorbent assay (ELISA) assays according to the manufacturer’s protocol. For human cytokines, IL-1β, TNF-α, IL-6, IL-1RA and IL-10 were measured after 24 h, and IL-17, IL-22 and IFN-γ were measured after 7 d of stimulation. For mouse cytokines, KC, IL-1β, IL-6, IL-10, IL-17 and IFN-γ were measured after 3 d and 7 d. Lactate was measured using the Lactate Fluorometric Assay Kit (Biovision). A detailed list of the ELISA assays used is provided in [209]Supplementary Table 6. Cytotoxicity assay. PBMCs (5 × 10^5) were added in 100 μl to flat-bottom 96-well plates (Greiner) and incubated with 50 μl of stimulus (RPMI, live C. albicans yeast (1 × 10^6 per ml) or different strains of live C. auris (1 × 10^6 per ml)) for 24 h. Cell viability was assessed using the CytoTox 96 non-radioactive cytotoxicity assay (Promega) according to the manufacturer’s instructions. Released LDH, which is a stable cytosolic enzyme that is released after cell lysis, was measured in the supernatant with a 30 min coupled enzymatic assay. The colour intensity, resulting from the conversion of tetrazolium salt (INT) into ref formazan, is proportional to the number of lysed cells. As a positive control, cells were lysed with 0.5% Triton X-100, reflecting maximal LDH release. Candida strains. C. auris strains from five clades were used (clade I, South Asia; clade II, East Asia; clade III, South Africa; clade IV, South America; clade V, Iran). C. albicans and C. auris strains were prepared by growing cells for 24 h in either a Sabouraud broth or on Sabouraud plates at 30 °C. Unless otherwise indicated, experiments were performed using C. albicans CWZ10061110, C. auris KCTC17810 reference, C. auris CWZ10051893 and C. auris CWZ10051895. Stimulations were performed using either live, heat-killed (12 h at 56 °C) or PFA-killed (4% PFA) microorganisms. Heat killing disrupted the outer layer of the Candida cell wall^[210]34, in turn enhancing β-glucan exposure^[211]35,[212]36. Whereas killing Candida by heat treatment disrupts the outer layer causing exposure of β-glucan in the cell wall, with PFA fixation and thimerosal fixation, the cell wall structure remains intact^[213]34,[214]51,[215]52. Detailed information on the Candida strains examined here is provided in [216]Supplementary Table 7. β-Glucan and mannan isolation from C. auris strains. For the cell wall experiments, a total of eight different C. auris clinical strains were used for glucan and mannan extraction, originating from three different clades (I, II and IV). Isolates were grown in 25 ml YPD (1% yeast extract, 2% dextrose and 2% peptone) at 30 °C for 48 h before parallel isolation of respective cell wall components under identical conditions. To help with β-glucan and mannan collection, cell walls were disrupted by three cycles of freeze–thawing (−20 °C) before the cell pellets were collected. A small aliquot of each culture was plated onto YPD medium to exclude the presence of viable cells, after which the cell pellets were suspended in 10 ml of 0.75 N NaOH at a final concentration of ~3 mg ml−1. Suspensions were heated to 105 °C for 15 min, cooled and separated by centrifugation (10 min at 863g). The supernatant, which contained the mannans, was collected and dialysed against 300 volumes of 18 MΩ water (2,000 molecular weight cut-off). After subsequent collection, the pH was confirmed to be neutral, then the sample was frozen and lyophilized to dryness. For β-glucans, the centrifuged extract was transferred, treated with 10 ml of 1.0 N H[3]PO[4] and heated 105 °C for 15 min. After cooling down, pellets were collected by centrifugation (10 min at 863g) and extracted once more with 10 ml 100% ethanol containing 1% (v/v) H[3]PO[4] at 90 °C for 15 min. β-Glucans were collected by centrifugation and purified by washing three times with 18 MΩ water, after which the final pellet was frozen and lyophilized to dryness. A detailed list summarizing the respective Candida strains from which mannans and β-glucans were isolated is provided in [217]Supplementary Table 8. C. albicans mannan isolation. Mannan isolation was performed using a modification of the Fehling method as previously described^[218]53. In brief, yeast cells, grown overnight at 30 °C, were delipidated by suspending the cell pellets