Humoral Immunity Against Aspergillus fumigatus

Abstract

Aspergillus fumigatus is one the most ubiquitous airborne opportunistic human fungal pathogens. Understanding its interaction with host immune system, composed of cellular and humoral arm, is essential to explain the pathobiology of aspergillosis disease spectrum. While cellular immunity has been well studied, humoral immunity has been poorly acknowledge, although it plays a crucial role in bridging the fungus and immune cells. In this review, we have summarized available data on major players of humoral immunity against A. fumigatus and discussed how they may help to identify at-risk individuals, be used as diagnostic tools or promote alternative therapeutic strategies. Remaining challenges are highlighted and leads are given to guide future research to better grasp the complexity of humoral immune interaction with A. fumigatus.

Introduction

Aspergillus fumigatus is an ubiquitous saprophytic mold, and it reproduces asexually forming spores (called conidia) [1]. When inhaled, these airborne conidia are rapidly cleared by the mucociliary function and resident macrophages without stimulating any undue immune response [2]. However, in certain patient populations with immunosuppression and underlying lung disease, this fungus can cause acute and/or chronic conditions, collectively called aspergillosis, with three main entities [2]. Invasive pulmonary aspergillosis (IPA) occurs in severely immunocompromised patients including hematopoietic stem cell and solid organ transplant recipients, with neutropenia or receiving heavy chemotherapy and/or corticosteroids for specific hematological malignancies and solid tumors [3]. Recently, IPA has been increasingly described in critically ill patients with viral acute respiratory distress syndrome [4, 5]. Chronic pulmonary aspergillosis (CPA) is responsible for various clinical presentations ranging from asymptomatic aspergilloma in patients with preexisting lung cavities (e.g., tuberculosis sequalae) to cavitary forms with an important clinical impact on general status in patients with underlying lung diseases [6]. CPA may evolve into lung fibrosis if untreated and/or subacute invasive forms when immunosuppressive factors are added [6]. Finally, hypersensitivity to Aspergillus antigens leads to allergic bronchopulmonary aspergillosis (ABPA), a progressive deleterious lung disease complicating asthma and cystic fibrosis [2]. High mortality of IPA (around 50% at 1 month) and CPA (around 50% at 5 years) [7] as well as diagnosis challenges and emergence of resistance made the World Health Organization to rank A. fumigatus a critically risky pathogen in the recent fungal priority pathogen list [8].

Understanding aspergillosis and its spectrum of diseases relies, therefore, mainly on understanding host immunity, whether insufficient or exacerbated. The host immune system is composed of both cellular and humoral components. While the cellular immunity (involving neutrophils, macrophages, dendritic cells and epithelial cells to some extent) has been extensively studied [9, 10], the humoral immune function has been understudied. Indeed, adequate cellular response may not occur in the absence of humoral components. In general, humoral components facilitate microbial phagocytosis by coating (opsonizing) pathogens, prevent pathogen entry into non-immune cells by neutralizing them, and may act as anaphylatoxins recruiting immune cells. Furthermore, humoral components can have a direct inhibiting impact on microbial growth and metabolism, up to direct killing.

Before exposing host-fungal interactions involving humoral immunity, it is necessary to understand the complex life cycle of A. fumigatus. Conidia, the infective propagules of this fungus, swell and germinate in enabling environments. The hyphae produced may then differentiate into conidiophores baring conidia prone to be dispersed. Each stage involves dynamic changes in their cell wall structure, mainly composed of polysaccharides and proteins [11, 12]. Importantly, A. fumigatus cell wall constitutes the first interface between host immune components and the fungus [13]. This review intends to summarize the available knowledge on major players of humoral immunity against A. fumigatus, how they may help in identifying at risk hosts for aspergillosis, could improve aspergillosis diagnosis, and can be exploited in therapy against aspergillosis. We also highlight challenges/research leads in humoral immunity that are being explored to bridge the gaps in A. fumigatus pathogenesis.

The Major Players of Humoral Immunity Against A. fumigatus

The major humoral components are the complement system, antimicrobial peptides, collectins, acute phase proteins and circulating antibodies. All studied components are summarized in Table 1.

Table 1.

Main characteristics of all soluble mediators studied against A. fumigatus.

Humoral factor	Family	Ligand	Status during aspergillosis	Specific biological role in anti-Aspergillus immunity	Reference
C1q	Complement system	NA	↓ in BAL ↑ in CNS	To be studied	[19, 41, 127]
C3b	Complement system	rodA (conidia) β-glucan galactomannan (hyphae)	NA	Opsonizes conidia and hyphae. Amplifies complement cascade. Increases phagocytosis and modulates immune response	[15, 20, 21]
C5a	Complement system	NA	NA	Recruits neutrophils at the site of infection	[24, 25]
Cathelicidin (LL-37)	AMP	NA	↑ in BAL	Discrepant results in the literature	[41, 56, 64, 65]
CRP	APP	NA	↓ in BAL	Promotes conidia phagocytosis by neutrophils	[41, 72]
Defensin (α-)	AMP	NA	NA	Fungicidal effect on swollen conidia and hyphae	[58, 61]
Defensin (ß-)	AMP	NA	NA	Inhibition of germination	[56]
Ficolin-1	Collectin	Chitin β-1,3-glucan	↑ in BAL	Opsonizes hyphae increase IL-8 secretion by lung epithelial cells	[40, 41]
Ficolin-2	Collectin	GlcNAc GalNAc	↓ or ↑ in BAL	Opsonizes conidia and decrease pro-inflammatory cytokines secretion by hMDM and neutrophils	[, 41, 43]
Ficolin-3	Collectin	NA	↓ or ↑ in BAL	May opsonize conidia and increase IL-8 secretion by lung epithelial cells	[29, 41, 44]
Galectin-3	APP	NA	↓ in BAL	Binds all morphotype of A. fumigatus and enhances neutrophil motility and their extravasation into the airway at the site of infection	[41, 75]
Histatin-5	AMP	NA	NA	Inhibition of hyphae metabolism	[57]
Histones	AMP	NA	NA	Major components of neutrophil extracellular traps. Inhibition of hyphae metabolism	[56]
Lactoferrin	AMP	NA	↑ in BAL	Iron depletion. Inhibition of germination	[41, 56, 57]
Lysozyme	AMP	NA	NA	Inhibition of hyphae metabolism	[56, 59, 60]
MASP-1/3	Complement system	NA	NA	Recruit ficolin-3 to Af, activate complement and facilitate phagocytosis	[28]
MASP-2	Complement system	NA	↓ in BAL	To be studied	[41]
MBL	Complement system Collectin	rodA (conidia) Galactomannan (hyphae)	↑ in cornea ↑ in serum	PRR binding to A. fumigatus and activating complement system through the lectin-pathway	[36, 48, 128]
PTX-3	APP	NA	↑ in serum and BAL	Opsonization promoting phagocytosis and killing by neutrophils	[74]
SAP	APP	NA	= in serum ↑ in BAL	Activates complement and enhances phagocytosis by neutrophils	[73]
SP-A	Collectin	NA	↓ in BAL	Opsonization, phagocytosis, dampening Th2 response	[53]
SP-D	Collectin	Melanin (conidia) Galactomannan GAG (hyphae)	↓ in BAL	Opsonization, phagocytosis, triggers pro-inflammatory response, fungistatic effect	[41, 48, 49]
Ubiquicidin	AMP	NA	NA	Inhibition of hyphae metabolism	[57]

