Mucormycosis is one of the most aggressive fungal infections with high mortality and an increase mobility in recent decades, especially during the pandemic of COVID-19, which has become a public health problem (1, 2). The members of the order Mucorales such as Rhizopus, Mucor, and Lichtheimia species are responsible for most cases while Cunninghamella, Apophysomyces, Saksenaea, Rhizomucor, Cokeromyces, Actinomucor, and Syncephalastrum cause fewer infections (3). Invasive mucormycosis usually occurs in individuals underlying medical comorbidities or immunosuppression (4, 5). Diabetes is a major risk factor and immunocompromised patients undergoing chemotherapy, cancer immunotherapy, hematopoietic stem-cell transplantations and solid organ transplantations are more susceptible to this infection (4, 5). In addition, glucocorticoid treatment, trauma, iron overload and malnourishment are also associated with the increased incidence of mucormycosis (4, 5).

The forms of mucormycosis mainly include rhino-orbital-cerebral, pulmonary, cutaneous and soft-tissue, gastrointestinal and renal mucormycosis (6, 7). Lungs are the third main location after the rhino-orbito-cerebral areas and skin (8). Pulmonary mucormycosis usually progresses rapidly and carries a high mortality rate (9). Liposomal amphotericin B is first-line therapy of antifungal agents and the step-down treatment includes oral isavuconazole or posaconazole. It is often difficult to cure by using drug therapy alone so surgical intervention is strongly recommended to improve prognosis and it is also critical to correct risk factors (6, 9). Nevertheless, the delayed diagnosis and limited pharmacological treatments result in poor survival rates (9). Therefore, it is necessary to improve our understanding of host immune response during mucormycosis, which can help exploring new approaches for prevention and antifungal therapies.

To date, the pathogenesis and immune mechanisms of pulmonary mucormycosis are not fully understood and remain unexplored. Mucormycosis is characterized by extensive tissue invasion and destruction and is the most angioinvasive of all fungal infections (1). Spores of Mucorales are dispersed by air and taken up by humans via inhalation; once the host immune function is compromised, these inhaled fungi are dreaded to cause fatal disease (10). After Mucorales invasion, the innate immune system works as the first defense line to prevent spore germination, including host barrier cells, alveolar macrophages, neutrophils and natural killer (NK) cells (1, 11). Although adaptive immunity is also activated, there is limited evidence for a critical role of it in combating mucormycosis (11). Host iron acquisition is another important factor for progression of mucormycosis (11). Given the increased incidence and limited therapeutic options, it is warranted to further explore the immune response in pulmonary mucormycosis to shed light on novel therapeutic interventions.

Assessing the transcriptomes and proteome during infection is a straightforward strategy to provide novel insights on the immune landscape of pulmonary mucormycosis. Previous studies have investigated the transcriptomic profile of COVID-19 associated mucormycosis (CAM) and reported that deficient phagocytosis, cellular iron overload and a hyperglycaemic state contribute to the progress of mucormycosis (12, 13). There are also a few researches focused on animal models (fruit fly, mouse and zebrafish models) of mucormycosis and infected cell models such as airway epithelial cells, human peripheral blood mononuclear cells (PBMCs) and macrophages (AMs), which revealed the antifungal immune response from different aspects (14–21). However, there is a paucity of studies on the pulmonary immune response of mucormycosis patients. As for proteomics, there is also a lack of related research. Thus, it is of great value to investigate the transcriptome and proteome of pulmonary mucormycosis.

In this study, we enrolled patients with pulmonary mucormycosis who underwent surgery and aimed to investigate immune landscape of infected lung tissues using transcriptomic and proteomic approaches. Transcriptomic analysis revealed that fungal infection induced a widespread up-regulation of genes associated with immune response and significant enrichment of multiple signaling pathways related to both innate and adaptive immunity. Nevertheless, we observed limited immune-associated proteins increased in proteomic profiling, suggesting compromised host immune function in protein level. Collectively, our results suggest impaired activation of host immune defense during pulmonary mucormycosis. The identification of altered transcripts and proteins could contribute to a better understanding of immune mechanisms and pathophysiological process and provide new perspectives for the prevention and treatment of mucormycosis.

