The pathways and the mechanisms by which Cryptococcus enters the brain

ABSTRACT

Generally, Cryptococcus initially infects the respiratory tract, but can spread, eventually crossing the blood-brain barrier (BBB) and causing meningitis or meningoencephalitis. Specifically, Cryptococcus invades the vascular endothelial cells of the BBB, from which it enters the brain. The main mechanisms through which Cryptococcus crosses the BBB are transcellular traversal, the paracellular pathway, and via Trojan horse. In this paper, the mechanisms by which Cryptococcus crosses the BBB were explained in detail. In addition to pathways of entry to the brain, this paper presents a discussion on some rare cryptococcal infections and provides some insights for future research directions.

KEYWORDS:

• Cryptococcus
• blood-brain barrier
• transcellular traversal
• paracellular pathway
• trojan horse mechanism

1. Introduction

Cryptococcus is a genus of basidiomycetous yeast, a pathogenic fungus that can cause cryptococcal meningitis and pneumonia. Cryptococcus neoformans (Cn) and Cryptococcus gattii (Cg) are the main pathogenic types (Francisco et al. 2021), both of which can cause nervous system disease. Cg commonly infects healthy hosts whereas Cn typically affects HIV-infected or other immunocompromised hosts (Sorrell et al. 2016; Yu et al. 2021). Some people infected with Cn may present with nervous system disease (Zaragoza 2019), especially immunocompromised individuals, who present with poorly controlled diabetes, malignant tumours, acute granulocytopenia, or severe burns; have a history of long-term antibiotic use; are receiving chemotherapy or glucocorticoids; or have undergone organ transplantation or have AIDS and therefore taking immunosuppressants long term (Thompson and Patterson 2012; Craig 2019). Globally, there are almost 250,000 cases of cryptococcal meningitis per year, and 181,000 deaths are attributed to cryptococcal meningitis annually, the mortality rate is 100% if the infection remains untreated (Iyer et al. 2021). Cryptococcal meningitis which is more common in adults with HIV (Ratemo and Denning 2023), accounts for 15% of AIDS-related deaths (Li et al. 2021).

Inhalation of cryptococcal basidiospores or yeast in the environment can result in pulmonary infection and meningitis if the infections spread to the central nervous system (CNS) (Leongson et al. 2013). Cryptococcal meningitis is most commonly caused by haematogenous dissemination, in which Cryptococcus spreads through the bloodstream from the lungs and crosses the blood-brain barrier (BBB), causing neurological disease. It also infects other parts of the body, resulting in cryptococcal laryngitis (Nadrous et al. 2004); however, occurrences of infection at alternative sites are uncommon. The CNS includes three important barriers: BBB, the blood-cerebrospinal fluid (CSF) barrier (BCSFB), and the CSF-brain barrier. These barriers share several functional similarities that protect the brain from toxins, maintain brain cell metabolism, and prevent pathogen entry (Dunn and Isaacs 2021), and among these barriers, the BBB has been the most extensively studied. Cryptococci crossing of the BBB is key to cryptococcal meningitis; therefore, it is important to study the interaction between cryptococci and BBB. Moreover, the sinuses are connected to the cranial cavity, making it easy for Cryptococcus to enter the brain. This review focuses on the different pathways of Cryptococcus entry into the brain and the mechanism by which Cryptococcus crosses the BBB.

2. The pathways by which Cryptococcus enters the brain

Cryptococci are inhaled and typically infect the lungs or CNS, after entering the respiratory tract, but can cause disease after accumulating in any part of the body (Figure 1). This means that cryptococci enter the CNS from different sites via multiple pathways possibly. For example, in the rare case of cryptococcal laryngitis, cryptococci can accumulate in abundance in the throat, where they can enter the blood circulatory system and ultimately cross the BBB into the brain. The mechanism is discussed in detail in the following part.

Figure 1. The pathways by which Cryptococcus result in infection. Cryptococcus were found in eucalyptus trees and pigeon droppings. Inhalation of cryptococcal basidiomycetes or yeast cells from the environment can cause pulmonary infection and meningitis. The intracranial infections originate from direct spread through bone defects and nerves, or indirect haematogenous diffusion, ultimately resulting in the pathogen crossing the blood-brain barrier (BBB) (By Figdraw).

