4.1. Phospholipases
The secretion of extracellular phospholipases (PL) by
C. neoformans was investigated by Chen et al., where a cluster of PL such as PLB, lysoPL hydrolase, and lysoPL transacylase/acyltransferase was detected and every
Δplb1 mutant usually marked with a drastic reduction in the activity of each enzyme compared to the
wt [
19,
20]. By popular demand, PLB is generally used to represent the three enzymes. By correlation, PL activity corroborated the mortality rate and mucosal invasion by
Candida albicans in animal models [
21]. Similarly,
C. neoformans may use these hydrolytic enzymes to invade tissue and cross the BBB. The production of PL appeared similar in environmental and clinical strains of
C. neoformans var.
gattii, but the environmental isolates of
C. neoformans var.
neoformans produced more PL than the clinical isolates [
19]. Various isolates of
C. neoformans possess a different degree of PL secretion, which correlates with the fungal virulence and tissue burden in the lungs and brains of mice inoculated intravenously [
19,
20].
PLs, best secreted at 37
oC [
22], are a highly diverse group of hydrolytic enzymes targeting the ester linkages of glycerophospholipids carbonyl linkage and can also transfer acyl chain to the lysophospholipid to form diacylglycerophospholipid. These three oligomeric proteins are borne on the same coding gene, the
Plb gene, just like in other yeasts such as
Penicillium notatum [
23],
S. cerevisiae [
24], and
C. albicans [
25]. Chen et al. showed that PLs are well secreted at about 6 – 10 hours of cell culture at 30
oC, with optimal activity at pH 3.5 – 4.5 and stability at pH 3.8. The enzyme activity was not affected by exogenous serine protease inhibitors such as leupeptin, phenylmethylsulphonyl fluoride (PMSF), divalent cations (Ca
2+, Mg
2+, and Zn
2+), and EDTA [
19]. The stability of this enzyme at acidic pH vividly reveals one of the strategies put in place by
C. neoformans to tolerate the micro-acidic environment of the phagosome, and, in so doing, the PLs may break down the phagosomal membrane leading to tissue dissemination.
The
Ste12α gene has been implicated in the regulation of extracellular PL. Deletion of the
Ste12α gene significantly reduced the production of PL in the egg yolk agar media [
26]. As crucial as PLs are to the virulence of
C. neoformans, the absence of these hydrolytic enzymes may not affect the masterminding virulence factors/phenotypes such as growth at 37
oC, urease activity, capsule, and melanin formation. However, such a mutant is less virulent in the mouse inhalation and rabbit meningitis models when compared to the
wt in so much that its growth is defective in a macrophage-like cell line [
20]. The attenuated virulence observed in the
Δplb1 mutant must have come from the lower density of the capsule; notwithstanding, the capsule and the cell wall dimensions are relatively retained compared to the
wt [
22].
The attenuated virulence observed in the
Δplb1 mutant is likely comparable to
Δure1 and
Δlac1. PLB and their components are generally associated with
C. neoformans virulence, yet an infection can still progress in their absence but at a slower rate. Increasing the temperature from 30 – 37
oC promotes transmigration of the PLB from the cell membrane to the cell wall with a concomitant increase in the cytoplasmic translation, perhaps due to constitutive secretion from the Golgi apparatus. However, secretion of PLB under this heat stress is highly minimised to maintain cell wall integrity and promote membrane homeostasis [
22].
The homologous
Ssn3/Ume5 gene was serendipitously discovered adjacent to the
Plb1 locus in
C. neoformans. Thus, like
S. cerevisiae, C. neoformans may be using the product of this low transcript to repress the transcription factor Hsp70 family. However, this repressing protein, known as Ume5p, is a cyclin-dependent protein kinase, which is sensitive to temperature- or oxidative-induced degradation via ubiquitin-mediated proteolysis [
27], and may not be playing any significant role in the virulence factors [
20].