Af Aspergillus fumigatus, AMP antimicrobial peptide, APP acute phase protein, BAL bronchoalveolar lavage, CNS central nervous system, GAG galactosaminogalactan, GlcNAc N-acetyl glucosamine MASP MBL-associated serine protease, MBL mannose binding lectin, GalNAc N-acetyl galactosamine, NA not assessed to the best of our knowledge, PRR pattern recognition receptor

The Complement System

The complement system is composed of several components. Activation of the complement cascade occurs through three major routes: classical, alternative and lectin pathways. The final outcome of these molecular cascades is the formation of the Membrane Attack Complex (MAC), being inserted in the microbial membrane leading to microbial lysis. However, the thick cell wall of A. fumigatus prevents MAC formation [14, 15]. Depending on the fungal morphotype the complement pathways will be activated. Dormant conidia, which are superficially covered by rodlet and melanin pigment layers, trigger alternative pathway while the exposure of cell wall polysaccharides as the conidia germinate progressively triggers the classical/lectin pathway (Fig. 1) [16, 17, 18]. Many steps of the complement cascade before MAC formation represent central tools in anti-Aspergillus host defense, especially in recruiting neutrophils [16]. Mice deficient for complement components (e.g. C3, C5, C1q) present a higher mortality rate and more severe pulmonary aspergillosis [15, 19, 20, 21]. Recognition by the complement system and activation of the cascade seems to interfere with fungal dissemination, supported by the fact that the level of complement deposition on different Aspergillus species correlates inversely with their pathogenicity [17]. A. fumigatus and A. flavus, the most virulent species, bind less C3 on their surface than non-pathogenic species [17]. Some specific complement components have been further studied to understand their importance in anti-Aspergillus immunity.

Fig. 1.

The three complement system pathway activated by Aspergillus fumigatus. While resting conidia can only activate the alternative pathway, hyphae can activate all three pathways. Created with BioRender.com

C1q

C1q is the target recognition protein of the classical complement pathway which indirectly recognizes pathogen through bound antibodies (Abs). C1r and C1s are then recruited to form C1 complex, necessary to activate C4 and C2 leading to C3 convertase [18]. C1q collectin-like structure, including a carbohydrate recognition domain could bind directly to Aspergillus cell wall polysaccharides and modulate immune response through a complement-independent pathway. Although not fully studied, C1q could play a major role in anti-Aspergillus immunity as suggested by the increased susceptibility of C1q^−/− mice to A. fumigatus infection [19].

C3

A central step of the complement cascade, independent of the activation pathway, is the production of C3 convertase, which activates C3 into C3b and C3a. C3b opsonizes A. fumigatus dormant conidia through RodA-rodlets and mycelia by binding to cell wall β-glucan and galactomannan [15]. This opsonization facilitates phagocytosis and mediate immune modulation [20, 21]. Moreover, C3b associated to factor Bb, functions as a C3 convertase, further amplifying C3 activation and deposition. C3b is also necessary to C5 convertase formation (C4bC2bC3b or C3bBbC3b), cleaving C5 with the resultant generation of C5a [22]. However, we showed that A. fumigatus can cleave C3 non-canonically, leading to the formation of inactivated C3b (iC3b) [23].

C5

C5a and C5b are generated via the cleavage of C5 by C5 convertase at a common checkpoint to all complement pathways. While C5b takes part in MAC, unlikely to harm A. fumigatus, C5a acts as an anaphylotoxin, able to recruit immune cells, particularly neutrophils.Thus, C5^−/− mice resulted in a significant higher mortality in a systemic intravenous mice model of aspergillosis [24]. Interestingly, C5 cleavage could also result from a non-canonical pathway triggered by A. fumigatus swollen conidia or hyphae [25], or by a fungal-associated metalloprotease [23]. C5a-receptor was found to be required to mount a protective immune response against systemic aspergillosis [24]. Of note, vilobelimab, an anti-C5a monoclonal antibody, has been shown to be beneficial in mechanically ventilated patients with COVID-19 patients [26]. If the use of vilobelimab becomes a standard care for COVID-19 patients, then it will be worth monitoring the incidence of COVID-19 associated pulmonary aspergillosis (CAPA) in those patients.

Mannose Binding Lectin (MBL)-Associated Serine Proteases (MASPs)

The lectin pathway of complement system is initiated by MBL-associated serine proteases (MASP-1, -2 and -3). MBL or other pattern recognition receptors of the collectin family (see below) bind to the pathogen surface, recruit these proteases, which further cleave key complement proteins, activating the complement cascade [27]. Interestingly, Rosbjerg et al. [28] observed direct binding of MASP-1 and MASP-3 to all morphotypes of A. fumigatus. MASP-1 and MASP-3 could recruit ficolin-3, which can bind alone to the fungus [29]. This reversed complex retained its ability to activate complement system, facilitating phagocytosis by neutrophils [28].