Moreover, the infectious particles in human cryptococcosis are yeast and spores (Velagapudi et al. 2009). Much attention has been paid to yeast cells, while how spores cause infections and disseminate to the brain remain poorly understood. Some studies have shown that spore-infected mice exhibit a higher brain fungal burden compared to yeast-infected mice (Walsh et al. 2019; Frerichs et al. 2021). It was found that such differences in outcome were the results of events occurring in the lungs before spreading to the bloodstream (Walsh et al. 2019). The current study suggests no difference between yeast cells and spores in the mechanism by which they cross the BBB from the blood into the brain. In addition, some studies have demonstrated that spores germinated into yeast inside alveolar macrophages and disseminate into the brain (Giles et al. 2009; Frerichs et al. 2021). We speculate that spores can also cause sinusitis infections, then germinate into yeast inside macrophages and enter the brain through direct spread.

In addition to the BBB, the barriers of the CNS include the BCSFB and the CSF-brain barrier. Is there a mechanism by which cryptococci enter the CSF after crossing brain capillaries? The choroid plexus is responsible for the production of CSF, suggesting that there is a mechanism by which cryptococci cross the choroid plexus of the brain and enter the CSF. Despite the discovery of cryptococcal choroid plexitis cases (Dubbioso et al. 2013), which suggests a possible mechanism for the entry of cryptococci into the cerebrospinal fluid through the choroid plexus, no yeast cells have been observed in the choroid plexus of mouse models (Chang et al. 2004), and no evidence of disruption to the epithelial layer or the basal lamina in choroid plexus sections has been observed (Charlier et al. 2005).

In addition, other pathways of entry into the brain have rarely been reported, for example, cryptococcal sphenoid sinusitis with meningitis, skull base osteomyelitis (Prendiville et al. 2016), and cryptococcal endophthalmitis (Crump et al. 1992; Amphornphruet et al. 2018). The sinuses are connected to the nasal cavity through an ostium and are separated from the brain by only a thin bony structure (Goldman-Yassen et al. 2021), which is an important site of intracranial complications of sinusitis. Inflammation of the paranasal sinuses may lead to complications, such as meningitis or endophthalmitis (Younis et al. 2002). Various fungi have been reported to be pathogenic and cause acute invasive fungal rhinosinusitis (AIFR), these fungi include Mucor/Rhizopus, Aspergillus, Alternaria, Scedosporium, Candida, and Fusarium (Luo et al. 2023). Accelerated reproduction of fungi may cause sinus bone destruction accompanied by intraorbital, pterygopalatine fossa, and even intracranial dissemination of the pathogen (Aribandi et al. 2007). However, AIFR is a rare disease and occurs mostly in immunosuppressed patients, and the incidence of intracranial infection is low (Thompson and Patterson 2012; Craig 2019). Intracranial infection usually originates from indirect haematogenous diffusion (De Vita et al. 2021), with the pathogen ultimately crossing the BBB (Figure 1). Direct pathogen spreading is rare (Carr 2016), and usually involves bone defects, the space between sinuses (Figure 1), the intracranial space (Jones et al. 2002; Younis et al. 2002), or the nerves in the brain (Sekaran et al. 2022), possibly via the sheath of the optic nerve (Ghobrial et al. 2016), olfactory nerve or trigeminal nerve (Hanson and Frey 2007, 2008).

These are some of the rare pathways through which cryptococcus enters the brain. Although these cases are rare, attention should be paid to these possibilities in the clinical, particularly when examining patients who are immunocompromised.

3. The blood-brain barrier

3.1. Anatomy and physiology of the BBB

The level of difficulty by which solutes in the blood move from brain capillaries into brain tissue varies: pass quickly, some pass slowly, and some cannot pass at all. The highly selective permeability of the brain suggests that a structure restricts the passage of solutes; this structure is the BBB (Verkhratsky and Pivoriunas 2023). The BBB separates blood plasma from brain cells and can prevent certain substances (mostly harmful substances or drugs) in the bloodstream from entering the brain (Zhang et al. 2022). This structure exhibits selective permeability, preventing or minimising brain tissue damage caused by harmful substances in the circulatory system blood, ensuring the dynamic balance of the environment within the brain, and helping maintain the physiological homoeostasis of the CNS (Zhang et al. 2022).