Covalently, PLB is usually attached to the cell wall chitin protruding β-1,3-linked glucan via its
β-1,6-linked glucan. This arrangement is anchored by the glycosylphosphatidylinositol (GPI) on the membrane lipid raft proteins. The release of PLB from the GPI-anchor is probably orchestrated by the phosphatidylinositol-specific phospholipase C (PI-PLC, putatively encoded by
Plc1 and
Plc2), and the inclusion of β-1,6-linked glucan to this release confirms the cell wall localisation of the PLB [
22,
28]. This shows that PLB constitutes part of the proteins involved in the cell wall integrity because
Δplb1 exhibited morphological cell wall defect, which is sensitive to disruption by SDS and Congo red; however, the growth is not affected by caffeine [
22]. The
Δplc1 C. neoformans var.
grubii strain H99 mutant, which failed to produce melanin and refused to grow at 37
oC, exhibited less secretion of PLC with concomitant accumulation of cytoplasmic PLB because of the impaired secretion of PLB [
28]. Furthermore, this mutant displayed high antifungal sensitivity (to 5-FC, azoles, and AmpB but not CpF), poor replication, and defective cell wall characterised by cell clumping and irregular morphology as confirmed by the diffusion of Calcofluor white staining and sensitivity to cell wall disruptors (Congo red, caffeine, and SDS – check
Supplementary 1 endnote for details) due to impaired activation of Pkc/MAPK pathway [
28].
Because the
Δplc1 mutant had poor growth at 37
oC, its virulence was attenuated in the mice study. Not only this, even at 25
oC, this mutant failed to kill a significant number of
Caenorhabditis elegans compared to the
wt and reconstituted mutant [
28]. Factually, the few numbers of
C. elegans killed by this mutant means that attenuated virulence associated with this mutant is independent of temperature; after all, other virulence determinants are simultaneously attenuated as well. The actual killing of the worms at 25
oC by the
wt and reconstituted mutant indicates that the Plc1 virulence can be temperature-independent and that certain other factors may augment this virulence. In all, the
Δplc2 mutant failed to exhibit any other major defect, and it is as active as the
wt.
The IP
3 produced by Plc1, besides the ER Ca
2+ release, serves as a substrate for IP
3 kinase (encoded by
Arg1) to produce inositol polyphosphate. Expectedly, the
Δplc1 mutant generally accumulates PIP
2 but a reduced level of IP
3, whereas in the
Δarg1 mutant, IP
3 will be accumulated [
29]. Invariably, these two mutants shared similar phenotypic defects, including impaired thermotolerance, defective cell walls, impaired virulence factors, improper cell division, and defective mating and filamentation. Further analysis showed that
Δarg1 accumulated larger intracellular vacuoles (excessive vacuolar fusion) than the
wt and
Δplc1 mutant [
29]. Therefore, Arg1p inositol polyphosphate anabolism (IP
4-8) is as important as the Plc1p catabolism for
C. neoformans virulence, and that IP
3 is a biological relay molecule for active Plc1p.
In another related work, efficient antioxidant control has been linked to PLB production. Mutation of Cu/Zn SOD encoded by the
Sod1 gene has been characterised by a severe reduction in the activities of phospholipase B, urease, and laccase leading to attenuated virulence manifested in the low brain colonisation and persistence. However, this failed to affect the capsule formation under LIM, and the mutant culture media were characterised by high mannitol production that probably acted as an antioxidant to complement the loss of Cu/Zn SOD activity [
30,
31] (for details on the phenotypic display of different cryptococcal cell mutants, check
Supplementary 1).