in 100 ml of acetone and incubating for 20 min. The supernatant was decanted after centrifugation for 5 min at 500 r.p.m. and pellets were air-dried for 30 min. To facilitate mannan extraction, the pellets were suspended in 200 ml of distilled H[2]O and supplemented with an equivalent amount of glass beads 0.5 μm to pellet weight, followed by bead beating (3 × 30 s pulses). Cell extracts were autoclaved for 2 h and subsequently centrifuged for 5 min at 5,000 r.p.m. Respective supernatant was split into two; one half was left untreated while the other was treated with 500 mg of pronase for 16 h at 37 °C to abolish glycosidic activity. To ensure the suppression of bacterial growth during pronase treatment, sodium azide was added at a final concentration of 50 nM. The pronase-treated and untreated samples were 1:1 diluted with freshly prepared Fehling solution and allowed to mix for 1 h at room temperature, followed by incubation for 20 min, allowing precipitation of the mannan–copper complexes. After decantation, complexes were treated with 5 ml HCl. When dissolved, 100 ml of a 1:8 mixture of glacial acetic acid:methanol was added for mannan precipitation. After 4 h, samples were washed repeatedly with the 1:8 mixture of glacial acetic acid:methanol until the remaining precipitate appeared colourless (lack of any blue/green) and followed by 3× methanol washes. The final precipitate was dissolved in 100 ml of distilled H[2]O and dialysed against 300 volumes of distilled H[2]O over 48 h and for 12 h against 200 volumes of ultrapure distilled H[2]O to remove any salts, acid, methanol and other low-molecular-mass contaminants remaining from the extraction protocol. Dialysed mannans were frozen and lyophilized for 48–72 h and stored at −20 °C. A detailed list summarizing the respective Candida strains from which mannans and β-glucans were isolated is provided in [219]Supplementary Table 8. RNA purification. PBMCs from three healthy donors (5 × 10^6 cells per ml) were stimulated in flat-bottom 12-well plates (Corning) with freshly counted live C. albicans (1 × 10^6 per ml) and C. auris (1 × 10^6 per ml), and purified cell wall components β-glucans (10 μg ml^−1) and mannans (10 μg ml^−1) isolated from both Candida species as described above. Note that the time between the Candida cell count and the start of the experiment did not exceed 15 min. PBMCs were cultured in the presence of 10% human pooled serum. At 4 h and 24 h, cells were lysed with RLT buffer. Before being processed using the RNeasy Mini Kit (Qiagen), lysates were homogenized using a 1 ml syringe with a 0.8 × 15 mm needle. RNA was subsequently extracted following the manufacturer’s protocol, including an on-column DNase digestion using the RNase-Free DNase Set (Qiagen). Quantification and quality assessment of extracted RNA was performed using the Qubit RNA HS assay (Thermo Fisher Scientific) and Agilent 2200 TapeStation (RNA HS Screentape, Agilent), respectively. The majority of samples that were assessed for quality revealed an RNA integrity number of ≥8. QuantSeq 3′ mRNA-seq. Libraries were generated from the extracted RNA using the QuantSeq 3′ mRNA-Seq Library Prep Kit-FWD (Lexogen) according to the manufacturer’s protocol. Three separate preparations were performed, split by PBMC donor, in turn limiting the number of samples to 14–18 samples per preparation. RNA input was normalized to 100 ng for donor A and to 250 ng for donors B and C. An aliquot (1:10) of double-stranded cDNA libraries was used for quantitative PCR, in turn indicating that 17–18 cycles was optimal for endpoint PCR (17, donor B; 18, donors A and C). Accurate quantification and assessment of quality of the generated libraries was performed using the Qubit dsDNA HS assay (Thermo Fisher Scientific) and Agilent 2200 TapeStation (HS-D1000 ScreenTape, Agilent). The cDNA concentration and average fragment size were used to determine the molar concentration of the individual libraries. Consequently, libraries were pooled equimolar to 100 fmol. After a final dilution of the pool to a concentration of 4 nM, the libraries were sequenced using a NextSeq 500 instrument (Illumina), with a 75 cycle (that is, 75 bp single-end sequence reads) high output kit with a 1.1 pM final loading