3. The Complement System: Effector and Immune HubTumors are more than only an aggregation of transformed cells. Instead, they represent a complex mass composed of multiple local and recruited cell types and soluble compounds [8]. A significant part of these elements belongs to the innate or adaptive immunity system [8,9]. Their activation is triggered by abnormalities of the malignant cells, such as altered cell morphology and generation of tumor-specific antigens that distinguish cancer cells from their non-transformed counterparts. The resulting inflammatory processes form a specific immunological tumor microenvironment (TME) that actively participates in multiple cancer hallmarks, such as cell proliferation and survival, tumor angiogenesis, metastasis, immune evasion, and resistance against therapies.One important contributor to the TME is the complement system. Complement is an evolutionarily ancient, multifunctional system of innate immunity. It consists of close to sixty soluble and membrane-bound proteins including activating proteins, regulators, and receptor molecules. All of them act together to form a tightly regulated cascade of cleavage and assembly reactions, thereby generating effector molecules that fulfill various functions [10,11,12]. The main production site of complement factors is the liver, but the local synthesis of various factors is found in most organs, including the pancreas [12,13,14,15,16]. Figure 1 gives an overview of the complement cascade with its three activation pathways.The classical pathway (CP) is initiated by the binding of complement factor C1q to antibodies/immune complexes on the surface of foreign target structures [17]; therefore, CP serves as an effector arm of the adaptive immune system and is referred to as “antibody-dependent”. The lectin pathway (LP) shares several steps and involved molecules with the CP, but is triggered by sugar molecules commonly found on the surface of microbial cells. The main starter molecule of LP is mannan-binding lectin (MBL), but also ficolins, are able to initiate the same processes [18]. Both CP and LP result in cleavage of the factors C4 and C2 that form the C3 convertase C4b, 2a.The alternative pathway (AP), often called ‘tickover’, is activated slowly and continuously in plasma by hydrolysis of factor C3. In the presence of foreign surfaces, a complex of C3b and the fragment Bb, derived from cleavage of factor B, is formed [19]. This C3b, Bb complex is stabilized by properdin (P) to generate the C3 convertase C3b, Bb, P of the alternative pathway [17]. In addition to spontaneous activation, the alternative pathway can propagate the complement cascade by entering the amplification loop [20].Both C3 convertases efficiently perform the central step of the complement cascade, the cleavage of complement factor C3 into the fragments C3a and C3b. C3b is an efficient opsonin that marks foreign structures for phagocytosis, but also participates in the further proceeding of the complement cascade. It attaches to one of the two C3 convertases to form the respective C5 convertases. These enzymes catalyze the second central step of the complement system, the cleavage of C5 into C5a and C5b. C5a, together with C3a, represents an important pro-inflammatory and chemotactic mediator and affects multiple processes in the body [21]. The fragment C5b initiates the terminal complement pathway that results in formation of the lytic C5b-9 complex, also termed membrane attack complex (MAC) or terminal complement complex (TCC). This complex forms a pore in the attacked surface, thus destabilizing the intracellular environment of pathogens or altered cells and resulting in their inevitable death.Apart from classical, lectin and alternative pathways, several bypass mechanisms also result in complement activation, such as those mediated by thrombin [22], cathepsin [23] or trypsin [24] (see also Figure 1). Since trypsin represents a central enzyme that is produced and released by pancreatic exocrine glands, the cleavage of C3 and C5 by this enzyme is of particular relevance in pancreas-associated conditions [25].The role of the complement system is often focused on destruction of pathogens and includes their direct lysis via the C5b-9 complex and their opsonic tagging for phagocytosis; the latter is achieved via deposition of C3b and C3b-derived fragments on the pathogen surface, facilitating their recognition and engulfment by phagocytes [26]. Furthermore, complement promotes pro-inflammatory responses to control infection, mainly via the anaphylatoxins C3a and C5a [27,28]. However, the spectrum of complement-associated functions is much broader than just antimicrobial attacks. Removal of apoptotic/dead cells as well as clearance of immune complexes from circulation to maintain and re-establish homeostasis are further core tasks of complement. Complement is also a key orchestrator of both innate and adaptive (B-cell and T-cell based) immunity and regulates their activity [27,28]. The multifaceted roles of complement in cancer are explicitly reviewed below and specified for pancreatic cancer.By nature, the complement system attacks everything that is not actively protected by a plethora of regulator proteins that emit the message ‘do not attack’ [21]. To avoid collateral damage of self-tissues, very tight regulation and protection of host cells are of utmost importance. This is reflected by the considerable number of soluble or membrane-bound regulatory proteins that exceed the number of complement factors. Only a few examples relevant for PDAC are mentioned here; for a detailed overview, see [29]. Although complement regulators work on multiple levels of the cascade, most of them affect the generation or function of either C3 or C5 convertases. Common membrane-bound regulators protecting the vast majority of body cells from complement attack are CD46, CD55 and CD59. Whereas CD46 and CD55 promote C3 degradation or decay of C3 convertases, CD59 inhibits the assembly of the C5b-9 complex. The soluble regulators C1 inhibitor (C1-INH) and C4 binding protein (C4bp) both down-modulate the proceeding of classical and lectin pathway, C1-INH by blocking the relevant serine proteases and C4bp by promoting the decay of C3 convertase of these pathways. The soluble regulator factor H (fH) targets the alternative pathway and accelerates the decay of the respective C3 convertase [29]. 4. Complement as Part of the TMECancer is much more than a mass of transformed cells that grow autonomously. The growing cell mass is sustained and protected by a tumor-associated stroma, composed of multiple infiltrated and local cell types as well as various proteins. This complex and diverse environment is referred to as tumor microenvironment (TME). The pivotal role of TME in cancer progression became obvious in the last years and is thoroughly reviewed elsewhere (e.g., [30,31]).