4.3. Iron Depletion
Strategically, human physiological conditions never allow invading pathogens easy access to iron because iron is not usually found circulating without being locked in a storage protein like transferrin or transport protein like haemoglobin. Hence, invading pathogens must strive to obtain iron through the arsenal of high-affinity iron uptake and acquisition proteins. The arrays of membrane transporters (such as siderophores) and intracellularly expressed regulatory proteins (such as Cir1p) are released for iron acquisition and homeostasis (for details of cellular events that induced/repressed different membrane transporters and permeases in
C. neoformans, check
Supplementary 2). Interestingly,
C. neoformans mainly released specific iron acquisition proteins based on the available iron sources
in vivo or
in vitro. Besides,
C. neoformans can direct the expression of
HapX or
Hap3 to shut down other iron-dependent functional pathways in LIM to maximise the utilisation of the scarce iron for more growth while capsule production is being initiated. Generally, capsule production enlarges as iron decreases with concomitant slower growth.
The expression of the GATA iron regulatory factor,
Cir1, is majorly responsible for iron uptake for growth at 37
oC and the formation of virulence factors in the host tissue [
40]. This expression is enhanced by the
Hap genes such as
HapX.
Cryptococcus ferroxidase (Cfo1) expression enables this fungus to survive LIM; however, Δcfo1 mutant or as a background mutant to the deletion of
Hap3, Hap5, or HapX cannot survive in the LIM unless feroxamine is supplied. Apart from the
Δcfo1 mutant that can survive in the media supplied with FeCl
3 and hemin, none of the doubly mutated iron gene mutants can survive in these two iron sources [
40].
It appears that
C. neoformans employed multiple alternative routes of acquiring iron from its environment for growth and proliferation. Though
Δcfo1 mutant may have attenuated virulence notwithstanding deletion of
HapX or
Hap3 gene failed to extenuate further the virulence of this mutant [
40]. Comparatively, the deletion of
HapX seemed to have a greater influence on various transcriptional genes than deleting
Hap3 or
Cir1 under iron-limiting or iron-repleting conditions [
40].
Paradoxically, though HapX seems to influence a greater percentage of other transcription factors under iron-limiting conditions, including siderophore transporters, yet deletion of
Cir1 had a greater influence on the high-affinity iron transporters; however, this function is overlapping to some extent under low iron conditions. HapX and Hap3 cooperatively repressed most iron-dependent mitochondrial functional pathways under low iron conditions, but this overlapping function is not found in the iron-repleting conditions. While HapX majorly controls the subset genes for siderophore transporters in different iron sources, Hap3 mainly inhibits other iron-dependent regulatory pathways, especially in the mitochondria [
40].
Though Cir1 and HapX positively regulate
Sit1 expression for siderophore transporter, only Cir1 appeared to regulate virulence genes such as
Lac1. Also, a report has shown that Cir1 positively affected the expression of Rim101 pH-responsive transcription factors, which further regulates the expression of various iron acquisition genes such as
Sit1 and
Cft1 [
40]. Under low iron content, HapX influences the expression of
Cir1 and
Rim101, but under iron-repleting conditions, HapX apparently loses its influence while Cir1 influences the expression of
rim101 but withdraws its influence on HapX to minimise redundancy of HapX perhaps. In a related study, Grx4, a 2Fe-2S cluster sensing and trafficking protein, was identified as an essential regulator of iron homeostasis capable of associating with Cir1p in the nucleus during iron depletion to induce multiple iron acquisition proteins [
41,
42]. This synergistic association may probably be under the influence of HapX during iron depletion. Reports showed that iron repletion promotes relocalisation of the Grx4p into the cytoplasm; besides, deletion of the
Grx4 gene alone is sufficient to dysregulate iron-dependent metabolisms leading to oxidative stress, mitochondria dysfunction, and impairment of DNA repair [
42].
The ability to utilise different sources of iron is peculiar to
C. neoformans and other invasive pathogenic fungi like
C. albicans. Iron is usually acquired in a reduced form from inorganic (hemin, feroxamine B, and FeCl
3) or organic/biological form (transferrin and heme), and the effective absorption is centred on the interplay of
Hap, Cfo1, and
Cft1 genes. Acquisition of iron involves the reduction of Fe
3+ by the cell surface reductases (Fre) to Fe
2+ and then oxidised by the ferroxidase (Cfo1), which is coupled to
Cryptococcus high-affinity iron permeases/transporters (Cft1 and Cft2). Besides,
C. neoformans can acquire the non-reduced Fe
3+ through the ferric-bound siderophores transporter (Sit1) presented by other closely associated microorganisms. This attribute is also common in other fungi, such as
C. albicans and
S. cerevisiae [
43].