concentration. Differential gene expression analysis. The quality of the acquired sequencing data was controlled using FastQC tool v.0.11.5 (Babraham Bioinformatics), and the adapter sequences and poly(A) tails were subsequently removed using Trim Galore! v.0.4.4_dev (Babraham Bioinformatics) and Cutadapt v.1.18 (ref. ^[220]54). On average ~6 million reads per individual library were retrieved. Filtered and trimmed reads were mapped to the human reference genome (hg38/GRCh38) using the STAR aligner v2.6.0a^[221]55 ([222]Supplementary Table 4). Less than 1% of all reads comprised of over-represented sequences and were uniquely mapped with a median of 4 million reads (74.1%). After generating gene-level count data using the HTSeq-count tool v0.11.0 (ref. ^[223]56), an additional filtering step was performed to ensure the exclusion of several non-coding RNAs—that is, mtRNA, lincRNA, snRNA, tRNA, miscRNA and snoRNA—in our dataset. Given the absence of sample replicates, PBMC donors were considered to be biological replicates. Thus, in the differential gene expression analysis using DESeq2 v.1.22.0, including logFold Shrinkage and apeglm^[224]57, the average PBMC donor response to the different stimuli were compared to their control condition, RPMI. Genes with a FC ≥ 2 and an adjusted P < 0.01 were considered to be DEGs. We performed a PCA analysis using DESeq2 normalized counts (normTransform) to enable us to identify the main principle components that underpin the majority of the variance. Pathway enrichment analysis. To distinguish between the responses triggered by both Candida species, DEGs were compared between species for the analogous stimulations (live, mannan and β-glucan), and the corresponding time points. This in turn resulted in a group of DEGs that overlap between the two species, and DEGs that were uniquely attributed to either one of the Candida species. Over-representation analyses were performed on all groups per stimulation (and time point) using Consensus PathDB^[225]58, including pathways as defined by pathway databases Kyoto Encyclopedia of Genes and Genomes KEGG^[226]59 and Reactome^[227]60. Minimum overlap in input was set at 2, together with a P-value cut-off of 0.01. For downstream analysis, pathways were considered to be enriched with a corrected P <0.01 (q value). Structural characterization of mannans by NMR spectroscopy. To gain insights into the mannan structure of both Candida species, isolated mannans were analysed using solution-state 1D ^31P-coupled and ^31P-decoupled ^1H NMR spectroscopy and 2D COSY NMR spectroscopy. Data acquisition and subsequent analysis were based on previously described methods^[228]61. In brief, ^1H NMR spectra were collected using a Bruker Avance III 600 NMR spectrometer operating at 331 °K (58 °C). Approximately 10 mg of mannan was dissolved in 600 ml of distilled H[2]O. Chemical-shift referencing was accomplished relative to trimethylsilylpropanoic acid at 0.0 ppm. Proton 1D NMR spectra were collected with 2 dummy scans, 256 scans, 65,536 data points, 20 ppm sweep width centred at 6.2 ppm and 1 s pulse delay. For the 1D ^31P decoupled ^1H NMR experiment, spectra were collected at 333 °K (60 °C) with 2 dummy scans, 1,024 scans, 65,536 data points, 21 ppm sweep width centred at 6.2 ppm and the ^31P decoupling pulse centred at 3.0 ppm. All 1D spectra were processed using exponential apodization with 0.3 Hz line broadening. COSY spectra were collected using 2,048 by 128 data points, 8 dummy scans, 32 scans and 6.0 ppm sweep width centred at 3.0 ppm, and processed with sine apodization in both dimensions and zero-filled to 1,024 data points in f1. Processing was performed using JEOL DELTA (v.5.0.4.4) and Bruker TopSpin (v.4.0.6). MM measurements. To determine the MM of mannans, isolates from C. albicans and C. auris strains were analysed using high-performance gel permeation chromatography (GPC) as previously described^[229]62. Using a Viscotek/Malvern GPC system, consisting of a GPCMax autoinjector fitted to a TDA 305 detector (Viscotek/Malvern). System calibration was achieved using Malvern pullulan and dextran standards. Mannan isolates, ranging from 3 mg ml^−1 to 6 mg ml^−1, dissolved in the mobile phase (50 mM of sodium nitrate, pH 7.3) were incubated