In this study, we described transcriptome and proteome of pulmonary mucormycosis patients to explore host immune landscape for the first time. Previous studies mainly focused on the transcriptome profiles of COVID-19-associated mucormycosis patients as well as cellular and animal models of mucormycosis. To date, there is a lack of proteomic investigations on mucormycosis. Our results highlighted the immune response in infected lung tissues intuitively and provided a new perspective for the pathogenesis of pulmonary mucormycosis.

It is well known that diabetes mellitus constitutes an important risk factor for susceptibility to mucormycosis (26). All five patients enrolled in this study had diabetes mellitus and two patients had ketoacidosis, which might be a potential cause for their infection. Hyperglycemia induces upregulation of glucose-regulated protein 78 (GRP78), advanced glycation end-products (AGE) and availability of iron as well as impairs phagocytic capacity and other microbicidal activity of immune cells (27–30). In this study, we employed a self-paired control design to explore immune mechanisms of pulmonary mucormycosis beyond established risk factor. Furthermore, the KEGG pathway of primary immunodeficiency was enriched among the infected group (Figures 2B, E; Supplementary Tables S4, S15), suggesting that in addition to diabetes, there may be underlying immunologic factors due to pathogen infection that impair the immune response. The underlying mechanisms warrant further investigation, as such immunopathogenic effects potentially contribute to the lack of clinical improvement in certain patient populations.

A notable finding of our study was the discordance between transcriptomic and proteomic immune responses in pulmonary mucormycosis. It was revealed that transcriptomics analysis showed activation of host immune defense function at the transcriptional level, whereas proteomic profiling suggested there were deficiencies in host immune response against infection at the protein level. These results are consistent with previous researches on mucormycosis. A study on CAM patients reported deficient phagocytosis in circulating monocytes with a significantly down-regulated of pathways related to endocytic pathways, phagosome maturation and the cytoskeletal regulation of phagocytosis (12). Similarly, another transcriptomic study demonstrated immune derangement in COVID-19 associated pulmonary mucormycosis (CAPM) patients (13). Similar phenomena were also observed in in vitro and in vivo models: R. oryzae infection induced a down-regulation of many PRRs genes and Toll-like receptor-signaling pathway in peripheral blood mononuclear cells (PBMCs) compared to C. albicans and A. fumigatus infection, and completely impaired the ability to generate reactive oxygen species (ROS) (15); in the drosophila melanogaster model, Zygomycetes infection also down-regulated several genes involved in pathogen recognition and immune defense compared to Aspergillus infection (16). Our results of proteomic profiling also revealed downregulation of many immune-related genes and pathways, collectively demonstrating compromised host immune defenses in mucormycosis. These results together explain the possible reason why mucormycosis is more aggressive and difficult to cure than other fungal infections.

Similar transcript–protein discordance has also been reported in some other multi-omics studies on infectious diseases. For instance, a tomato spotted wilt virus infection model in insect gut tissues showed discordance between DEGs and DEPs, indicating distinct and dynamic regulatory mechanisms of transcript and protein abundance (31). In human infection contexts, DEGs and DEPs exhibited opposing regulatory patterns in immunocompromised populations following SARS-CoV-2 infection; the host responses in SARS-CoV-2-infected lung epithelial cells also revealed overlapping yet distinct immune signatures across the transcriptome and proteome (32, 33). In addition, complex host immune dysregulation reflected by transcriptomic and proteomic signatures was also reported in children with severe malarial anemia (34). Mechanistically, although RNA-seq is sensitive for transcriptional activation, protein abundance is regulated not only by transcription but also by post-transcriptional, translational and protein degradation regulation (35). For example, in a multi-omics study on macrophage activation model, protein secretion could be independent of transcription due to processes such as ectodomain shedding and proteolytic processing, leading to the protein-level responses might lag behind or diverge from transcript-level changes (36).

In our study, this transcript–protein discordance was further supported through integrative analyses that transcriptomic and proteomic findings were not completely contradictory but rather exhibited a partial discordance (Figures 3D, 4G, H; Supplementary Tables S12, S14). On the transcriptomic layer, broad induction of immune-defense programs was observed, including iron-homeostasis mediators (e.g., HAMP, HP, TF/TP, LTF) and multiple pathogen-sensing/immune-regulatory genes (e.g., PRR, cytokines, chemokines), along with pathways related to innate and adaptive immunity. In contrast, proteomic profiling showed mainly down-regulation of immune- and barrier-related pathways. Moreover, integrative overlap analyses indicated the shared down-regulated signals were related to structural and signaling modules (e.g., tight junction, focal adhesion, and ECM–receptor interaction), suggesting lung tissue damage and compromised barrier integrity during pulmonary mucormycosis. Collectively, these results suggested that although transcriptomic profiling indicated an attempted immune activation, the protein-level execution of host defense might be constrained in severely infected lung tissues, consistent with compromised barrier integrity and impaired immune responses during pulmonary mucormycosis.