The BBB is composed of a continuous layer of brain endothelial cells connected with tight junctions (TJs) between cells, as well as a complete basement membrane with pericytes, glial limiting membrane surrounded by astrocyte endfeet (Zhang et al. 2022). Together, these components constitute the brain endothelium and the main structure of the BBB (Figure 2). Compared with capillaries in other tissues and organs, brain capillaries differ in structure and function: (a) Brain capillaries lack pores (or have a few small pores) that non-brain capillaries have; (b) Endothelial cells overlap each other and are tightly connected through TJs (Abbott et al. 2006; Pivoriunas and Verkhratsky 2021), effectively preventing large molecules from crossing between endothelial cells; (c) Human brain microvascular endothelial cells (HBMECs) exhibit a low endocytosis rate (Daneman and Prat 2015; Pivoriunas and Verkhratsky 2021), and (d) Endothelial cells are surrounded by a continuous basement membrane, and approximately 85% of the surface of BMECs outside the basement membrane is surrounded by the end feet of perivascular astrocytes (Daneman and Prat 2015). At present, it is generally believed that the permeability of the BBB depends mainly on BMECs. However, some studies have shown that astrocytes secrete bioactive molecules and extracellular vesicles that can be taken up by endothelial cells to regulate the permeability of the BBB (Verkhratsky and Pivoriunas 2023), possibly controlling the integrity of the BBB by regulating the expression of TJ proteins (Delpino et al. 1995; Wey et al. 2008; Pivoriunas and Verkhratsky 2021; Woo and Martinez 2021). In conclusion, the function of astrocytes and their interactions with cryptococci need to be further explored. Other studies have shown that the loss of pericytes may increase the permeability of the BBB, and pericytes exhibit different recovery rates after the BBB is damaged (Geranmayeh et al. 2019). However, there are only a few related studies, with most studies having been focused on HBMECs.

Figure 2. Anatomy and physiology of the blood-brain barrier (BBB) (By Figdraw).

CNS diseases can cause marked changes in the structure and function of the BBB, such as disruption of TJs, enlargement of the intercellular space, and a large increase in the permeability of the barrier, allowing large molecules such as plasma albumin to pass through the barrier (Zhang et al. 2022). Damage to the BBB leads to the entry of pathogens into the brain, causing CNS diseases (Shi et al. 2012). In addition, pathogen infections abrogate the integrity and permeability of the BBB and lead to increased entry rates of viruses, immune cells, or cytokines into the brain, leading to increased inflammatory responses in the CNS and further damage to the brain (Liu et al. 2019). A study revealed that cryptococcal meningitis in patients with AIDS can lead to reduced activation of astrocytes (Tripathi et al. 2014), which can further damage the BBB.

3.2. The in vitro model of BBB

In various studies on the mechanism by which microorganisms cross the BBB, in vitro models are often used (Kim 2006; Sorrell et al. 2016). The establishment of an in vitro model takes advantage of the speciality of BMECs, that is, TJ formation, as well as the ability to form a polar monolayer in culture (Weksler et al. 2005). The simplest model is a single-cell culture model in which a culture medium is added to the upper chamber and a lower chamber of a permeable insert, with BMECs inoculated onto the membrane that separates the chambers (Kim et al. 2021). In this model, the upper chamber represents blood and the lower chamber represents the brain, therefore, the system simulates the process of microorganisms penetrating BMECs into the brain. Coculture models, including double-layer coculture models (BMECs and astrocytes) and triple-layer coculture models (BMECs, astrocytes, and pericytes), can also be established. Notably, in vitro models represent static conditions, not dynamic conditions simulating blood flow in capillaries (Kim et al. 2021).

4. The mechanisms by which Cryptococcus crosses the BBB

In most cases, cryptococcal meningitis is believed to result from haematopoietic dissemination, with cryptococci colonising the lungs and then travelling through the bloodstream to the BBB, which they penetrate to enter the brain directly (via transcellular traversal or a paracellular pathway) or indirectly (via a Trojan horse mechanism) by transversing BMECs, causing CNS diseases. A study on Cn-infected mouse models showed that paravascular microabscesses formed after cryptococci entered the brain and moved through cortical microvessels and blood vessels (Perfect 2004), with the development of meningitis after secondarily spreading to the meninges. Regarding the mechanisms underlying cryptococci crossing the BBB, debates have focused on whether cryptococci are transported through the blood and across the BBB as free cells or encased in phagocytes (the so-called Trojan horse mechanism). In cellular and animal models of cryptococcosis, it has been shown that free cryptococci can either engage in specific ligand-receptor interactions and then cross the BBB via transcytosis (Jong et al. 2007a, 2008a, 2008b, 2012) or cross the BBB after mechanical blockage mediated by cryptococci in capillary branches followed by mechanical or biochemical disruption of the capillary walls, which consist of endothelial cells and their TJs (Olszewski et al. 2004; Shi et al. 2010). An increasing number of recent studies have shown that cryptococci are transported through the bloodstream within phagocytes and then cross the BBB via the Trojan horse mechanism.