Other Complement Proteins

Opsonizing A. fumigatus conidia with human serum or bronchoalveolar lavage (BAL) from healthy donors indicated that several other complement proteins [C6, C7, C8, C9 (but not multimers, suggesting the absence of MAC formation), Factor-D and Factor-P] can directly interact with conidia. The biological importance of these interactions needs to be studied [15].

Complement Regulatory Proteins

A. fumigatus tries to evade complement attack by binding to several of the complement regulatory proteins. A recent study showed that A. fumigatus allergen, enolase, binds to Factor H, Factor-H-like protein-1 (FHL-1) and C4b binding protein (C4BP)[30]. Both Factor H and FHL-1 have Factor I cofactor function, and result in the inactivation of C3b, thereby decaying C3 convertase, whereas C4BP leads to C4b inactivation, an essential component of the classical/lectin pathway.

Collectins

Collectins are the Ca²⁺-dependant (C-type) family lectins that functions as soluble pattern recognition receptors (PRR) [31]. Their common structure contains a collagenous-like region linked to a carbohydrate recognition domain (CRD) (Fig. 1B from reference [22]). CRD enables binding to oligosaccharides, the main constituent of A. fumigatus cell wall. Some of these lectins such as MBL and ficolins are central molecules of the complement system lectin-pathway.

Mannose Binding Lectin (MBL)

MBL is produced by the liver and present in serum at concentrations of 0.2–2 µg/mL in healthy individuals [32]. Neth et al. [33] found MBL to bind A. fumigatus although the morphotype is not detailed. A recent study found no binding of MBL to resting conidia suggesting that the MBL-CRD binds to polysaccharides exposed after germination and therefore not involved in the direct interaction of inhaled conidia [15]. However, MBL provides a protective role in a corticosteroid immunosuppressed mice model of IPA [34] but not in an immunocompetent mice model of systemic aspergillosis [35]. Furthermore, some MBL genetic haplotypes are responsible for protein deficiency and were associated to both IPA and CPA in patient cohorts [36, 37]. Conversely, polymorphism responsible for increased MBL serum level was associated with bronchial asthma with allergic rhinitis and ABPA [38].

Ficolins

There are three types of ficolins in human: ficolin-1 (M-ficolin), -2 (L-ficolin) and -3 (H-ficolin), all functioning as recognition molecules in a Ca²⁺-dependent manner in the lectin complement pathway. Ficolin-KO mice showed significantly higher fungal loads in the lungs 24 h post-infection compare to wild-type (WT) mice that could be explained by decreased pro-inflammatory cytokines secretion, independent of complement activation [39].

Ficolin-1 is primarily produced by bone marrow derived cells and lung epithelial cells, and was found not to bind resting conidia of A. fumigatus [29] but hyphae through chitin and β-1,3-glucan in a Ca²⁺-dependent manner [40]. In vitro, ficolin-1 did not inhibit A. fumigatus growth but opsonization increased IL-8 secretion by A549 lung epithelial cells [40]. In human lung tissue samples, ficolin-1 was increased around aspergilloma and in BAL fluid from infected patients [40, 41].

Ficolin-2 is mainly produced by the liver. The homologue in mouse, ficolin-A, binds to conidia in a Ca²⁺-independent manner and enhances interaction with A549 cells [42], human monocytes derived macrophages (hMDM) and neutrophils [43]. This interaction was responsible for a decrease in pro-inflammatory cytokines production by hMDM and neutrophils [43]. Human ficolin-2, confirmed this binding pattern to conidia and opsonisation effect on cytokine production [29, 43]. The concentration of L-ficolin in BAL from IPA patients (n = 19) was significantly higher than in control [43]. Another recent study found no ficolin-2 by proteomic in BAL samples from infected patients (n = 10) compared to controls, suggesting ficolin-2 may have been consumed [41]. Of note, the control population was immunocompromised lung transplant recipients and non-immunocompromised patients with lung infection or cancer suspicion in the first and second studies, respectively [41, 43].

Ficolin-3 is synthesized in both liver and lungs with highest expression in the lungs. Binding of ficolin-3 to conidia is unclear: Ma et al. [29] did not identify any interaction by flow cytometry, while Bidula et al. [44] observed binding by immunofluorescence microscopy. Ficolin-3 interaction with hyphae still remains to be studied. Data regarding ficolin-3 levels in BAL of infected patients compared to controls are discrepant in the literature with the same control population bias as ficolin-2 studies [41, 44].

Surfactant Proteins

Surfactant proteins A and D (SP-A/SP-D) have a collectin structure and are constitutively expressed by type II alveolar cells and Clara cells in the airways [31]. Their primary function is to reduce surface tension at the air–liquid surface of the lung alveoli. They may also function as PRR opsonizing and participating in various pathogens clearance [31]. Both SP-A and SP-D were shown to bind to A. fumigatus conidia [45]. However, in a mice model of IPA only SP-D was found to have a protective role and increased survival rate if administered intranasally during the course of infection [46]. In addition, SP-A^−/− mice challenged with A. fumigatus conidia showed less mortality than WT mice while SP-D^−/− mice showed increased susceptibility to IPA [47].

SP-D binds melanin pigment on conidia and galactomannan as well as galactosaminogalactan on A. fumigatus hyphae cell wall [48]. SP-D opsonized conidia are phagocytosed more efficiently and stimulated the secretion of pro-inflammatory cytokines by hMDMs [48]. Furthermore, SP-D exerts direct fungistatic activity by restricting hyphal growth and induces hyphal surface modifications associated to increased susceptibility of the fungus to voriconazole [49].

On the allergic side, both SP-A and SP-D could bind A. fumigatus allergens and prevented their binding to specific IgE [50]. Furthermore, in sensitization/ABPA-like murine models, treatment with SP-A and SP-D significantly lowered blood eosinophilia, pulmonary infiltration, specific antibodies and Th2 cytokines [51, 52, 53]. Corroborating these results, SP-A^−/− and SP-D^−/− mice exhibited intrinsic hypereosinophilia and a Th2 bias of the immune response [47]. Overall, SP-A and SP-D appear to have opposite immunomodulatory functions and effects in IPA but a similar protective role against ABPA. More studies are needed to improve our understanding of these essential players of lung immunity.