In mucormycosis, severe tissue destruction, hypoxia, metabolic dysregulation and fungal immune-modulatory strategies may further compromise protein synthesis or stability in infected lung tissues (37, 38). The possible mechanisms include severe tissue destruction and disrupted cellular composition at infection sites, strong post-transcriptional/translational regulatory constraints during overwhelming infection and enhanced protein degradation. Collectively, these mechanisms could contribute to activated transcriptome but insufficient effector proteome, which was consistent with impaired host defense during pulmonary mucormycosis. Therefore, we speculated that during pulmonary mucormycosis, transcriptional upregulation might reflect an attempted or compensatory host response, but failed to translate into effective protein-level responses, which were required for effective host defense against mucormycosis. Given that the patients in our study required surgical intervention after failure of systemic antifungal therapy, the observed suppression of immune-related proteins could reflect the pathophysiological state of pulmonary mucormycosis. The failure to increase expression of functional immune-related proteins might contribute to a critical mechanism underlying disease severity and poor clinical outcomes.

Different from proteomic findings, our transcriptomic profiling showed a broad activation of immune-related genes and pathways. For instance, cytokine/chemokine-associated signaling was induced; pathogen-sensing and innate immune regulation were activated, such as increased transcripts of PRRs. Also, adaptive immune activation was reflected by enrichment of B- and T-cell related pathways, together with upregulation of some lymphocyte markers and regulators. Notably, iron-homeostasis emerged as an important signature, exemplified by marked upregulation of HAMP along with HP, TF/TP and LTF. Collectively, these transcriptomic findings indicated a broad “attempted” immune activation program in infected lung tissues at the mRNA level, which provides an important context for interpreting the limited or suppressed immune-protein responses observed by proteomics. These results were consistent with studies focusing on transcriptome of macrophages, epithelial cells as well as murine and zebrafish infection models (14, 18–21). While a subset of these up-regulated genes and pathways contributes to antifungal immunity, the functions and mechanisms of many remain uncharacterized.

Pulmonary mucormycosis is initiated by inhalation of Mucorales spores. In the healthy host, the first-line immune barrier relies on airway/alveolar epithelial cells, resident alveolar macrophages, recruitment of neutrophils as well as cytokine/chemokine production that modulates innate immune functions and subsequent adaptive immune responses (1, 9). As we all know, pulmonary mucormycosis is an opportunistic mycosis that infects hosts with qualitative or quantitative defects in immunity, including patients with severe neutropenia, recipients of corticosteroids or other immunosuppressive medications, poorly controlled diabetes mellitus and those with iron overload states (39). Consequently, therapeutic strategies aimed at reversing host immune compromise could potentially improve survival outcomes of patients with severe infections.

Iron homeostasis plays an important role in antifungal immunity, particularly in mucormycosis (40–42). Prior studies have reported serum unbound iron is a critical factor for mucormycosis susceptibility in diabetic ketoacidosis (DKA) patients and limiting iron utilization by fungus is beneficial for host defense (14, 30, 43). Our transcriptomics analysis found several DEGs (HAMP, HP, TP and LTF) involved in iron homeostasis was up-regulated. HAMP was the most significantly up-regulated DEGs, which encode hepcidin and play a central role in iron metabolism. Similarly, a previous study has found iron-related genes were differentially expressed in infected macrophages and the association between HAMP gene polymorphisms and CAPM susceptibility has been investigated (14, 44). Hepcidin was also strongly induced in murine models of Candida and Aspergillus infection and it contributed to host defense against Candida albicans infection (45, 46). Haptoglobin encoded by HP, transferrin encoded by TP and lactoferrin encoded by LTF, can also restrict iron acquisition by fungal pathogens. Haptoglobin is up-regulated in paracoccidioidomycosis, aspergillosis and fusarium keratitis as well (47–49). Transferrin has antimicrobial activity against several bacterial and fungal pathogens, such as Candida albicans and Histoplasma capsulatum (50–53). Lactoferrin has already been found antifungal effects against mucor species in combination with Amphotericin B (54, 55). The precise functional role of these iron-related proteins in mucormycosis and other fungal infection diseases needs further exploration. Our results indicated a discordance between transcriptional upregulation and protein expression of these iron-metabolism proteins and supplementing these antibacterial-active proteins may constitute a novel adjunctive antifungal strategy (56).