There are two putative orthologues of ferroxidase, Cfo1 and Cfo2, with the former being actively involved in the acquisition of iron while the latter plays a minor role. During infection, the austere iron withheld by the host tissues enables the fungus to express Cft1, which is trafficked to the plasma membrane under the influence of Pka1 to utilise transferrin for growth and virulence expression. The
Δcfo1 and
Δcfo1Δcfo2 mutants, irrespective of the temperature, are characterised by growth defect, attenuated virulence, copper toxicity, increased susceptibility to FCZ and AmpB due to impaired heme formation as a cofactor needed for ergosterol biosynthesis [
44]. However, an exogenous supply of heme and feroxamine (siderophore) but not FeCl
3 can restore some of these phenotypic deficiencies meaning that
Δcfo1 mutant can still utilise iron through the low-affinity uptake proteins such as Cfo2 and perhaps Cft2. Iron assimilation mediated by the Cfo1 expression is independent of the Sit1 (feroxamine) and heme; nevertheless, when supplied at 100 μg/mL, these two iron sources can restore growth and relative resistance against FCZ in
Δcfo1 mutants [
44].
The
Cft1-Cfo1 expression is needed to utilise iron from hemin and FeCl
3 but not from heme or feroxamine, and any mutant with single/double deletion of these genes may not survive in the low-iron environment [
44,
45]. Thus, it seems the
Cft1 and
Cfo1 may be clustered on the same locus as observed in serotype D, and their transcripts are synergistically elevated with other iron regulatory genes in a low-iron condition [
46]. Apart from these tandem iron-uptake-clustered genes, Lian et al. further suspected, in the LIM, other yeast orthologue genes involved in the intracellular iron transport and storage, and this includes:
- ❖
a low copy Yfh1 transcript (encoding frataxin that mediates mitochondrial iron efflux cytosolic unbound iron),
- ❖
Utr1p (augmenting ferric reductase activity),
- ❖
Atm1p/ABC inner mitochondria iron transporter,
- ❖
Mrs3p and Mrs4p (inner membrane protein suppressing mitochondrial RNA (mtRNA) splicing defect in the yeast, and also function as inner mitochondrial iron transporter),
- ❖
fet5 (encoding Fet3-related integral membrane multicopper oxidase required for high-affinity iron uptake across the vacuolar membrane in conjunction with Cft1p to mobilise iron from the storage), and
- ❖
ccc1 (encoding vacuolar membrane transporter to facilitate cytosolic reduction of iron while storing the iron in the vacuole) [
46,
47,
48,
49,
50,
51,
52].
The evidence that
Δhap3Δcfo1 and
ΔhapXΔcfo1 failed to utilise hemin and yet are relatively virulence in mice experiment means that other additional iron acquisition transporters, apart from the Cfo1, may be orchestrating iron acquisition during infection [
40].