Sorrell et al. (2016) suggested that Cn typically causes meningitis in HIV-infected hosts. In addition, of 86 Cg-infected patients in Australia between 2000 and 2007, most patients (85%) presented with meningoencephalitis (Chen et al. 2012; Miyazato 2016). However, no studies have revealed whether Cn and Cg cause meningitis differently or whether Cn or Cg is more likely to cause meningitis via a preferential route through the BBB. Many studies have shown that HIV promotes cryptococci crossing the BBB, leading to cryptococcal meningitis in patients with AIDS. Moreover, the HIV gp41 protein has been found to stimulate the binding of Cn to BMECs, thereby increasing the risk of brain infection (Jong et al. 2007b; Huang et al. 2011b). In addition, corresponding studies have shown that HIV infection increases the ability of peripheral monocytes to enter the brain (Wang et al. 2008). Relevant mechanistic studies have revealed that the high-affinity FC-γ receptor 3A on phagocytes promotes the phagocytosis of cryptococci, thereby causing cryptococcal meningoencephalitis (Rohatgi et al. 2013). MYOC, a cytoskeleton-related gene identified as an important factor in the Trojan horse mechanism, is one of the central regulators of the miRNA-mRNA regulatory network and is significantly downregulated during Cn infection. In addition, MYOC is overexpressed in THP-1-derived macrophages in HIV-infected patients infected with cryptococci, which increases the phagocytosis activity and migration ability of macrophages, thereby inducing more cryptococci to be taken up macrophages and enhance the Trojan horse mechanism mediated by macrophages (Li et al. 2022).

4.1. Transcellular traversal

Transcellular traversal refers to the penetration of barrier cells by cryptococci through cellular endocytosis (Charlier et al. 2005; Chang et al. 2011). By this mechanism, cryptococci must adhere to and be internalised from the surface of the brain microvascular lumen (blood side) by BMECs, after which cryptococci migrate through the endothelial cytoplasm to the parietal lumen side of the BMECs (the brain side) to cross BBB. Chang et al. (2004) directly observed the whole process of cryptococci passing through BMECs in vitro via electron microscopy. After cryptococci adhered to the surface of endothelial cells, the formation of microvillus-like membrane protrusions was observed on the surface of BMECs, and then, the yeast cells were gradually enveloped. However, this process was only directly observed in vitro, not in vivo. The histopathology results obtained by Chang et al. (2004) showed that cryptococci initially accumulated near brain capillaries and then appeared in the brain parenchyma far away from these capillaries, and ten days later, cryptococci were observed near the meninges. The adherence of cryptococci to HBMECs is the first step in pathogen crossing the BBB via transcellular traversal. Cryptococci promote their transcellular traversal across the BBB via different mechanisms described in detail below.

4.1.1. Inositol

Inositol is one of the most abundant metabolites in the human brain. A high concentration of inositol has been detected in astrocytes directly associated with the BBB, and inositol can be released from cells rapidly under hypertonic conditions (Isaacks et al. 1994; Fisher et al. 2002). It has been suggested that the high abundance of inositol in the host brain is a factor contributing to the high incidence of cryptococcal meningitis (Wang et al. 2011a). Liu et al. (2013) found that cryptococci leverage a complex system to obtain inositol from the environment and that they use host inositol, which binds to their inositol transporter, thereby promoting their adhesion to HBMECs (Figure 3). It has also been reported that inositol is absorbed from the inositol transporter on the cryptococcal cell membrane, followed by high CPS1 gene expression, which stimulates the expression of hyaluronic acid on the cryptococcal cell membrane (Figure 3). Hyaluronic acid binds CD44 receptors on BMECs of the CNS, further increasing the adherence of cryptococci to HBMECs (Liu et al. 2013; Miyazato 2016).

Figure 3. Model of transcellular traversal by which Cryptococcus crosses the blood-brain barrier (BBB). Cryptococci use host inositol which binds to their inositol transporter. Inositol is absorbed from the inositol transport tube on the cryptococcal cell membrane, followed by activating genes such as CPS1, which induces the expression of hyaluronic acid on the cryptococcal cell membrane. Hyaluronic acid binds to the CD44 receptor on the surface of HBMECs. Cryptococci induce EPH-EphrinA1 (EphA2) phosphorylation through transactivation of CD44, which promotes GTPase-dependent signalling. It induces the up-regulation of cytoskeleton-related proteins, reorganises the actin cytoskeleton, and internalises cryptococci via endocytosis. Cryptococci secrete metalloproteinase Mpr1, which binds to the Mpr1 target protein AnnexinA2 (AnxA2) inside the HBMECs. This also induces the up-regulation of cytoskeleton-related proteins, which triggers the reorganisation of actin cytoskeleton, thus cryptococci transcytosis and exit through HBMECs to cross BBB (By Figdraw).