for 60 min at 60 °C, then sterile filtered (0.2 μm) and injected into the GPC (100–200 μl). The samples were analysed in duplicate or triplicate and data analysis was performed using Viscotek OmniSec (v.4.7.0.406). Binding interaction of mannans with PRRs rh-dectin-2 and rh-mannose. Mannan binding interactions to the recombinant Dectin-2 and Mannose receptors (R&D systems) were assessed using an Octet K2 BLI instrument (ForteBio) in 10× kinetics buffer (pH 7.4) at 30 °C and 1,000 r.p.m. Increasing concentrations of the respective ligands (3.125–400 μg ml^−1) were used to generate the respective saturation curves, after which the binding affinities were calculated for mannans isolated from C. albicans and C. auris strains as previously described^[230]63. The Ni-NTA biosensor was equilibrated for 3 min before being exposed for 10 min to 0.1 μg ml^−1 His-tagged rh-dectin-2 or rh-mannose receptor proteins and, finally, dissociated for 10 min in 10× kinetics buffer to measure the BLI signal, consistently 20 s after transferring; this was subsequently followed by a series of 8 similar 5 min exposures to an increasing concentration (twofold) of carbohydrate. To control for receptor dissociation during the experiment, a parallel biosensor with the immobilized receptors was placed in the 10× kinetics buffer without respective carbohydrate exposure. Data analysis was performed using GraphPad Prism 7.0 and the dissociation constant K[D] is presented as the mean value with the 95% confidence interval. ROS assay. The induction of ROS was measured by oxidation of luminal (5-amino-2,3-dihydro-1,4-phtalazinedione) and determined in an automated LB96V Microlumat plus luminometer (EG & G Berthold). In brief, PBMCs (5 × 10^5 per well) or neutrophils (2.5 × 10^5) per well were seeded into white 96-well plates and incubated in medium containing either RPMI + zymosan (100 μg ml^−1), heat-killed opsonized C. albicans or C. auris yeast (10^7 c.f.u. per ml). Luminol (20 μl of 1 mM) was added to each well to start the chemiluminescence reaction. Each measurement was performed, at least, in duplicate. Chemiluminescence was determined every 145 s at 37 °C for 1 h. Luminescence was expressed as relative light units per second. The relative light units per second values within the AUC were plotted against time and analysed using Graphpad Prism v.7.0. FITC-labelling of Candida. To label cells with FITC (CAS number, 3326-32-7), 1 × 10^8 per ml of thimerosal fixed cells was sonicated and resuspended in 0.1 mg ml^−1 of FITC in 0.1 M carbonate–bicarbonate buffer (pH 9.6). After incubation for 30 min on a tube roller at 4 °C in the dark, unbound FITC was washed away by centrifugation (3,000 r.p.m. at 4 °C for 10 min) three times in PBS. Before use, cells were resuspended to a concentration of 1 × 10^7 per ml in PBS, aliquoted and stored in the dark at −20 °C. Phagocytosis assay in human cells. To test the uptake of Candida strains by human monocytes, 4 × 10^6 cells per ml of thimerosal-killed FITC-labelled C. albicans and C. auris were preopsonized with 20% human pooled serum for 1 h at 37 °C + 5% CO[2] and subsequently incubated with 2 × 10^5 PBMCs per well (m.o.i., 2:1 fungal:human cells) for either 30 min or 2 h at 37 °C + 5% CO[2]. After the incubation period, cells were gently washed with PBS (1% BSA) and then stained in a total volume of 50 μl using CD14 monoclonal antibodies (Mouse-anti-Human CD14 Pacific Blue, Beckman coulter, clone RMO52, dilution 1:20) for 30 min at 4 °C on ice in the dark. Cells were then washed, and the fluorescence signal of extracellular non-phagocytosed Candida was quenched using 0.1% Trypan blue solution (Sigma-Aldrich; CAS Number 72-57-1). Cells were subsequently measured using a CytoFLEX flow cytometer (Beckman Coulter) and the data were analysed using the Kaluza Analysis software v.2.1. To determine the uptake of C. albicans and C. auris by human monocytes, the percentage of CD14-positive cells that had phagocyted FITC-positive Candida (as a percentage of FITC-positive cells in the CD14-positive population) was calculated. The detailed gating strategy is provided in [231]Extended data Fig. 7a. Phagocytosis assay in BMDMs. Bone marrow was extracted from femurs and tibias of C57BL/6 mice (aged 8 weeks) and