Antimicrobial Peptides

Antimicrobial peptides (AMPs) are naturally produced key components of innate immunity. They control micro-organisms at early onset of infection and modulate inflammation. Their expression varies according to each tissue with some AMPs being ubiquitous and others, tissue- or cell-specific. About 147 human antimicrobial peptides are described in the antimicrobial peptide database [54]. Several of the antimicrobial peptides, known but not discussed below, remains unstudied regarding their potential effect on A. fumigatus [55].

Lactoferrin is a glyco-protein released by activated neutrophils and mucosal epithelial cells. It mediates iron depletion and was shown to inhibit conidial germination but not hyphal metabolism [56]. Proteolytic hydrolysis of lactoferrin results in various AMPs. A synthetic fragment of lactoferrin possessed direct activity and damaged A. fumigatus hyphae [57]. Intravenous injection of a lactoferrin peptide increased survival time in a mice model [58]. Histatin-5 and ubiquicidin are cationic peptides secreted by salivary glands and airway epithelial cells, respectively, also demonstrated anti-Aspergillus activity [56]. Lysozyme, the most prevalent AMP in nasal secretions displayed fungicidal effect on A. fumigatus at physiological concentrations [56, 59, 60].

Defensin family is an evolutionary conserved group of small cationic peptides in plants and animals. α-Defensins are stored in neutrophil granules, while β-defensins are specific of epithelial tissues. α-Defensins showed a fungicidal effect on swollen conidia and hyphae but not on resting conidia [58, 61]. α-Defensins 1 was associated to increased survival in a mice model [58]. β-Defensins also had an inhibiting effect on germination, but not on already formed hyphae [56]. Their production by epithelial cells in vitro was upregulated when exposed to A. fumigatus [58, 62]. Moreover, the expression of drosomycin-like defensins, displaying activity against broad spectrum of Aspergillus spp. was observed in several human tissues [63].

Cathelicidin are expressed in all mammals. In humans, this peptide is abundant in specific granules of mature neutrophils and activated by cleavage by neutrophils or microbial proteases into LL-37. LL-37 contributes to both intracellular and extracellular killing of micro-organisms. LL-37 was shown to bind all morphotypes of A. fumigatus [64]. However, its inhibitory effect is debated by two studies showing opposite results [64, 65]. At similar concentrations with similar protocols but different fungal strains, media and incubation period, LL-37 was shown to either promote growth [65] or have a fungistatic effect on A. fumigatus [64]. A recent study did not find any antifungal effect of LL-37 in two experimental designs targeting both hyphal metabolism or conidial germination [56].

Histones play essential roles in organization and regulation of chromatin. Although not traditionally seen as AMPs, they possess antimicrobial activity [66]. They are the major components of neutrophil extracellular traps (NETs), a key antifungal strategy of neutrophils against A. fumigatus hyphae. In a recent study, they inhibited hyphal metabolic activity with varying susceptibility among isolates [56].

Acute Phase Proteins

Acute phase proteins (APPs) are proteins whose plasma concentration either increase or decrease in response to inflammation and are thus called positive or negative APPs [67, 68]. C-reactive protein (CRP) is the first discovered and the most commonly known APP. Yet, many proteins discussed above, such as several proteins from the complement system (e.g. C3, C4, MBL), are APPs [67]. Some APPs directly interact with A. fumigatus and may opsonize and facilitate phagocytosis [22].

Pentraxins

Pentraxins are a superfamily of PRR known to cooperate with complement in the handling of pathogens. Short pentraxins include CRP and serum amyloid protein (SAP) while pentraxin-3 (PTX3) is a typical long pentraxin. Their action against A. fumigatus have been recently reviewed [69, 70] and is therefore only shortly summarized here. CRP (also known as pentraxin-1) was found to bind A. fumigatus hyphae [71] and promotes conidial phagocytosis by human neutrophils in vitro [72], but its role in fungal pathogenesis remains unknown. SAP (pentraxin-2) bound to conidia resulted in complement activation and enhanced phagocytosis by neutrophils [73]. Consistently, SAP^−/− mice showed increased susceptibility to IPA [73]. Using complement component depleted serum or serum from mice knocked out for specific complement components, it was shown that SAP recruitment on A. fumigatus conidia facilitates complement activation [73]. This data suggests that multiple humoral components can interact to mount antifungal host defense. Similarly, PTX3, secreted by phagocytes and non-immune cells at sites of inflammation has a non-redundant role in the immune response against A. fumigatus. Indeed, immunocompetent but PTX3^−/− mice show enhanced susceptibility to IPA due to defective recognition of conidia by neutrophils, macrophages and dendritic cells associated to a biased Th2 response [19]. This was further explained by conidial opsonization by PTX3 promoting conidial phagocytosis and killing by human and mouse neutrophils in vitro [74].

Galectins

Galectins are soluble mammalian lectins that bind specifically to β-galactoside sugars. One of them, galectin-3, released by both immune and stromal cells following cellular injury was found to interact directly with all morphotypes of A. fumigatus and released in response to the fungal presence in vivo [75]. Galectin-3-deficient mice were more susceptible to infection. The underlying mechanism was explained by galectin-3 being a non-redundant key player in enhancing neutrophil motility and their extravasation into the airway at the site of infection [75]. No other galectins were studied for their role against A. fumigatus although we recently showed that galectin-10 was significantly increased in the BAL samples of the patients infected with A. fumigatus compared to controls. Conversely, we found galectin-9 to be completely absent from this first group leading to suspect either its consumption by the immune response or its degradation by the fungus [41].