4.1.2. CD44 receptor

Hyaluronic acid binds to CD44 receptors (Vu et al. 2013) (Figure 3). CD44, a cluster of differentiation glycoprotein, is located in lipid rafts/vesicles on the surface of BMECs (Huang et al. 2011a; Long et al. 2012). It has been shown that CD44 knockout mice infected with Cn survived longer than infected wild-type mice and presented with a lower brain fungal burden (Jong et al. 2012), suggesting that CD44 may play a critical role as the host receptor during cryptococci crossing the BBB. The expression of hyaluronic acid on the surface of the cryptococcal capsule is closely related to the expression of the CPS1 gene, and when the hyaluronic acid on the capsule is abrogated, the ability of cryptococci to adhere to HBMECs is significantly reduced (Jong et al. 2007a). After successful adhesion of cryptococci to the surface of BMECs, protein kinase Cα/dual-specificity tyrosine phosphorylation-regulated kinase 3-mediated downstream signalling pathways [the CPS1-CD44-PKCα/DYRK₃-actin filament signalling pathway (Jong et al. 2008a)] and the CD44-EPH-EphrinA1 (EphA2) tyrosine kinase receptor signalling pathway (Aaron et al. 2018) are activated, triggering actin cytoskeleton reorganisation and phagocytosis (Jong et al. 2008b, 2012) (Figure 3).

4.1.3. Metalloprotease Mpr1

A secreted metalloprotease Mpr1 was recently discovered in the extracellular proteome of Cn (CnMpr1), and some studies have indicated that it is required for fungal infection of the CNS (Vu et al. 2014; Pombejra et al. 2017; Aaron et al. 2020). Vu et al. (2014) found that Mpr1 facilitates the passage of Cn through BMECs and into the CNS by promoting the attachment of cryptococci to the endothelial cell surface, suggesting a key role for M36 protease in fungal pathogenesis, as it plays a key role in transcellular traversal. In addition, Pombejra et al. (2017) identified 62 protein targets of Mpr1 on the surface of the BBB, and among these proteins, transcytosis of cells lacking AnnexinA2 (AnxA2) and exit through HBMECs are prevented (Figure 3). Moreover, they found that Mpr1 promoted cytoskeletal remodelling of HBMECs and then facilitated BBB permeation mediated by the AnxA2-MPR1 interaction.

4.1.4. Molecules related to the cytoskeleton

Host actin cytoskeleton rearrangement was needed for Cryptococcus binding, invading, and migrating into HBMECs. Vu et al. (2013) found upregulated expression of cytoskeleton-associated proteins, such as myosin, transgelin, annexin A2, and S2A100 (proteins involved in the recruitment of factors to the cytoskeleton), which may have been a result of the rearrangement of the cytoskeleton in BMECs after the attachment and internalisation of Cn. These cytoskeleton-associated proteins have also been identified in other studies (Chen et al. 2003, 2011; Wang et al. 2011b). Kim et al. (2012) found that the three major Rho-GTPases, namely RhoA, Rac1, and Cdc42, were closely associated with the rearrangement of the actin cytoskeleton and that cryptococci induced the phosphorylation of focal adhesion kinase (FAK), ezrin, and protein kinase C alpha (PKCα), all of which are involved in the rearrangement of the host actin cytoskeleton. Their study showed that downregulation of FAK, ezrin, or PKCα significantly reduced the ability of cryptococci to cross a HBMEC monolayer, meaning that Cn-activated Rho-GTPases, FAK, ezrin, and PKCα to facilitate their subsequent crossing of the HBMEC monolayer, which is a critical step in the development of cryptococcal brain infection and meningitis (Kim et al. 2012). In addition, Li et al. (2021) found that Pdlim2 was associated with cytoskeletal factors in the cytoplasm, while myosin phosphatase Rho-interacting protein (Mprip) targeted myosin phosphatase to the actin cytoskeleton. Pdlim2 and Mprip were shown to cooperatively regulate intracellular traversal and exocytosis of Cryptococcus in HBMECs.