Recognition of fungal pathogens by PRRs is critical in host immunity. Previous research has identified altered expression of dectin-1, TLR2, and TLR4 in circulating monocytes of mucormycosis patients and demonstrated that TLRs were essential for pro-inflammatory responses against mucor species (57–59). Our results showed elevated transcripts of four PRRs (TLR10, ZBP1, CLEC17A and AIM2). The genetic polymorphism of TLR10 was associated with susceptibility to aspergillosis (60, 61). ZBP1 was necessary for inflammatory PANoptosis induced by C. albicans and A. fumigatus infection (62, 63). CLEC17A, a member of C-type lectin receptor (CLR) family, might be capable of recognizing fungal β-glucans, but its role in fungal infections remains uninvestigated (64). AIM2 was beneficial for host defense against Aspergillus infection but enhanced Candida infection (65, 66). It was also up-regulated in Malassezia folliculitis (67). In mucormycosis, the functions of these PRRs remain unknown and warrant further mechanistic investigation.

Cytokines and chemokines promote inflammatory response and recruitment of immune cells during infection. Cytokine-cytokine receptor interaction was significantly enriched in our KEGG analysis (Figures 2A, B). GO analysis also showed a number of terms related to cytokines enriched, including cytokine-mediated signaling pathway, regulation of cytokine production, cytokine production, positive regulation of cytokine production, cellular response to cytokine stimulus, response to cytokine, cytokine-mediated signaling pathway (Supplementary Table S5). Previous studies have found a few cytokines and chemokines induced during mucormycosis infection, such as TNF-α, interferon-γ, GM-CSF, interleukins, and so on (37, 68–70). Our data found elevated expression of several transcripts in chemokine family, including CCL11, CXCL13, CCL19 and its cognate receptor CCR7, which contribute to the recruitment of eosinophils and lymphocytes. CCL11(eotaxin) serves as a major chemoattractant for eosinophils and Th2 cells. It could be induced by A. fumigatus, which was associated with allergic airway responses and probably participated in the pathogenesis of allergic bronchopulmonary aspergillosis (71–73). Notably, some case reports have documented allergic-like manifestations in mucormycosis, but the underlying mechanisms are unknown (74–78). The expression of CXCL13 was also up-regulated in Pneumocystis and Candida albicans infection, though its precise function remains incompletely understood (79, 80). CCL19 expression was elevated during Aspergillus infection and its receptor CCR7 was also up-regulated in Aspergillus and Paraccocidium brasiliensis infection; however, CCR7 deficiency reduced host susceptibility to aspergillosis (81–89). In summary, the functions of these chemokines remain not fully defined in mucormycosis.

Beyond chemokines, transcriptional upregulation was observed in some members of interleukin receptor family (IL2RA and IL17REL) and TNF receptor superfamily (TNFRSF7, TNFRSF9, TNFRSF13C and TNFRSF17). Current studies have documented IL-2 and IL-17 up-regulated during mucormycosis and they strengthened host defense (90–92). For instance, IL-2-stimulated natural killer (NK) cells could damage mucormycetes (92). In murine mucormycosis model, IL-17 signaling contributed to pathogen clearance and antifungal immunity. Concordantly, our KEGG enrichment analysis also found significant upregulation of IL-17 signaling pathway (Figures 2A, B). TNFRSF regulates a spectrum of immune cell responses. TNFRSF9 (CD137) and TNFRSF13C (BAFFR) exerted protective roles in Candidiasis and Pneumocystis pneumonia, respectively (93–95). Moreover, our analysis showed some interferon-related genes (IRF4, IFITM10, IFRD1 and SYNDIG1) significantly up-regulated. Interferon-γ has been found to improve fungal clearance in vitro and murine models of mucor infection and a lot of clinical case reports have documented interferon as adjunctive antifungal therapy in mucormycosis (69, 96–107). In summary, the mechanisms by which these cytokines regulate anti-fungal immune responses in mucormycosis remain incompletely understood. Given our proteomic findings demonstrating compromised host immune function at infection sites, supplementation of effective cytokines may constitute a potential therapeutic strategy to enhance host immunity against fungal infection. Consequently, further research is warranted to develop novel adjunctive immunotherapeutic approaches.