Besides extracellular matrix proteins (collagens, proteoglycans etc.) and signaling molecules such as cytokines and chemokines, the complement system is a main player of the TME protein components. Its role in modulation of tumor development and in orchestration of TME is reviewed and experimentally confirmed for PDAC in the following chapters (see below). The most relevant cell types for the role of complement in the TME of PDAC are (beside the cancer cells themselves) cells related to the innate immune system such as myeloid-associated cells such as tumor-associated macrophages (TAMs) and tumor-associated neutrophils (TANs), as well as cancer-associated fibroblasts (CAFs) and pancreatic stellate cells (PSCs) which are crucially involved in stroma formation.

TAMs represent a highly abundant immune cell type within tumors that are able to fulfil a broad repertoire of functions via diverse phenotypes [32]. TAMs can be categorized into M1 and M2 subsets that differ in their contribution to tumor development. The M1-like pro-inflammatory phenotype mediates a potent tumor-suppressive immune response, whereas M2-like TAMs contribute to tumor-supporting procedures such as angiogenesis, immunosuppression and metastasis. The TAMs that reside within the tumor generally show characteristics of the M2 phenotype [33].Tumor-associated neutrophils (TANs) also show N1 or N2 phenotypes. After being recruited, neutrophils can be polarized by various cytokines or chemokines to one of the phenotypes and subsequently play an either tumor-suppressing or tumor-promoting role [34]. The antitumor N1 phenotype enhances cytotoxicity against tumor cells and attenuates the local immunosuppression, whereas the protumor N2-TANs participate, e.g., in metastasis development.Pancreatic stellate cells (PSCs) are pluripotent mesenchymal cells that exist typically in a quiescent state. In a normal pancreas, they reside around acinar cells and contain large amounts of vitamin A-containing lipid droplets [35]. After being activated by a variety of triggers, PSCs subsequently undergo a morphological change into myofibroblast-like cells and change their functionality spectrum. Nowadays, it becomes more and more clear that they represent critical players in pancreatic pathophysiology, pancreatitis and pancreatic cancer [35,36]. Activated PSCs harbor immunological functions, can recognize pathogen-associated molecular patterns and engulf pathogens [37]. Furthermore, activated PSCs highly upregulate the expression of stroma proteins and thus crucially contribute to the excessive fibrosis of aggressive PDAC. The synthesized desmoplastic stroma promotes the formation of a microenvironment that favors malignant transformation and facilitates the capacity of cancer cells to survive and invade [38].Cancer-associated fibroblasts (CAFs) are derivatives from PSCs, but also from other cell types in the TME. They represent an important cellular component of the TME and are also main producers of the various stroma proteins [39]. CAFs are heterogeneous and can be divided in different subpopulations: myofibroblast CAFs (myCAF), inflammatory CAFs (iCAF), antigen-presenting CAFs (apCAF) and complement-secreting CAFs (csCAF) [39,40]. CAFs are involved in immune regulation of the TME by inhibiting CD8+ cells and producing immune-modulatory cytokines/chemokines. Via secretion of these molecules, CAFs are supposed to affect cancer cell proliferation and functionality of other immune cells (TAMs, TANs, PSCs) in the TME [39]. 