4.1.5. Others

Meredith et al. (2014) found that, in association with increased mitochondrial metabolism induced by glucose, the binding of advanced glycation end products (AGEs) to receptors for advanced glycation end products (RAGEs) led to increased BMEC permeability. This finding may explain the greater susceptibility of diabetic patients to cryptococcal meningitis. In addition, some studies have shown that inhibition of TLR-4/NF-κB/MMP-9 signalling pathways attenuates neuroinflammation reactions and reduces damage to the BBB (Nighot et al. 2019; Fu et al. 2021; Geng et al. 2022; Liu et al. 2022), but it is unclear whether Cryptococcus infection causes alterations in these signalling pathways. Given these findings, transcellular traversal causing Cryptococcus crossing the BBB is a hot topic, but there are still more mechanisms to be explored.

4.2. Paracellular pathway

The paracellular pathway refers to cryptococci crossing the BBB into the brain via spaces between damaged or weakened endothelial cell TJs (Kim 2008; Stie et al. 2009; Zhang et al. 2022). Notably, some studies have reported the temporary relaxation of TJs caused by osmotic or bioactive agents (Barar et al. 2016), such as bradykinin, cereport, and histamine. However, it is not clear whether Cryptococcus infection activates these alterations and thus promotes pathogen crossing BBB via the paracellular pathway. There are several studies on Cryptococcus crossing the BBB via the paracellular pathway.

4.2.1. The urokinase-plasminogen-plasmin pathway

Urokinase is a plasminogen activator that activates the conversion of plasminogen to plasmin in the body. Plasmin degrades the fibrin-enriched extracellular matrix and basement membranes and activates other zymogen-protease systems that affect vascular permeability, thereby enhancing pathogen invasiveness. Stie et al. (2009) found that cryptococcal cell wall surface receptors bound to host plasminogen, and in the presence of tissue-derived plasminogen activator (tPA), cryptococci that bind plasminogen induced its conversion into plasmin, which binds and breaks down the extracellular matrix and cell membrane, thereby enhancing the ability of cryptococci to invade the extracellular matrix. Moreover, coculturing plasminogen-bound Cryptococcus with BMECs promoted the conversion of plasminogen to plasmin without the addition of tPA and enhanced the ability of cryptococci to invade the extracellular matrix (Stie and Fox 2012), suggesting that HBMECs may have the ability to activate plasminogen. This group also found that HBMECs in coculture with live Cryptococcus were induced to secrete urokinase and that urokinase activated plasminogen on the cryptococcal surface into plasmin, enhancing the ability of cryptococci to cross the BBB, while siRNA-mediated silencing of urokinase gene expression or the use of specific inhibitors of urokinase activity eliminated plasminogen-to-plasmin conversion and reduced the capacity of cryptococci to cross the BBB (Stie et al. 2012). These findings suggest that Cryptococcus can use the host urokinase-plasminogen-plasmin pathway to promote its traversal of the BBB (Figure 4a).

Figure 4. Model of the paracellular pathway by which Cryptococcus crosses the blood-brain barrier (BBB). (a) HBMEC was induced to secrete urokinase by Cryptococcus stimulation. Urokinase activates the conversion of plasminogen on the surface of cryptococci to plasmin, which degrades the fibrin-enriched extracellular matrix and basement membranes. This facilitates Cryptococcus to cross the BBB. (b) Cryptococcus secretes urease, which converts urea into ammonia. The local production of ammonia may damage endothelial cells, impair tight junctions, and widen endothelial cell gaps, thereby causing the migration of cryptococci into the brain (By Figdraw).

4.2.2. Urease

Olszewski et al. (2004) utilised C. neoformans H99, the urease knockout strain H99 (ure1), and urease recovery strain ure1+URE1-1 to explore the role of cryptococcal urease in lung-to-CNS cryptococci transmission and invasion. They speculated that urease potentiates yeast adhesion to BMECs during haematogenous transmission, thereby facilitating blood-to-brain invasion. In addition, Charlier et al. (2005) employed horseradish peroxidase (HRP) to determine whether the entry of cryptococcal yeast cells into the brain parenchyma was associated with functional lesions in the BBB, and they found that damage causing lesions in the BBB and BMECs were evident as early as 6 hours in mice inoculated with Cn and observed leakage of HRP as well as alterations in the basal lamina and endothelium cells. They thus speculated that functional and morphological changes in the BBB appear shortly after cryptococcal fungemia and may lead to the disruption of TJs between HBMECs, causing yeast cells to invade the brain through the intercellular spaces formed. Shi et al. (2010) injected fluorescently labelled urease-mutant strain C. neoformans intravenously into mice and performed cerebrovascular imaging of the mice. The results showed that the ability of the Cn urease mutant strain to enter the brain was significantly reduced. This result suggested that urease may cause biochemical disruptions to brain endothelial connections (Vu et al. 2013). Another study found that accessory proteins Ure4, Ure6, and Ure7 were all required for the function of urease, while mutants of the core urease protein Ure1, accessory protein Ure7, and transporter protein Nic1 all attenuated the invasion of pathogen into the CNS, suggesting that urease activity may directly affect the integrity of TJs between the endothelial cells of the BBB (Morrow and Fraser 2013). Urease, which can convert urea to ammonia, may act be one of the enzymes that dissolves tight junction proteins. Specifically, the local production of ammonia may damage endothelial cells, impair tight junctions, and widen endothelial cell gaps, thereby increasing the permeability of HBMECs and leading to the migration of cryptococci (Yang et al. 2017) (Figure 4b).