The Cft1 and Cft2 are the two orthologues of high-affinity iron permease similar to the Ftr1 of
S. cerevisiae, and their expression is influenced by Pka. The Cft1 expression is more effective and highly involved in the growth and iron uptake from FeCl
3 and holo-transferrin than Cft2; however, the deletion of the two genes
(Δcft1Δcft2) attenuated the virulence accompanied by low tissue burden (for details on the phenotypic display of different cryptococcal cell mutants, check
Supplementary 1). In the presence of ≤1.0 mM FeCl
3, the
Δpka1 and
Δpka2 mutants significantly failed to raise the expression level of
cft1 compared to the
wt, and even the
Δpkr1 mutant further displayed reduced expression of
cft1 by 2 – 3-fold. More significantly, the
Δpka1 mutant displayed a more increased expression of
cft2 compared to the
Δpka2 mutant. In contrast, the
Δpkr1 mutant failed to show any significant effect on the expression of
cft2 when compared to the
wt [
45]. Though no significant contribution has been traced to Cft2 high-affinity iron uptake, the upregulation of this gene by Pka, however, might suggest the importance of this seemingly insignificant iron transporter. The reduced growth observed by Lian et al. with
Δcft1 mutant in the low- and high-iron media [
46] is redolent that Cft2 and perhaps Cfo expressions may be rescuing the mutant. This is like an SOS message received by the low-affinity iron uptake proteins to rescue the
Δcft1 mutant.
In the absence of ferrous iron,
C. neoformans may be compelled to use a non-reductive ferric-bound siderophore complex
in vivo or
in vitro as the iron source. Several pathogenic fungi have evolved to possess various specific siderophore transporters (Sitp) through which ferric-bound siderophores can be utilised by their competitors. These competitors may be other microorganisms capable of producing siderophore as deferoxamine, which can subtly steal iron from transferrin or heme and present it as siderophore feroxamine bearing Fe
3+. A putative
sit1 transcript identified in
C. neoformans encodes a siderophore transporter, which is highly expressed in cells cultured in the LIM alone or with a ferrous iron chelator, bathophenanthrolinedisulfonic acid (BPDA), supplemented with deferoxamine. Experimental observation from Tangen et al. showed that
Δsit1 mutants (either serotype A or D) could utilise ferrichrome but not feroxamine B as an iron source under iron-limiting conditions [
53].
Furthermore, deletion of the
Sit1 gene in the serotype D background enhances melanisation better than the serotype A
Δsit1 mutant. The melanin production seemed better with increasing DOPA concentration in the medium. However, 1.0% glucose repressed melanin formation in the two serotype mutants. Additionally, the
Sit1 expression appears to control copper availability to Cu
2+-binding pocket of laccase. So, the loss of
Sit1 gene apparently attenuates ionic homeostasis and reduces laccase activity more significantly in the
wt than the mutants. This means that laccase in the mutants has more access to copper because of the absence of Sitp, which ordinarily could have distributed the copper among other copper-dependent proteins. In this case, Tangen
et al. exogenously supplied copper and discovered an elevated laccase activity in the
wt and reconstituted
Δsit1 strain, which is comparable to the
Δsit1 mutant [
53]. Notwithstanding, an exogenous supply of copper or iron in the absence of glucose generally seems to favour melanisation in serotype D
ΔSit1 mutant.
In serotype A, the lack of the
Sit1 gene showed no significant effect on the
Lac1 transcript, even if copper or iron was supplied in a non-glucose medium. However, in serotype D, this mutant showed a significantly high level of
Lac1 transcript than the
wt and reconstituted mutant – only when copper or iron is exogenously supplied in the absence of glucose [
53]. With Calcofluor white, SDS, and Congo red in low-glucose or
L-DOPA medium, serotype D
Δsit1 mutant only showed a slight growth sensitivity to these compounds compared to the
wt, but no significant sensitivity was observed in the serotype A
Δsit1 mutant. Furthermore, serotype D
Δsit1 mutant showed increasing growth defect from 30 – 37
oC, but such temperature sensitivity is not conspicuous in the serotype A
Δsit1 mutant [
53]. This serotype differential phenotypic expression in response to copper and iron from