differentiated for 7 d with RPMI Medium 1640 Glutamax (Gibco) supplemented with 10% heat-inactivated fetal calf serum, 100 U ml^−1 penicillin–streptomycin and 15% L929-cell-conditioned medium at 37 °C with 5% CO[2]. BMDMs were added to a 8-well μ-Slide (Ibidi) at 0.5 × 10^5 cells per well to adhere overnight. C. albicans and C. auris strains were prepared by growing cells for 24 h in Sabouraud broth at 30 °C, followed by three washes in PBS. Fixed Candida cells were prepared by incubating the Sabouraud-grown yeast overnight at room temperature in 50 Mm thimerosal (Sigma-Aldrich) followed by five wash steps in PBS. Phagocytosis dynamics were determined after adding 3:1 yeast:BMDMs. The intended m.o.i. was calculated by cell count (haemocytometer) and the actual m.o.i. was observed from videos. Live imaging of macrophage interactions with live or fixed C. albicans and C. auris were performed using a Nikon Ti Eclipse microscope with an objective ×20 magnification set to acquire images at 1 min intervals using Volocity (v.6.3, PerkinElmer), with thanks to the University of Aberdeen Microscopy Core Facility. Videos generated from 3 h interactions were analysed to determine over time the proportion of macrophages that phagocytosed yeast (percentage uptake), the number of yeast phagocytosed per 100 macrophages (phagocytic index), the proportion of macrophage death after 3 h (macrophage lysis) and the distribution of yeast contained within individual macrophages. Experiments were performed on three occasions; a total of nine videos were generated for each condition. Phagocytic index data are based on yeasts that were fully inside macrophages. Yeasts adhering but not internalized were not included in cell counts for phagocytic index. Statistical analyses were performed using analysis of variance (ANOVA) with GraphPad Prism (v 7.0). Cell wall staining. Fixed Candida yeast were stained for exposed cell wall β-glucans using Fc–dectin-1 (a gift from G. Brown, University of Aberdeen) and secondary F(ab′)2 anti-human IgG AlexaFluor 488 conjugate (Life Technologies). ConA-Texas Red conjugate (Life Technologies) was used to detect cell wall mannans. Cells were counted, and 2.5 × 10^6 yeast were combined with FACS wash (1% bovine serum albumin and 5 mM EDTA in PBS) with either Fc–dectin-1 at 1 μg ml^−1 or ConA at 25 μg ml^−1. After incubation for 30 min on ice, cells were washed twice in FACS wash, then incubated with secondary F(ab′)2 (for Fc–dectin-1 only) on ice for 45 min, with a further two wash steps. Flow cytometry was performed using an LSR Fortessa cytometer (BD), with thanks to the University of Aberdeen IFCC Core Facility. Data were analysed using FlowJo (v.10.0.8). The detailed gating strategy is provided in [232]Extended Data Fig. 7b. In vivo experimental model of disseminated candidiasis. For the in vivo experiments, C. auris strain 10051895 was selected as representative of clade I, which was the first C. auris clade to be identified and associated with bloodstream infections and high mortality rates^[233]17. Experiments were conducted with a total of 200 C57Bl6 male mice (aged 7–8 weeks), which were purchased from Pasteur Institute (Athens, Greece, EL 25 BIObr 011). Mice were allowed to acclimatize for 7 d before start of the experiments. Animals were housed in cages, with no more than five mice per cage under constant temperature (21 °C) and humidity with a 12 h–12 h light–dark cycle. All of the animals had ad libitum access to food and water. Analgesia was achieved with paracetamol suppositories. Other analgesics were avoided to avoid interactions with the immune system. Healthy mice were i.v. challenged through the tail vein with 1 × 10^7 c.f.u. per mouse log-phase inoculum^[234]45,[235]46 of C. albicans 10061110 (n = 11) and C. auris 10051895 (n = 10) following slight anaesthesia with methoxyflurane (2,2-dichloro-1,1-difluoroethyl methyl-ether in butylated hydroxytoluene 0.01% w/w). Mice were split into groups through a randomization table. Survival was recorded for 14 d; 3 d and 7 d after challenge, mice were euthanized through intramuscular injection of ketamine. To evaluate the fungal burden at day 3 and day 7 after the inoculation of 1 × 10^6 c.f.u. per mouse (n = 4–5 mice per group per time point), we