Notably, these experiments provide only indirect evidence of the paracellular pathway, and there is no direct evidence showing that cryptococci enter the brain through this pathway. As a result, finding direct evidence or even observing cryptococci invading the brain via the paracellular pathway via image-assisted technology is a future research suggestion.

4.3. Trojan horse mechanism

The Trojan horse mechanism refers to pathogens, after being internalised by phagocytes, that are not digested but survive within the phagocytes (Alanio et al. 2015) and enter the brain as the phagocytes cross the BBB (Charlier et al. 2009; Vu et al. 2013). The Trojan horse mechanism has been a hot research topic in recent years.

The Trojan horse mechanism has been shown to exist in bacterial pathogen infections, such as beta-haemolytic Streptococcus (Nizet et al. 1997) and Listeria monocytogenes infection (Drevets et al. 2001, 2004; Drevets and Bronze 2008; Disson and Lecuit 2012), and viral pathogen infection, such as human immunodeficiency virus infection (Fischer-Smith and Rappaport 2005). In addition, different research teams have provided strong evidence for the migration of cryptococci via the Trojan horse mechanism. Charlier et al. (2005, 2009). found that depletion of circulating monocytes induced by chlorophosphate reduced the severity of cryptococcal meningitis and the brain fungal burden. This means that a decrease in the phagocytosis rate leads to a decrease in brain fungal burden. In mice injected with monocytes infected with Cn in vitro showed increased brain fungal burden. In this regard, Santangelo et al. (2004) performed a similar experiment in which lung macrophages and peripheral blood mononuclear cells from mice infected with Cn via inhalation were injected into healthy mice, which developed cerebral cryptococcosis. Sorrell et al. (2016) performed experiments with FITC-labelled cryptococci, and after phagocytes engulfed the FITC-labelled cryptococci, the unengulfed cryptococci were stained with Uvitex to distinguish them from the engulfed cryptococci. Then, fluorescence microscopy was used to observe the process of phagocytosis and BBB crossing in vitro. The result suggested that the cryptococci engulfed by phagocytes crossed the BBB via phagocyte migration (Sorrell et al. 2016) (Figure 5a). In addition, they found that compared with Cg, more Cn crossed the BBB via the Trojan horse mechanism (Sorrell et al. 2016), which may support the idea that Cn is more likely to cause neurological disorders. However, this experiment illustrated only the difference between the two fungal isolates that cross the BBB via the Trojan horse mechanism; the exact reason for this difference is still unknown. Moreover, this study illustrated only the difference between Cn and Cg penetration rates of the BBB through the Trojan horse mechanism, the exact reason for which is still unknown, and the differences in the infectivity through the other two pathways remain to be verified.

Figure 5. Model of Trojan horse mechanism by which Cryptococcus crosses the blood-brain barrier (BBB). The phagocytosis of Cn inhibited the chemotaxis of macrophages stimulated by CX3CL1 and CSF-1. Phagocytes infected with Cryptococcus adhere to HBMEC and utilise their nuclear lobes to create gaps between (a) or within (b) endothelial cells. Nuclear lobes insert into the HBMEC gap, and then phagocytes enter the brain. Then cryptococci in phagocytes are carried to the side of the brain by this behaviour of phagocytes, and cryptococci continue to be wrapped in phagocytes or be released.

In addition, the interpretation of the Trojan horse mechanism is difficult and complex because cryptococci can grow both inside and outside phagocytes and can move to both sides of the phagocyte membrane while remaining viable in vivo. In other words, cryptococci may be transported through the bloodstream within phagocytes but are expelled after phagocytes reach the BBB, or after they are expelled before phagocytes attach the BMECs. That is, free cryptococci cross the BBB. The first possibility seems to have been verified by Santiago-Tirado et al. (2017), who directly observed phagocytes containing cryptococci crossing the BBB via live-cell microscopy, providing evidence of the Trojan horse mechanism. In addition, they observed that phagocytes moved from the blood to the brain through a “donut hole” formed in endothelial cells without entering the cytoplasm and moving between HBMECs (Figure 5b). However, this finding has not been verified by others. Whether phagocytes pass the BBB through a “donut hole” formed in the endothelium, through intercellular spaces, or both remains to be explored.