Circulating Antibodies

Antibodies/immunoglobulins are synthesized by the immune system to identify and neutralize foreign bodies. They mediate protection against fungi in general by enhancing complement activation and deposition, promoting phagocytosis and antibody-dependent cellular cytotoxicity [76]. They may also inhibit fungal growth, adherence, germination, and biofilm formation [76]. A dichotomy exists between regular antibodies (Abs) and natural antibodies (NAbs). While Abs are produced by mature B-lymphocytes upon antigenic exposure from infection or vaccine, NAbs are produced spontaneously by naïve B-lymphocytes [77]. NAbs, although less specific, provide protection against infection and eliminate cell debris and damaging molecular species [77]. Specific anti-Aspergillus Abs are detected in nearly everyone at low levels and increased from childhood to adulthood [78]. High levels are detected in patients with ABPA and CPA in which they are key diagnostic features [78]. In patients with IPA, Abs take a mean of 10.8 days to appear, making serology useless for early diagnosis [79]. Furthermore, sensitivity is higher in non-neutropenic patients (48%) than neutropenic patients (6%) [79]. B-cell deficient mice survive an intra-tracheal challenge with 2 × 10⁸ conidia [80]. Although not essential, several antigens (e.g., β-glucans [81], a 100-kDa cell wall glycoprotein [82], allergen Asp f3 [83]) were shown to elicit protective antibody response [84].

Regarding NAbs, a recent study suggests that complement C3 enhances NAbs pool contributing to innate defense against aspergillosis delaying mortality in a mouse model [24]. Furthermore, an essential axis was identified between innate B1a lymphocytes, A. fumigatus binding NAbs and lung neutrophils explaining the increased susceptibility to IPA of non-neutropenic patients with severe viral pneumonia [85]. Briefly, viral pneumonia depleted innate B cells and corticosteroids (often used to dampen hyperinflammation in these patients) contributed to the decrease of NAbs secretion. NAbs were found to be essential mediators of host defense through the enhancement of neutrophil-mediated phagocytosis in vivo [85]. Understanding Abs response may lead to therapeutic strategies such as the use of monoclonal Abs or intravenous immunoglobulins for passive therapy or vaccination, which is discussed in a dedicated part of this review (see section IV).

Humoral Immunity: A Tool to Identify at-Risk Population for Aspergillosis?

Several genetic polymorphisms in essential genes of humoral immunity players against A. fumigatus have been identified. The increased access and decreased cost of high throughput sequencing, whole genome sequencing and algorithm-based bioinformatic analysis are paving the way to personal medicine. Incorporating genetic markers into clinically valid processes to stratify the risk of fungal infection and to identify patients who might benefit the most from antifungal prophylaxis holds promise and groundbreaking innovation for patients at risk of aspergillosis [86]. So far described polymorphisms associated with increased risk for aspergillosis in humans are listed in Table 2. Polymorphisms were studied for MBL, PTX3, SAP and SP-A genes. PTX3 and SAP polymorphisms appear to be the most promising targets with well identified nucleotide substitutions linked to increased risk of IPA in immunocompromised patients. Prospective studies with reinforced prophylaxis protocol and/or monitoring could be undertaken. MBL polymorphisms may be of help to identify patients at risk for CPA and ABPA but more studies with increased number of well-defined patients are required to specify its potential.

Table 2.

Study identifying polymorphism statistically associated to aspergillosis in humans

Humoral factor	Haplotype/polymorphism or mutation	Findings	Reference
MBL	codon 52 polymorphisms	Increased frequency of R52C mutation in CPA patients (n = 11) compared with controls (n = 72)	[37]
		Increased frequency of C868T mutation in CPA patients (n = 15) compared with controls (n = 82)	[129]
	G1011A	Increased in frequency in asthma (n = 49) and ABPA (n = 11) compared with controls (n = 84)	[38]
		Increased in frequency in ABPA (n = 11) compared with controls (n = 20). Polymorphism associated with higher IgE levels and eosinophilia in ABPA patients	[130]
PTX3	Rs3816527 C > A	Increased risk of invasive mold infection (n = 26) in SOT recipients with A/A genotype compared to controls (n = 1075)	[131]
	Rs2305619 A > G	Increased risk of invasive mold infection (n = 26) in SOT recipients with A/A genotype compared to controls (n = 1075)
	h2/h2 homozygous haplotype (combinaison of minor allele of rs3816527 and rs2305619)	Associated with increased risk of invasive IPA in patients undergoing hematopoietic stem-cell transplantation (n = 268). Further confirmed patients with IPA (n = 107) compared with matched controls (n = 223)	[117]
		Increased risk of lung colonization (n = 14) and IPA (n = 13) compared with controls in lung transplant recipients	[132]
		Increased risk of invasive mold infection (n = 26) compared with controls (n = 133) in acute leukemia patients undergoing intensive chemotherapy	[133]
	Rs1840680 G > A	Increased risk of CPA (n = 36) compared with controls (n = 137) in chronic obstructive pulmonary disease patients with A/A genotype	[134]
SAP	Rs2808661 G > A	Increased risk of IPA in HSCT recipients if donor (but not recipient) displays homozygous A/A genotype (n = 372)	[73]
	Rs3753869 C > A	Increased risk of IPA in HSCT recipients if donor (but not recipient) displays homozygous A/A genotype (n = 372)
SP-A	SP-A1 (C1416T and T1492C) SP-A2 (C1649C and A1660G)	Increased frequency in ABPA (n = 22) patients compared with controls (n = 23). Mutations in SP-A2 showed stronger association with ABPA and were associated with clinical markers of severity	[135]

ABPA allergic broncho pulmonary aspergillosis, CPA chronic pulmonary aspergillosis, HSCT hematopoietic stem cell transplant, IPA invasive pulmonary aspergillosis, MBL mannose binding lectin, PTX3 pentraxin-3, SAP serum amyloid protein, SOT solid organ transplant, SP-A surfactant protein A

Humoral Immunity: A Promising Tool to Aid Aspergillosis Diagnosis?

Diagnosis of aspergillosis regardless of the clinical entity IPA, CPA or ABPA, remains challenging. Distinguishing disease from colonization is difficult when only a respiratory sample culture is positive for Aspergillus. Biomarkers (e.g., galactomannan antigen, nucleic acids) in serum have poor sensitivity especially in non-neutropenic patients to diagnose IPA and of no use in CPA and ABPA. Available serological tools display many drawbacks including disputed cut-off values and non-optimal choice of antigen used to detect specific antibodies [79]. On the other hand, humoral components described above vary during infection/colonization by Aspergillus. Known variations from human samples are presented in Table 1.