6. Fungal Dysbiosis and PDAC DevelopmentIn recent years, more and more attention was pointed at the role of the microbial flora in the initiation and progression of pancreatic cancer. Interestingly, the role of the fungal flora, so called mycobiota, arose as a key factor in tumorigenesis, and a distinct mycobiome signature was demonstrated to characterize several types of cancers [7].Not only with regard to tumors of the hollow organs of the gastrointestinal tract, like colorectal [89] and esophageal cancer [90], specific fungal signatures have been described. Additionally, within the human pancreas, an organ formerly thought to be sterile, there is increasing evidence of the presence of a specific mycobioma [91,92]. Initially described in 2019 by Aykut et al. and recently confirmed by other independent studies, 18S rRNA sequencing and immunohistochemistry analysis revealed a change in the human intrapancreatic mycobioma not only in the presence of PDAC but also in the presence of precursor lesions like PanIN [7,68,93]. However, all these techniques demonstrate only the presence of fungal components like nucleic acid or membrane compounds like beta-glucan, and up to now, any attempt to cultivate living fungi from tumor specimens has been unsuccessful [7].Further metagenomic characterization of the mycobiome showed absolute abundance of Malassezia spp. within human and mouse PDAC [7,68] and also of Alternaria spp. [7]. Malassezia is a known skin commensal with capability of gut colonization in humans [94], and its presence within the pancreatic gland is thought to occur through a direct migration from the duodenum via the main pancreatic duct [68]. Of note, Malassezia spp. encodes some secreted enzymes—similar to those of Candida spp.—that have been shown to act carcinogenically [95,96].From a clinical point of view, fungal dysbiosis was demonstrated to be related to the tumor burden. The ablation of the mycobiota resulted in prolonged survival and reduction of tumor progression in an orthotopic mouse model for PDAC. Interestingly, only the species Malassezia showed the capability of promoting tumor growth in PDAC, while other fungi like Aspergillus spp. or Candida did not [68].Based on these findings, different pathogenetic models have been proposed to understand the pathways involved in this fungal-driven tumorigenesis [92]. Principally, the link between the immune milieu within the tumor and its response to fungal components are thought to represent the crucial aspect of this process. Alam et al. showed that increased immune-suppressive cell populations like Th2 and innate lymphoid cells 2 (ILC2)—which characterize the tumor microenvironment (TME) of PDAC and can lead to its progression—are recruited by IL-33 secreted by PDAC cells. Of note, the secretion of IL-33 was demonstrated to be dependent on the composition of the intratumoral mycobioma [93]. So far, the link between the fungal presence and the extracellular release of IL-33 has not yet been defined.Aykut et al. observed a crucial role of the complement system in relation to the tumorigenic processes within the pancreas. More specifically, the model proposed by this study group implies the recognition of Malassezia ssp. by MBL and the activation of the lectin pathway of the complement with consequent production of C3. These findings are corroborated by the positive correlation between MBL and C3 with poor survival in PDAC patients [68].

These two independent studies showed interesting links between the intrapancreatic mycobiota and the interdependent pro-tumorigenic immunity shape. However, many questions remain open. It is still unclear which receptors and specific mechanisms can shape the TME in such a pro-tumorigenic way. Furthermore, no specific tissue-related analysis concerning the expression of complement factor within pancreatic tissue was performed. It is still unclear whether the tumor cells themselves can produce complement factors or activate their cascade.

Curiously, the role of the reactive species of the oxygen (ROS) in response to intrapancreatic fungal colonization remains quite unexplored. Still, these compounds could reveal new interactions between the different tumorigenic pathways. On the one hand, ROS can activate extracellular release of IL-33 [97,98]. On the other hand, lectin complement pathway activation can occur in a context of oxidative stress [99]. Moreover, IL-33 and complement anaphylatoxins are known to interact on mast cells [100] synergistically. Whether this synergism happens on other immune cells, or even on PDAC cells themselves, is an intriguing question.