removed, weighed and homogenized the kidneys and livers. The number of fungal counts were measured by serial dilutions of 1:10 in 0.9% saline and expressed as log[10][c.f.u. per g]. For collecting the organs homogenates, at day 3 and day 7 after the injection of 1 × 10^6 Candida c.f.u. per mouse (n = 5–6 mice per group per time point), after a midline incision under aseptic conditions, the entire spleen was removed and segments of the right kidney and of the liver were cut and placed into separate sterile containers. Statistical analysis. Statistical analysis, except where otherwise indicated, was performed using GraphPad Prism 7. All of the experiments were performed at least in duplicate. In experiments with a sample size <4, no statistical testing was performed owing to the small sample size. Datasets with a sample size of >8 were tested for normality using the D’Agostino–Pearson omnibus normality test and, when normally distributed, were analysed using one-way ANOVA with Holm–Sidak multiple comparison test, as specified in corresponding section of the Methods and the figure captions. Given the experimental set-up—in which the same donors were used, yet exposed to different Candida species and strains (stimulations)—measurements were considered to be paired/dependent. Thus, we applied the Wilcoxon signed-rank test for nonparametric matched data to non-normally distributed data. To enhance insights in the C. auris-induced host response, all statistical analyses were performed comparing respective C. auris strains with C. albicans, which we consider here to be the control species. In the case of non-normally distributed data and comparison with multiple control strains, Kruskal–Wallis tests with Dunn multiple comparison were applied. For statistical testing of IL6, IL1B, and IL1RN expression levels, an ordinary one-way ANOVA was performed comparing the log[2]-transformed FC values and corresponding s.e. between the respective C. auris and C. albicans (live, β-glucans and mannans) conditions. Moreover, in vivo mouse data were an exception, for which we performed log-rank tests to assess survival and Mann–Whitney U-tests to assess fungal burden due to sample independence. In all cases, P < 0.05 was considered to be significant. Reporting Summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Extended Data Extended Data Fig. 1 |. Transcriptomic profiling PBMCs stimulated with live C. albicans or C. auris and respective cell wall components β-glucans and mannans for 4 and 24 hours. Extended Data Fig. 1 | [236]Open in a new tab a, Principal component analysis (PCA) performed on normalized count data (normTransform, DESeq2) demonstrates the main component introducing variance in the dataset is time (40%), as indicated by a clear split between the early 4-hour host induced response (left, triangle) and the late 24-hour response (right, circle). To a lower extent (15%), the second component introducing variance appears to be inherent to the stimulus (color). b, At 4 h, PCA reveals a clear donor clustering (shape) irrespective of stimulus (color), indicating the main variance in the early host response reflects inter-individual differences (left). PCA on the late response, 24 h, is predominantly influenced by the respective stimulus (38%, color), and to a lower extent by the donors (19%, shape), indicated by the scattering of stimuli together with a rough clustering amongst donors. c, Pathway enrichment plot displaying the top 20 enriched pathways for both C. albicans live and C. auris live (color) at 24 h. Enrichment determined using Consensus PathDB, including pathways as defined by KEGG and Reactome (shape), considering a p-adjusted value < 0.01 (indicated as ‘q-value’) significant. Size of the geometric points indicates the amount of DEG in relation to the pathways’ size. The exact q values and the data used to make this figure can be found in [237]Source Data Extended Data Fig. 1. Extended Data Fig. 2 |. Comparative LDH secretion, LDH cytokine gene expression and phagocytosis dynamics between C. albicans and C. auris. Extended Data Fig. 2 | [238]Open in a new tab a, Assessment of Candida induced cell death of PBMCs after 24 h stimulation without (RPMI; negative control) or with C. albicans, several