PTX3 is the most studied humoral component as a potential diagnostic marker. Serial monitoring of PTX3 plasma levels in pediatric leukemia patients with pulmonary fungal infections had significantly higher values compared with three other groups (healthy controls, those with bacterial/viral infections, and leukemic patients at the end of their induction phase of treatment) [87]. In addition, response to anti-fungal therapy correlated with normalization of PTX3 [87]. In a non-neutropenic cohort diagnosed with aspergillosis (n = 112, IPAs and CPAs), the plasma levels were ~ fivefold higher than in controls (n = 324, lung cancer or other microbial infections) with a sensitivity and specificity of 79.8 and 72.1%, respectively, and a cut-off value of 2300 pg/mL [88]. Such sensitivity is promising compared to galactomannan in non-neutropenic patients where a prospective study found a sensitivity and specificity of 37.8 and 87.1%, respectively [89]. In BAL fluid (BALF), a study analyzing 322 samples found significantly higher level of PTX3 in patients with invasive aspergillosis (n = 15) compared to colonized patients (n = 38) and healthy controls (n = 269) [90]. This was confirmed by a Czech and Chinese cohort proposing an optimum BAL PTX3 cut-offs for IPA of 2545 and 1900 pg/mL respectively when control groups are composed of patients with other microbial infections or lung cancer. These cut-offs were associated to a sensitivity/specificity of 100/98% and 86.3/82.5%, respectively [88, 91]. Consensual cut-offs remain to be determined but PTX3 may help differentiate between Aspergillus colonization and disease. However, PTX3 is not highly specific for Aspergillus infection; its level also rises during bacteriemia, dengue fever, leptospirosis, septic shock and has predictive value for the disease severity and increased mortality [92]. We recently undertook comparative proteomic analysis of the BALF from A. fumigatus infected or colonized (Asp +) hosts versus controls, to identify humoral components altered by the presence of this fungus [41]. Our results confirmed increased PTX3 in Asp + BALF. Interestingly, our study identified that specific humoral components were absent (i.e., C1q subunit A, MASP-2, SP-D, ficolin-2, complement factor H related proteins 2 and 5, complement-8) or exclusively present (i.e., ficolin-1) in Asp + BALF giving a potential panel of humoral components that could be combined to increase diagnosis specificity [41]. Although, not the subject of this review, interleukins have also been studied as potential diagnostic markers [93–95]. However, being chemical messengers, their levels vary with the disease status. The specificity of each humoral component as diagnostic marker remains to be studied in comparison with other fungal infections and more broadly with viral and bacterial infections. Furthermore, categorized analysis of IPA, CPA and colonized patient samples are needed. However, a combination of increased/decreased immune markers in blood and/or BAL may arise as a signature of aspergillosis.

Boosting Humoral Immunity as a Therapeutic Strategy Against Aspergillosis

Humoral components are essential to anti-Aspergillus immunity; some have been shown to increase survival if administered during invasive aspergillosis in animal models, awaiting confirmation in human clinical trials for treatment/prophylaxis. Indeed, therapeutic strategies could be completed by new approaches aiming to boost the host immune response by increasing availability of key humoral components. Furthermore, using natural proteins in their recombinant form could spare the use of antifungal compounds such as azole for prophylaxis and treatment in order to avoid side effects. The most promising and already available components are described in this review. Vaccine development to prevent aspergillosis, involving adaptative immunity, will not be discussed here but have been reviewed in-depth with the conclusion that much more work is still needed before it becomes a reality [96].

PTX3

Recombinant PTX3 (rPTX3) has been shown to effectively treat aspergillosis in both a mouse and a rat model of IPA [97, 98]. In mice, rPTX3 administered 5 days before Aspergillus challenge was preventing IPA, concomitantly survival rate was 100% with a dose ≥ 0.2 mg/kg. However, rPTX3 administration 5 days after the fungal challenge only delayed the death. The effect was similar if rPTX3 was administered intraperitoneally or intranasally independent of the dose [97]. In rats, rPTX3 (0.15 mg/kg) administered intraperitoneally had similar effect as voriconazole (20 mg/kg) on survival, reduction of fungal burden and improvement of respiratory functions [98]. A combination of both had an additional effect on all parameters. To the best of our knowledge rPTX3 has yet not been tested in any human disease and therefore require several steps before it can be considered for use in the treatment of IPA.

SP-D

SP-D was studied in mice models of both ABPA and IPA [46, 51]. Exogenous administration of human recombinant SP-D intranasally in a murine model of pulmonary hypersensitivity induced by A. fumigatus antigens lowered blood eosinophilia, pulmonary infiltration and specific Ab levels up to 16-days after the dose [51]. In a murine model of IPA, survival in the amphotericin B- and SP-D-treated mice was the same and significantly increased compared to controls [46]. There is a body of preclinical data supporting significant treatment effect in various pulmonary diseases [99]. Full length recombinant SP-D (AT-100) produced by Airway Therapeutics, received an orphan designation status for bronchopulmonary dysplasia in the US and European Union [99]. It has passed preclinical development steps and is currently being tested in a phase I clinical trial (ClinicalTrials.gov number NCT04662151) targeting bronchopulmonary dysplasia in preterm neonates. During aspergillosis, SP-D might be completely consumed and be a limiting factor of anti-Aspergillus immunity as identified in our proteomic study where SP-D was absent in Asp + patients BAL. Considering its availability for clinical trial, SP-D is a good candidate for investigation in aspergillosis alternative treatment strategy.

MBL

Despite showing some promising direct interaction with Aspergillus and the link observed between MBL polymorphisms and CPA/ABPA, MBL was only used once in a mice model of IPA [34]. Administered intranasally to mice on day 1 after infecting with A. fumigatus, a recombinant human MBL (rhMBL) significantly increased survival from 0 to 80% [34]. The use of rhMBL in clinical studies was undertook in phase Ib studies in the 2000s for the prevention of severe infections in individuals with multiple myeloma or undergoing lung transplant with deficient levels of MBL (ClinicalTrials.gov numbers, NCT00388999 and NCT00415311). A total of 46 and 24 patients were enrolled between 2006 and 2009 respectively before termination by the sponsor. The results have not been published and Enzon Pharmaceuticals discontinued further development of rhMBL[100]. Although MBL has not met the strict criteria established by this program, it may nevertheless be an important target in the management of aspergillosis.