Currently, whether adaptive immunity plays a critical role in mucormycosis remains uncertain. Our data showed a number of up-regulated DEGs and enriched pathways related to adaptive immune activation (Table 2; Figure 2). For instance, T cell receptor signaling pathway was enriched in KEGG analysis (Figures 2A, B). Notably, the expression of CTLA4, a coinhibitory receptor for T cell activation, was also up-regulated in other fungal infections, including paracoccidioidomycosis, candidiasis and Lacazia loboi infection (108–112). Blocking CTLA-4 confered protective effects in Paracoccidioides brasiliensis and Cryptococcus neoformans infections as well as fungal sepsis (Candida infection) (113–118). Similarly, blockade of PD-1/PD-L1 pathway demonstrated therapeutic efficacy in a murine model of pulmonary mucormycosis (119). These results suggest that immune checkpoint inhibitors (ICIs) could enhance T cell immune response to improve antifungal defense. CD28, a critical co-stimulatory molecule of T cells, is associated with diverse fungal infections. Low CD28⁺CD8⁺ T-cell counts were associated with mortality of invasive fungal infection, such as candidiasis and aspergillosis (120–122). In vitro studies demonstrated CD28 participated in immune responses against Blastomyces dermatitidis, Pneumocystis and Paracoccidioides brasiliensis infections (123–127). However, its role in mucormycosis remains unexplored. Moreover, our KEGG enrichment analysis also revealed Th17 cell differentiation pathway up-regulated (Figures 2A, B) and a study on CAM patients revealed that reduced Th17 cell frequency potentially correlated with susceptibility to mucormycosis (128). Our results of KEGG analysis showed B cell receptor signaling pathway was up-regulated as well. CD19 activate B cell immune response and modulate humoral immunity. However, CD19⁺ B cells were found reduced in CAM patients (129). Whether they exert antifungal roles requires further exploration.

In summary, our transcriptomic profiling revealed a broad upregulation of transcripts related to immune and inflammatory response, a substantial portion of which remain uncharacterized in fungal infections and need further investigation. Nevertheless, proteomic analysis demonstrated that the majority of proteins involved in immune defense failed to be induced. Only a limited subset of proteins was up-regulated, mainly enriched in innate immune pathways, such as phagosome, neutrophil extracellular trap formation and lysosome, suggesting their potentially critical roles in fungal clearance.

Because of the difficulty in obtaining lung specimens from healthy population, we took a self-controlled study design, which had intrinsic limitations and could not fully reflect normal pulmonary status. Due to pulmonary mucormycosis is a rare opportunistic infection, the number of cases that could be enrolled was inevitably limited. Our study only included surgical patients, who had severe disease manifestations and poor response to systemic antifungal therapy alone. Thus, our results could not represent the whole population of pulmonary mucormycosis. Additionally, the substantial clinical heterogeneity of mucormycosis should be considered, such as variations in infection sites and Mucorales species. Our study population had a single risk factor, diabetes mellitus, which further limited the generalizability of our results. Importantly, due to the limited availability of surgically resected human lung tissues, additional functional validation experiments were not feasible. Also, the limited sample size and heterogeneity of human surgical specimens reduced statistical power for single-gene discovery after multiple-testing correction. Therefore, raw p-value–based DEGs and enriched pathways were used for hypothesis-generating analyses, with FDR values provided as a sensitivity analysis. In conclusion, further studies with larger sample sizes, independent validation approaches and experimental models are warranted to identify the functions of DEGs and DEPs. Despite these limitations, our work provided deeper characterization of host immune response signatures during pulmonary mucormycosis, which may contribute to development of novel therapeutic strategies.

Our study investigated the transcriptome and proteome of pulmonary mucormycosis for the first time. Transcriptomic profiling found upregulation of numerous genes and pathways related to immune defense, whereas proteomic analysis revealed dysfunction of antifungal immunity, collectively indicating a failed activation of host immune response in pulmonary mucormycosis.