C. auris strains originating from all five geographical clades or a positive control (dead cells). Lactate dehydrogenase (LDH) was detected as measure of cell death (Mean ± SEM, n = 6, pooled from two independent experiments). b, Log[2]Fold Change (Log[2]FC) of IL-6, IL-1β, and IL-1RN (encoding for IL-1Ra) gene expression in PBMCs from 3 donors stimulated for 24 h with C. albicans (1006110) and C. auris (KCTC17810, clade II) and their respective cell wall components, β-glucans (left) and mannans (right). Graphs represent Log[2]FC from DEG analysis. * p < 0.05, ** p < 0.01, *** p < 0.001, 1-way ANOVA with correction for multiple comparison. c, The BMDM phagocytic capacity of Thimerosal-fixed C. albicans or C. auris strains in the course of 3-hours. BMDM engulfment depicted as the percentage of macrophages having phagocytosed at least one fungal cell (left), and the phagocytic index, here considered as the number of fungal cells engulfed per 100 macrophages (right); graphs represent mean, n = 9, pooled from at least two independent experiments. d, BMDM phagocytic capacity of Thimerosal-fixed C. albicans or C. auris strains after 1 h. Engulfment is depicted as the percentage of macrophages having phagocytosed at least one fungal cell; graphs represent mean ± SEM, n = 9, pooled from at least two independent experiments. e, BMDM phagocytic capacity of live C. albicans or C. auris strains after 1 h. Engulfment is depicted as the percentage of macrophages having phagocytosed at least one fungal cell. Graphs represent mean ± SEM, n = 9 (n = 7 for C. auris 10051893), pooled from at least two independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001, d 1-way ANOVA with a Holm-Sidak’s multiple comparison test, e Kruskal Wallis test with two-sided Dunn’s multiple comparison. f, Distribution of phagocytosed Thimerosal-fixed fungal cells per macrophage in a 3-hour period, n ≥ 100 observations per condition. Data used to make this figure can be found in [239]Source Data Extended Data Fig. 2. Extended Data Fig. 3 |. Relative C. auris induced ROS production and heat-sensitivity of the cell wall components responsible for the C. auris induced cytokine production. Extended Data Fig. 3 | [240]Open in a new tab a, Neutrophil ROS release after 1-hour stimulation without (RPMI; negative control) or with heat-killed C. albicans, C. auris strains or zymosan (positive control), depicted in relative light units (RLU) either as time-course (left) or as area under the curve (AUC, right), n = 9. b, PBMC ROS release after 1-hour stimulation without (RPMI; negative control) or with heat-killed C. albicans, C. auris strains or zymosan (positive control), depicted in RLU either as time-course (left) or as AUC (right), n = 6. c, TNF-α, IL-6, IL-1β, and IL-1Ra levels in the supernatant of PBMCs after stimulation without (RPMI; negative control) or with heat-killed C. albicans and C. auris from all five geographical clades for 24 h, n = 8. d, PBMC production of cytokines IFN-γ (n = 10; n = 7 for C. auris 10051895), IL-10 (n = 6), IL-17 (n = 6), and IL-22 (n = 14; n = 6 for C. auris 10051893; n = 11 for C. auris 10051895) after stimulation without (RPMI; negative control) or with heat-killed C. albicans and C. auris for 7 days. Graphs represent mean ± SEM, data are pooled from at least two independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p = 0.001, a, b Time curves (left panels) were assessed for statistical differences between C. auris strains and C. albicans by a two-way ANOVA, Area Under curve (AUC) means (right panels) were compared using the two-sided Wilcoxon signed rank test, c, d two-sided Wilcoxon matched pairs signed-rank test comparing respective C. auris strains with C. albicans as control or reference species. Data used to make this figure can be found in [241]Source Data Extended Data Fig. 3. Extended Data Fig. 4 |. Transcriptional changes induced by purified cell wall components and their respective exposure on C. albicans and C. auris surface. Extended Data Fig. 4 | [242]Open in a new tab a, Heatmap displaying the Log[2]Fold change (color scale) of the top 50 DEG of C. albicans live, for both Candida species and their cell wall components, β-glucan and mannan, at 4 h (left panel) and 24 h (right panel). b, Flow