SAP

Key findings about SAP and aspergillosis are recent and promising [73]. Administration of human SAP, both intraperitoneally (4 mg/kg) and intranasally (50 µg), in a mouse model of IPA at 2 h after inoculation significantly increased survival and reduced lung fungal burden [73]. A recombinant human SAP (PRM-151) has been developed and evaluated for the treatment of idiopathic pulmonary fibrosis [101, 102, 103]. Pharmacokinetic and pharmacodynamic data are available for intravenous administration [101]. PRM-151 appears to be well tolerated and a promising therapeutic option [102, 103] that could be tested as an adjuvant therapy or a prophylactic drug against aspergillosis.

Natural Abs and Monoclonal Abs

Natural Abs (Nabs) are major components of intravenous immunoglobulin (IVIg) therapy, delivered as pools from serum of healthy donors. It is used mainly as replacement therapy for patients with primary/secondary immune deficiency disorders or as immunomodulatory therapy in a wide range of autoimmune and inflammatory conditions, with a variable evidence base [104]. In liver transplant recipients, IVIg therapy was associated with a significant reduction of fungal infections [105], while it was not the case in hematopoietic stem cell transplant recipients [106]. As discussed above, they may be essential in certain clinical conditions only such as viral acute respiratory disorder syndrome where the cellular immunity producing NAbs such as innate B1a lymphocytes may be depleted [85]. However, clinical trials to assess efficacy of IVIg in these populations to specifically prevent IPA in hematology are lacking. One major drawback yet of IVIg is their impact on false positivity of galactomannan and β-D-glucan antigenemia, which could impair diagnosis and monitoring of patients with invasive aspergillosis [107]. Interestingly, experimental use of IVIg in allergy has been described and protective effects have been shown in a mouse model of ABPA both as prophylaxis and therapy [108].

Monoclonal Abs (MAbs) are produced by single B-lymphocyte clones and have high specificity in targeting its corresponding antigen(s). They have been produced for the treatment of cancers, auto-immune and allergic diseases as well as infectious diseases [109] including COVID-19 [110]. Various MAbs have been produced against Aspergillus spp. with scarce pre-clinical data regarding their efficacy as therapeutic agents with in vivo models [111]. In an intravenous mouse model of IPA, Mab R-5 raised against the cell surface protein enolase significantly reduced fungal burden in the kidneys [112]. Although MAbs could be a promising therapeutic strategy, much more data are needed before conceiving clinical trials.

Filling the Gaps and Understanding Challenges in Humoral Immunity Against Aspergillosis

Although humoral components are essential in linking A. fumigatus and immune cells, it has been far less studied. A simple Pubmed database search since 2000 comparing [Aspergillus AND humoral immunity] and [Aspergillus AND cellular immunity] finds a mean number of publications per year of 2.5 and 13.1 respectively. Consequently, many unknowns remain, and many promising leads and topics are awaiting researchers in the field of aspergillosis.

Lung-Specific Immunity Needs a Higher Place in Aspergillosis Research

Immune responses are influenced by various stressors depending on the localization in the body. Studies in animal models and human have revealed how immune cells are compartmentalized into tissue sites and exhibit tissue-specific subset composition. These studies have provided a new framework for approaching future challenges to human health and have readdressed partially solved topics such as aspergillosis [113]. A. fumigatus is an airborne pathogen and is primarily handled by lung-specific immunity. Lungs have the second largest surface area of human tissues after the digestive tract, with constant exposure to pathogens, allergens, particles as well as resident microorganisms requiring a complex immune balance to maintain homeostasis. Lung-specific innate immunity include a particular proportion of alveolar macrophages, lymphoid cells, dendritic cells and essential epithelial cells with multiple roles in the fight against respiratory pathogens [114]. Again, lung specific-humoral immunity literature is scarcer, but surfactant proteins have been identified as major and specific players. When we opsonized conidia [15] with BAL or serum followed by shotgun proteomic analysis of the conidial bound proteins, C3 was the major complement component binding to conidia [15]. Unsurprisingly SP-A and SP-D were only present in BAL opsonized conidial extract, whereas complement proteins C2, MBL, factor I, MASP-1 and MASP-2 were only found in serum opsonized conidial extract. This suggested that only the alternative pathway may be operational in the alveolar space. Furthermore, these differentially opsonized conidia elicited distinct cytokine responses from hMDMs [15]. This was the first study to shed light on lung-specific innate response against A. fumigatus. A common practice in performing in vitro cell culture assays is to supplement medium with serum. Therefore, most in vitro work regarding cellular immunity against A. fumigatus used serum as a surrogate for humoral soluble mediators. Relevance of these data may need to be revisited carefully.

Crosstalk Between Humoral Factors

Humoral immune components against A. fumigatus have been studied to the extent described in part I, yet one by one. These components may interact, cooperate and amplify microbial recognition or modulate the final immune response. For e.g., ficolin-2 was found to interact with PTX3 in a calcium dependent manner. Binding of ficolin-2 to A. fumigatus was enhanced by PTX3 and vice versa [29]. This promoted C4 deposition on the surface of A. fumigatus indicating an activation of complement lectin pathway [29]. C1q interaction with pentraxin family proteins is well document, yet the effect of such complexes remains to be studied with phagocytes in vitro [69]. Other interactions (e.g., MBL-PTX3) have been described to modulate immune response with other fungal pathogens but remains to be studied with A. fumigatus [31]. Conventional research tools may be inadequate to grasp a clear picture of the complexity of the humoral immune response and its multiple crosstalk possibilities. New technologies could help us quickly identify potential interactions to be confirmed secondarily.