Sit gene re-emphasises the existence of cellular morphotypes in
C. neoformans.
To further emphasise the importance of
Sit1 expression in cell wall integrity,
Δsit1 mutants generally displayed a less dense and heterogenous cell wall thickness and polysaccharide fibril distribution. In addition, the serotype D
Δsit1 mutant cell wall is easily distorted compared to the serotype A
Δsit1 mutant, which reaffirms the phenotypical disadvantages earlier described in the serotype D
Δsit1 mutant. Therefore, it is not surprising to observe that serotype D
Δpka1 mutant could have a preponderant level of
Sit1 transcript and that the serotype A
Δpka1 and
Δpka2 mutants equally showed an elevated level of
sit1 transcript [
53]. This observation is similar to the negative regulation of iron uptake by the cAMP-Pka event in
S. cerevisiae [
54]. To this effect, the
Δsit1 mutant failed to improve defective capsule formation or induction in the
Δpka mutant cultured in the low-iron or iron-supplemented medium. Also, the deletion of
Sit1 did not initiate unilateral mating defect in serotype A
MATα Δsit1 x
MATa or serotype D
MATα Δsit1 x
MATa [
53].
Expectedly, the hypermelanised serotype D
Δsit1 mutant would have shown higher virulence than the
wt or reconstituted mutant; however, Tangen et al. could not observe any significant difference in the virulence among these strains. Not only this, but the virulence from this serotype mutant is generally weak [
53]. Still, serotype A
Δsit1 mutant, alongside the
wt and the reconstituted mutant, actually induced infection almost equally with no significant difference in the survival rate of experimental animals after the infection, tissue colonisation/burden within the brain and lungs, and extrapulmonary dissemination [
53]. Thus, this outcome generally presents the
Sit1 gene as not particularly required for virulence and infection in an animal model.
While Cir1 positively influences the expression of
Cft1,
Cft2 is negatively regulated. The transcription level of Cft1 in
Δcft2 mutant decreases just like in the
wt as FeCl
3 concentration increases. However, the
Cft2 transcripts level in
Δcft1 mutant increases with a threshold at 10 mM FeCl
3 but reduces in the
wt as the iron level increases. The
Sit1 transcript, on the other hand, is generally reduced in the
wt but becomes better responsive to FeCl
3 level ≤100 μM in
Δcft1 than
Δcft2; however, concentration beyond this makes Sit1 entirely unresponsive and drastically reduced in the
Δcft1 [
45]. This means that the loss of Cft1 potentially reduces the intracellular iron content, and this condition makes Sit1 transcription unresponsive even in the presence of 1.0 mM of FeCl
3. This, therefore, means that
Cft1 expression may come to rescue
Δcfo1 background mutant during infection. Not only this, but evidence also showed that
C. neoformans could secrete reductant metabolites such as
α-ketoglutarate, 3-hydroxyanthranilate, and melanin to reduce Fe
3+ to Fe
2+ nonenzymatically for extracellular absorption [
55]. This alternative reduction method might be responsible for late organ invasion observed with
Δcft1 mutant in experimental animals [
45]. The possibility also exists that though
Δcft1 and
Δcft1Δcft2 mutants cannot easily utilise transferrin
in vitro, heme might be unavoidably used to supply iron in the late systemic infection. Nevertheless, the brain fungal burden of these mutants is constantly lower compared to other systemic organ invasions, perhaps due to transferrin availability in the brain parenchyma.
Evidence showed that the pH of cryptococcomas (mass lesions within the brain parenchyma caused by the dissemination of cryptococcal cells) is between 5.4 – 5.6, and this favoured the accumulation of pro- and anti-inflammatory metabolites such as polyols, acetoin, dihydroxyacetone, and GABA [
56]. Accumulation of superoxide in this neutrophil micro-acidic milieu may have subtly deceived the host tissue into deploying iron into this environment. This condition favours iron acquisition through low-affinity/ferrous iron-specific systems in the
C. neoformans. Convincingly, the growth of
Δhap3Δcfo1, Δhap5Δcfo1, and
ΔhapXΔcfo1 mutants are better rescued at pH 5.0 than 7.0 in media supplemented with FeCl
3 or hemin. In addition, all the
Δhap and
Δcfo1 mutants failed to survive low iron irrespective of the pH, but all can utilise feroxamine at pH 5.0 and 7.0. The better growth of
Δcfo1 mutants in the presence of FeCl
3 and hemin, but the poorer growth of
Δhap mutants, showed that Hap proteins function better to utilise different iron sources in a mutant lacking
Cfo1 for high-affinity and reductive iron uptake system [
40].