cytometry plot based on forward scatter component (FSC) and side scatter component (SSC), demonstrating C. auris strains are slightly smaller and of higher complexity than C. albicans. c, Flow cytometry-based comparison of cell wall components of C. albicans and C. auris strains. Mean fluorescent intensity (MFI) of thimerosal-fixed Candida cells stained for Fc-Dectin-1, a marker for β-glucan (left), and ConA, a marker for mannans (right). Graphs represent mean ± SEM of the 3 means, each performed with three replicates in three independent measurements, * p < 0.05, Kruskall Wallis test with two-sided Dunn’s multiple comparison test was performed comparing the respective C. auris strains with the two C. albicans reference strains. Data used to make this figure can be found in [243]Source Data Extended Data Fig. 4. Extended Data Fig. 5 |. Evaluation of cytokine production upon C. albicans and C. auris mannan stimulation. Extended Data Fig. 5 | [244]Open in a new tab a, PBMC production of cytokines TNF-α, IL-6, IL-1β, and IL-1Ra after 24 h stimulation without (RPMI; negative control) or with purified mannans from C. albicans and C. auris strains in the presence of 10% heat-inactivated human serum, n = 7. b, PBMC production of cytokines IFN-γ (n = 6), IL-17 (n = 9), and IL-22 (n = 9) after 7 days hours stimulation without (RPMI; negative control) or with purified mannans from C. albicans and C. auris strains in the presence of 10% human serum. Graphs represent mean ± SEM, data pooled from at least two independent experiments. * p < 0.05, two-sided Wilcoxon matched pairs signed-rank test, comparing respective C. auris strains with C. albicans as control or reference species. Data used to make this figure can be found in [245]Source Data Extended Data Fig. 5. Extended Data Fig. 6 |. Cytokine levels in plasma and organ homogenates from C.albicans and C. auris-infected mice. Extended Data Fig. 6 | [246]Open in a new tab a, IL-6 production in plasma and supernatants from liver homogenates. b, IFN-γ production in supernatants from kidney and spleen homogenates. c–e, IL-1β (c), IL-17 (d), and IL-10 (e) production in plasma and supernatants from liver, kidney, and spleen homogenates. Mice have been infected i.v. with 1×106 c.f.u. of C. albicans or C. auris. Graphs represent mean ± SEM, n = 6 per group per time-point pooled from two independent experiments. Data used to make this figure can be found in [247]Source Data Extended Data Fig. 6. Extended Data Fig. 7 |. Applied gating strategies across flow cytometry experiments. Extended Data Fig. 7 | [248]Open in a new tab a, Gating strategy for FITC-labelled Candida in PBMCs (linked to [249]Fig. 2b). All events were plotted based on forward scatter (FS) and side scatter (SS) characteristics. In the upper plot (2.1) the region of cells positive for FITC-Candida was highlighted (green gate) while in the bottom plot (2.2) CD14 positive cells are represented (red gate) gated within the total PBMCs population (1). Within the CD14 + cells selection, the amount of phagocytosed FITC positive Candida was examined by plotting (3) the FITC signal against the CD14-PB450 signal (blue gate) and the percentage of cells and mean fluorescent intensity (MFI) were used for analysis. b, Gating strategy for Thimerosal-fixed Candida cells stained for either β-glucan using Fc-Dectin-1 or ConA as marker for mannans ([250]Extended Data Fig. 4c). Supplementary Material Suppl. Fig. 1 [251]NIHMS1810177-supplement-Suppl__Fig__1.pdf^ (1.3MB, pdf) Suppl. Fig. 3 [252]NIHMS1810177-supplement-Suppl__Fig__3.pdf^ (1.3MB, pdf) Suppl. Fig. 2 [253]NIHMS1810177-supplement-Suppl__Fig__2.pdf^ (1.4MB, pdf) Suppl. Fig. 5 [254]NIHMS1810177-supplement-Suppl__Fig__5.pdf^ (876.8KB, pdf) Suppl. Fig. 6 [255]NIHMS1810177-supplement-Suppl__Fig__6.pdf^ (1MB, pdf) Suppl. Fig. 4 [256]NIHMS1810177-supplement-Suppl__Fig__4.pdf^ (1.2MB, pdf) Supple. Fig. 7 [257]NIHMS1810177-supplement-Supple__Fig__7.pdf^ (1,023.7KB, pdf) Suppl. Tables [258]NIHMS1810177-supplement-Suppl__Tables.pdf^ (10.4MB, pdf) Suppl. video 1 [259]Download video file^ (20.5MB, mp4) Suppl. video 2 [260]Download video file^ (24.5MB, mp4) Suppl. video 3 [261]Download video file^ (24.2MB, mp4) Suppl. video 4 [262]Download video file^ (24.2MB, mp4) Acknowledgements