Understanding Immunity at the Era of Omics

‘Omics’ are primarily aimed at the detection of genes (genomics), mRNA (transcriptomics), proteins (proteomics) and metabolites (metabolomics). ‘Omics’ strategies have already shown great interest in the study of host–pathogen interactions [115, 116]. Data remains scarce regarding the understanding of aspergillosis and humoral immunity.

• (i)Genomics. Genomic-wide association studies (GWAS) enable correlation of genomic variation with phenotype including susceptibility to infection and clinical outcome. To date, studies regarding genetic variation of humoral immune factor were genotyping studies targeting the gene of interest (Table 2)[73, 117]. A GWAS with well define substantial cohort of proven/probable IPA/CPA including controls could bring many new leads and be used for several purposes depending on the question asked (e.g., prognosis, response to treatment, etc.).
• (ii)Transcriptomics. A. fumigatus stimulates the host response to produce variable effector defense mediators such as acute phase proteins via a massive gene reprogramming in host immune cells. Genome-wide gene expression levels is nowadays possible and techniques like single cell RNA sequencing are routinely used although it requires acquisition and analysis of high-throughput data. To specifically analyze transcription of genes related to the immune response, several microarray strategies are available. Feys et al. [118] have recently analyzed expression of 755 genes linked to myeloid innate immunity in 166 patients with viral acute respiratory syndrome with or without associated aspergillosis. They identified that genes encoding proteins related to complement functions and PTX3-mediated phagocytosis of conidia were downregulated in patients with influenza associated pulmonary aspergillosis [118].
• (iii)Proteomics. Mass spectrometry has been applied to perform proteomic analyses and other -omics analyses (e.g., secretomics, lipidomics) of host–pathogen interactions. Several in silico or in vitro strategies allow to identify specific host–pathogen and/or immune system protein–protein interactions [116]. Machata et al. [119] have compared host and fungal proteins expressed during IPA in mice and humans identifying 11 common host factors significantly increased in both including the acute phase protein CRP. We recently compared BAL proteome of patients infected or colonized with A. fumigatus (Asp +) compared to controls with a specific focus on innate humoral factors [41]. Beyond the identification of differentially expressed proteins, we found several proteins being completely absent in the Asp + condition (i.e., SP-D, C1q, ficolin-2, C8, etc.) suggesting their importance in host–pathogen interaction by being either consumed, degraded by the fungus or initially deficient and a potential risk factor. These findings confirmed and deepened previous knowledge on certain humoral components and identified promising leads that resulted in major findings, especially on SP-D [49].
• (iv)Metabolomics. A new field, immunometabolism, has recently emerged to investigate metabolism dynamics in the context of immune processes. Indeed, the inflammatory response is an energy-intensive process that involves a switch from a resting to a highly active state [115]. To the best of our knowledge, no studies has yet been undertaken in the field of aspergillosis and humoral immunity. Combining ‘Omics’ technologies in multi-omic strategies provide orthogonal information to achieve a systems-level understanding of the host–pathogen interactions in basic, translational and clinical research. This will allow us to picture global immune phenotype during aspergillosis and identify major crosstalk between humoral mediators. Feys et al.’s study, combining transcriptomic and protein data in human samples infected with A. fumigatus, is the first of a kind and paves the way to future avant-garde work. While the costs of such studies remain high and analysis of high throughput data challenging, these studies may serve multiple purposes depending on the question asked, the investigation’s angle and could lead to fructuous research collaborations with a common dataset.

Identifying the Impact of Microbiota and Mycobiota on Humoral Immunity

Recent studies identified associations between microbiota and mycobiota or specific bacterial and fungal taxa with several lung diseases including infectious, systemic or allergic diseases [120]. However, correlation does not imply causality and to decipher whether the microbiota and/or mycobiota is the cause, or the consequence of a condition requires more studies to assess microbial community function in parallel with host response and clinical disease features. Furthermore, microbiota and mycobiota are rarely studied in parallel although bacterial and fungal organisms are known to interact and a wider picture is needed to understand these complexe dynamics. Many angles may be subject for future investigations, the first being the impact of gut microbiota and mycobiota on humoral response to Aspergillus. The impact of the gut microbiota on the development and function of humoral immunity has been established although findings need to be deepened, opening the possibility of microbiome engineering for optimal immune response [121]. For e.g., the gut microbiome and its metabolites, such as short-fatty acids can modulate pulmonary host defense and decrease allergic pulmonary inflammation, thereby could play a role in Aspergillus asthma and ABPA [122]. Furthermore, the gut microbiota plays a clinically relevant role in drug disposition including microbial interaction with the host immune system that can have broad effects on immunomodulatory drugs [123].

The second angle is to study the impact of lung microbiota and mycobiota on humoral response against Aspergillus. In contrast to the oral and gut microbiome, studies on the pulmonary microbiome faces challenges, since the lower respiratory tract is not easily accessible and contains a low microbial biomass [124]. A recent study by Hérivaux and co-workers provides evidence for marked alterations in lung microbiota composition and diversity in patients with IPA with a significant decrease in Prevotella and Veillonella taxa [125]. They found that lung dysbiosis worsens the prognosis of patients with IPA, yet specific underlying immune mechanism remains unknown [125]. Some cytokine levels (IL-8 and IL-27) were negatively associated with bacterial diversity in the entire cohort of IPA and controls while the abundance of Finegoldia was positively associated with level of IL-2 specifically in cases of IPA. The third angle would be to study the impact of Aspergillus presence among the microbiota and mycobiota (i.e., colonization) on humoral factors. Overall, understanding how the microbiome interact with the immune system is essential to uncover the molecular and cellular mechanisms underlying pathogenesis of IPA, CPA and ABPA.

Conclusion

We reviewed the current knowledge regarding humoral immunity against A. fumigatus. Although not addressed here, A. fumigatus displays mechanisms to counteract and evade humoral immunity that could be interesting therapeutic target and reviewed elsewhere [126]. Cytokines were deliberately not discussed as they are not directly involved in immune recognition and interaction with A. fumigatus. Many challenges remain to get a full picture of the complexity of humoral immune response against A. fumigatus. Yet, the potential impact of such research is significant as it could help anticipate, diagnose and treat the very different entities of aspergillosis spectrum.