Strong evidence suggests a link between fungal dysbiosis and immunologic alterations resulting in a pro-tumorigenic milieu within the pancreas. With the complement system being crucially involved in these mechanisms, analyzing the expression of the complement factors within pancreatic tissue is of primary importance to further understand its role in tumor development.

7. Pilot Study: Complement and Malassezia in PDACAs illustrated above, new insights underline the relevance of complement in the context of PDAC development. Furthermore, the model of fungus-driven complement activation (described in Section 6) corroborates the role of complement as a main orchestrator in the TME. However, additional aspects need to be further clarified. In the study of Aykut et al. [68], the focus is set on the role of the lectin pathway, but no consideration has been made in regard to the classical and alternative pathway of complement. Moreover, the relation between the colonization by Malassezia and the effective level of complement expression and/or secretion remain still unknown.

Intending to further characterize this interplay between the complement system and fungal dysbiosis, we set up a pilot study based on the retrospective analysis of formalin-fixed paraffin-embedded (FFPE) specimens of pancreatic lesions. Following approval from the ethics committee of the Medical University of Innsbruck (study number 1188/2021), 19 pancreatic lesions, resected between 2017 and 2019 at the Department of Visceral, Transplant and Thoracic Surgery of the Medical University of Innsbruck, were included in the study.

The intent of the pilot study consisted in analyzing the specimens (1) for the presence of DNA fragments of Malassezia spp., which had been recently described to be involved in PDAC progression [68], and (2) for the expression of selected relevant complement proteins. More precisely, our examination focused on C1q as starter molecule of the classical pathway, on factor B as specific for the alternative pathway, and on MBL as a pattern recognition molecule of the lectin pathway. In addition, we analyzed the presence of C3, the central complement factor of all three pathways, acting as an opsonin and as starter molecule of the terminal pathway. From the histopathological point of view, we focused on entities arising from the exocrine component of the pancreas, including PDAC, as well as premalignant and benign lesions. Stroma, epithelium and inflammatory cells of the lesions were evaluated separately. For all 19 cases, lesional as well as perilesional specimens were available.Following DNA extraction from each FFPE specimen, specific multiplex PCR using both primers for the ITS region of Malassezia spp. and species-specific primers for M. globosa and M. restricta were performed. The expression analysis of the different complement proteins was done by immunohistochemistry (see Appendix A for a detailed Material and Methods description).Among the 19 examined specimens, 10 were PDAC, and 9 were benign or premalignant pathologies, including 5 intraductal papillary mucinous neoplasms (IPMN), 2 cases of chronic pancreatitis, 1 pseudotumor in autoimmune pancreatitis and 1 serous cystic neoplasm (5.3%). The patients’ demographic data are reported in Table 1.

With regard to the fungal DNA, out of 19 specimens, positive Malassezia DNA isolation occurred in 14 cases. Malassezia DNA was present within the lesion in six cases, and in eight cases in the perilesional area. No specimen showed positivity for Malassezia DNA, both within the lesion and in the perilesional area simultaneously. Considering the distribution in relation to the histopathologic diagnoses, all 10 PDAC were positive for Malassezia, either in the lesion or in the perilesional area, while only four non-malignant lesions showed positivity for Malassezia (p = 0.006). Interestingly, three of these four non-malignant specimens containing fungal DNA were lesions known to harbor a higher risk of malignant transformation, namely one pseudotumor in autoimmune pancreatitis, one IPMN, and one chronic pancreatitis. No correlation was observed between the presence of Malassezia DNA and known risk factors for PDAC (age: p = 0.354, BMI: p = 0.853, diabetes: p = 0.144, tobacco: p = 0.865, and alcohol consumption p = 0.865).

The expression of each complement protein was analyzed separately in three different components of the pancreatic lesions, namely the tumor stroma, the tumor epithelium and the cellular infiltrates. All complement proteins analyzed (C1q, fB, MBL and C3c) revealed correlations either with the dignity of the lesion or with the presence of Malassezia DNA or with both.

Tumor epithelial cells displayed significantly higher C1q, C3c and MBL expression than benign/premalignant lesions (p = 0.027, p = 0.034, and p = 0.027, respectively; Figure 3 and Figure 4).