There are many reports on the connection of iron level with major virulence factors and cell surface-associated proteins in
C. neoformans. With SAGE, iron-enriched media elevate various transcription factors connected to iron storage, glycolysis, mitochondria oxidative function, lipid and amino acid metabolisms, and calmodulin-calcineurin signalling events. However, other important transcription factors are well elevated in the LIM. Lian et al. reported an increased level of
Cap60 gene, mannoprotein MP88, and GPI transamidase (which mediates cross-linking of cell wall mannoprotein to
β-1,6-glucan) under iron-limiting conditions [
46]. The transcript level of
Cig1 encoding an extracellular glycoprotein (like mannoprotein) needed for capsule attachment to the cell, together with
α-1,3-glucan (synthesised by
α-1,3-glucan synthase), is usually found upregulated under the iron-limiting condition as well [
46,
57]. Though this gene responded to low-iron conditions where
Δcig1 mutant is, the mutant is characterised by poor growth in a low-iron medium and failed to suppress capsule formation under an iron-supplemented medium. This means that Cig1p may be involved in the iron-sensing relay process or probably participates in maintaining membrane iron retention through the siderophore transporter. Still, its role in capsule formation is dispensable.
4.6. Oxygen Depletion(Hypoxia)
C. neoformans is an encapsulated and environmental obligate aerobe (growing optimally under atmospheric oxygen at about 21% of the total air); therefore, adaptation to the relatively lower oxygen level of 1 – 5% in the human brain and lungs is inevitable to the survival of this fungus to cause infection. Oxygen availability is paramount to all the oxidative pathways, including mitochondria metabolism in this fungus and many other obligate aerobic basidiomycetes. These reactions are iron- and copper-dependent for electron transport leading to energy production. According to whole-genome microarray-based transcriptional profiling of
C. neoformans under hypoxic conditions, about 347 transcripts are differentially regulated [
68] and more than 100 expressed genes are statistically higher in the
wt than the
Δsre1 mutant under 1% O
2 level in 3 hours. However, more than 414 genes are significantly expressed in the
Δsre1 mutant than the
wt under the same condition [
69].
The major upregulated genes are involved in stress, iron and copper homeostasis, and the metabolism of sterol, heme, sphingolipid, sugar, and amino acids. In contrast, the downregulated genes are significantly involved in vesicular trafficking, cell wall formation, capsule synthesis, and ribosomal metabolism. Surprisingly, stress and antioxidant-related genes are highly upregulated in Δsre1 mutant via the cAMP/Pka or Hog1 pathway than the wt under the hypoxic condition. These genes include Tco2, Cat1, Cat3, Cas34, Hsp12, Hsp31, Gst1, Ecm4, and other genes involved in reductive pathways such as Gre2, Pkp1, and Ena1, some of which are evolutionarily conserved in the yeast.
The expression of
Gst1, Gst2, and
Gst3 encodes glutathione-
S-transferase to protect the cells against reactive oxidants. Also, upregulation of
Cat3 (encoding catalase A) in the
Δsre1 mutant is highly significant in the hypoxic condition to degrade the H
2O
2 to water, and the speculation for this is that
Δsre1 mutant may constantly be under more stress than the
wt whenever the oxygen is limited [
69]. In addition,
Hem13 expression appeared unchanged under hypoxic conditions either in the
wt or
Δsre1 mutant; however, because of the strong connection between the Sre1p, ergosterol synthesis, and ionic homeostasis, the
Sit1, Ctr4, Fre7, Erg12, and
Erg25 transcripts are usually drastically reduced in
Δsre1 mutants.