A recent study found that the phagocytosis of Cn inhibited the chemotaxis of macrophages stimulated by CX3CL1 and CSF-1 (Luo et al. 2009), and Zhang et al. (2015) suggested that this effect may slow the crawling of Cn-containing macrophages along the cerebral vasculature, thereby facilitating the transport of Cn (Figure 5). To induce an appropriate immune response to infection, macrophage metabolism is increasingly switched to aerobic glycolysis with the simultaneous decrease in the oxidative phosphorylation rate, so that the end product of glycolysis, lactic acid, accumulates in the extracellular environment; this process is similar to the Warburg effect (Cruz-Gregorio et al. 2023), which refers to the fact that pyruvate produced by massive glucose uptake by tumour cells does not undergo oxidative phosphorylation, but is lysed in the cytoplasm to form a large amount of lactate in the normoxic environment. Moreover, a study revealed that lactate significantly increased bacterial clearance from human macrophages infected with Mycobacterium tuberculosis (Maoldomhnaigh et al. 2021). The accumulation of lactic acid helps macrophages clear cryptococci and thus reduce the brain fungal burden is a possibility worthy of deeper investigation. In studies on the mechanisms by which leukocytes cross the endothelial barrier, it has been shown that the nuclear lobes in leukocytes are the sites of gap and pore generation between and inside endothelial cells, enabling reversible bending of endothelial actin stress fibres and leukocyte movement through the endothelial barrier via rapid turnover of endothelial actin filaments, not endothelial cell contraction (Barzilai et al. 2017) (Figure 5). Whether a similar mechanism enables phagocytes to cross the cerebral vasculature and BBB should be further investigated.

The studies on the Trojan horse mechanism have produced mostly indirect evidence, and the direct evidence that has been presented to date is not sufficient. Moreover, the experiments related to the Trojan horse mechanism have been mostly based on in vitro models because the results of experiments performed in vivo are more difficult to visualise. Recent research on this mechanism has largely fallen short of confirming that cryptococci cross the BBB via this mechanism, and therefore, the true mechanism underlying entry into the brain remains to be identified.

5. Conclusions

This paper focuses on the interaction between cryptococci and the BBB and describes the three mechanisms related to cryptococci crossing the BBB that have been reported thus far, but the human body is a complex system with a multilayered structure, and there are still more mechanisms to be explored.

In addition, among numerous case reports of cryptococcal infections, both Cn and Cg cause cryptococcal meningitis. Do these passages follow different routes of entry into the brain? Most studies have focused on Cn and not Cg, so we know little about the mechanisms by which Cg crosses the BBB and the reasons for the differences between Cn and Cg infection rates. We speculate that there are differences in the mechanisms by which Cn and Cg cross the BBB. It may be manifested in the speed of crossing the BBB, and it may be reflected in the difficulty crossing the BBB. In another, Cn is more likely to infect immunocompromised patients, the deficiency of immunocytes that interfere with the clearance of Cn, and mechanisms by which HIV promotes Cn to cross the BBB are also related. In subsequent studies, attention should be paid to whether both Cg and Cn penetrate the BBB by all three mechanisms or whether only one or two mechanisms are involved, differences in the amount or rate of Cg and Cn entering the brain via the same mechanism need to be resolved and whether Cg or Cn is more likely to cause meningitis needs to be determined. Furthermore, it is inconclusive which of the three mechanisms for crossing the BBB predominates, although transcellular traversal was the first to be identified. In addition, due to the high mortality rate of cryptococcal meningitis, the development of therapeutic drugs is urgently needed. There are currently a few drugs for the treatment of fungal infections caused by Cryptococcus, such as Amphotericin B, which is a commonly used antifungal drug (Mahendrarajan and Bari 2022). The elucidation of the mechanism by which cryptococcal cells cross the BBB will facilitate the development of relevant drugs that impede them into the brain. This type of prevention will greatly reduce the incidence and mortality of cryptococcal central nervous system diseases. In conclusion, blocking cryptococcal brain entry may facilitate the development of a new therapeutic intervention that will benefit patients suffering from cryptococcal meningitis.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Key Research and Development Programme (2021YFC2302100), Beijing Hospitals Authority Clinical Medicine Development of Special Funding Support (XMLX202115).