In a similar pattern, a significantly higher expression of C1q, fB and MBL was detected in epithelial cells of lesions positive for fungal DNA compared to those not harboring fungal DNA (p = 0.003, p = 0.048 and p = 0.003, respectively), with C3c reaching almost statistical significance (p = 0.053).

With regard to the tumor stroma and the inflammatory component of the lesions, expression of C1q and MBL was significantly higher in Malassezia DNA-positive lesions compared to DNA-negative lesions (p = 0.041, p = 0.041, and p = 0.046, p = 0.046, respectively; Figure 5 and Figure 6).The low number of analyzed specimens in this pilot study must be considered when interpreting these data. Still, the important correlation between fungal DNA and the observed malignant pathologies confirms earlier findings regarding a putative tumorigenic role of Malassezia spp. It is noteworthy to highlight that not only all malignant specimens but also three premalignant lesions were found positive for fungal DNA. This suggests that the Malassezia spp. plays a pivotal role in tumor progression, an observation already made by the group of Aykut, who described in a mouse model the presence of Malassezia spp. as crucial for tumor development [68].The results presented here suggest an intriguing interaction between the mycobioma and the complement system. In this regard, Aykut et al. already observed in PDAC activation of the complement system via the lectin pathway in the presence of Malassezia. As reported in Section 5.5, the deletion of MBL resulted in significantly lower tumor mass in a PDAC mouse model [68]. The present pilot study corroborates this observation by showing a correlation between MBL and Malassezia DNA in PDAC lesions. However, it also points at a possible involvement of other complement pathways.Since not only MBL, but also C1q and C3c expression in epithelial cells correlated with Malassezia positive malignant lesions, activation of more than one complement pathway as pro-tumorigenic cause of fungal dysbiosis has to be taken into consideration. In fact, these findings are in accordance with previous reports of tumor cells [44] or recruited immune cells like TMAs [4] or CAFs [48], showing higher expression and production of different complement factors including not only MBL but also C1q and C3 (see Section 5.1). Similarly, these three complement factors have also been observed to be activated in tumoral microenvironment ([40,53], see Section 5.2) In addition, the higher expression of complement factors C1q and C3 found in our prelimianry study is in line with previous findings addressing the complement as promising a sensitive and specific biomarker for PDAC (see Section 5.9). More precisely, the elevated presence of C1q and C3 in the serum of PDAC patients [3,4,75] correlated with the higher expression of the same factors within the pancreatic tissue. So far, there is still no evidence for the clinical use of serum MBL as possible marker for PDAC. However, in the light of our preliminary findings a characterization of MBL levels as well as of other complement factors in patients’ serum could be an appealing project aiming at testing the specificity and sensitivity of these molecules in the PDAC setting but also at stratifying patients according to their tumor prognosis.In contrast, despite the observed correlation with the presence of fungal DNA in epithelial cells, there was no difference in fB expression between malignant and benign or premalignant samples. This is in conflict with the observation of fB as a highly-specific serum marker for PDAC [47]. Not having access to patients’ blood samples in this retrospective study, we cannot exclude significant differences in the serum of these patients. Interestingly, our analysis revealed an increased expression of fB in the stroma of Malassezia ssp. positive lesions, although not reaching statistical significance. Even though not directly confirming the observations of Shimazaki et al., who described a high stromal fB expression in PDAC stroma with unfavorable clinical outcomes [59], this finding could point to an interplay between fB and Malassezia ssp. as a promoting factor for further tumor development.The separate analysis of three distinct components of the specimen (epithelial cells, stroma, and inflammatory infiltrates) showed not only a correlation between complement factor expression and fungal DNA in the tumor epithelial cells, but also in the cells of the stromal and of the inflammatory compartment of the analyzed lesions. This confirms recent findings that complement activation does not only occur extracellularly when recognizing membrane abnormalities (see Section 5.2). In fact, in human T-cells, intracellular stores of complement proteins have been shown to be responsible to guide T-cell differentiation towards different subsets [101]. This suggests that the activation of the complement cascade could happen intracellularly in the epithelial pancreatic compartment, like in colorectal cancer, where intracellular complement participates actively in the tumorigenic process [102]. So far, mechanisms underlying its intracellular activation and regulation and its functional outcomes, in particular concerning the pancreas, are largely unexplored [36,37,38,39,40,41,42,43,44,45]. However, this aspect opens a new scenario regarding innate immune surveillance in relation to PDAC development.