C. neoformans expresses the
Sre1 gene to sense oxygen and regulate sterol synthesis. This expression is a combination of endoplasmic reticulum membrane-bound Sre1p (homologue of mammalian sterol-response element-binding protein, SREBP) and its associated sterol- and oxygen-sensor activator (Scp1p orthologue of SCAP) with other possible accessory proteolytic enzymes such as a metalloprotein site-2-protease (Stp) and Kex2-like protease. Hypoxia (<3% O
2 level) or specifically low sterol level induces the proteolytic activation of Sre1p by Scp1p in the Golgi apparatus lumen and subsequently processed by Stp and Kex2 proteases to release a ≈75 kDa N-terminal transcription factor [
69]. Any antifungal that blocks ergosterol biosynthesis will induce Sre1p activation; however, extremely low oxygen levels (0.2% O
2 level) will drastically slow down
C. neoformans growth due to the slow activation of Sre1p [
69]. The nuclear-translocation of this basic helix-loop-helix (bHLH) leucine zipper DNA-binding domain of N-terminal activated Sre1p induces the plethora of transcription factors that enable the fungus to cope with oxygen-limiting conditions alongside sterol, iron, and copper depletions [
69,
70,
71,
72].
The interdependence of oxygen, copper, and iron to facilitate the environmental survival of
C. neoformans and cause infection in the human hypoxic tissues cannot be overemphasised. Heme is a cofactor for sterol production in fungi. The production of heme is iron-dependent, which is induced by hypoxia. Hence, it is unsurprising that
Δsre1 mutant, which failed to sense oxygen, is characterised by impaired ergosterol synthesis, iron, and copper uptake [
69]. Under normal conditions,
Δsre1 mutants tend to accumulate more sterol intermediates while the ergosterol level is reduced to basal need. These sterol intermediates become higher than in the
wt in hypoxic conditions as the ergosterol level reduces further because of the absence of
Sre1 expression. Thus, downstream accumulation of oxygen-dependent sterol intermediates under hypoxic conditions is necessitated in the
Δsre1 mutant except for some upstream oxygen-independent intermediates such as 24-methylene lanosterol, which seemed to be unperturbed in the mutant and
wt [
69].
Apart from the
Sre1 gene, Chun et al. established a hypoxia-sensitive trait in
Δtco1 mutant, a surprisingly hypermelanised avirulent strain with normal capsule production [
73]. All mutants generated, such as
Δsre1, Δscp1, Δstp1, Δtco1, and
Δsre1Δtco1, displayed normal growth in YPD at 37
oC with no
in vitro significant defect in the capsule formation but are sensitive to hypoxia condition, and the highest sensitivity is found in the double knockout mutant [
68]. Furthermore, these mutants displayed a slight defect in melanin formation (except the
Δtco1 mutant with hyper-melanisation), defective to moderate tissue proliferation vis-à-vis dissemination, antifungal susceptibility (except
Δtco1 mutant), and impaired ergosterol biosynthesis (except
Δtco1 mutant) [
68]. Enzymatically, the
Δsre1 mutant showed an insignificant defect in the activities of laccase, urease, and phospholipase, and the infection by this mutant failed to progress to terminal infection in an animal study due to limited brain lesions caused by poor growth in the brain though the mutant disseminated equally as the
wt [
69]. This
in vivo weak infection may also be corroborated by the relatively smaller capsule size of this mutant
in vivo [
69].
Strictly, brain regions are endowed with self-secure barriers and nutrient stinginess to invading pathogens. Also, the outer regions of the brain, leptomeninges and cortex area are generally white with lower blood flow and iron content compared to the grey matter. Although Beard et al. observed a higher concentration of iron in the white matter, probably because of the preponderant transferrin circulation [
74] yet the availability of this ferrous iron to the invading pathogen might not be corroborated. Furthermore, oxygen supply will be extremely low and stringently controlled in these brain regions because of the lower blood capillary network. Definitely, this self-care brain region might present a first-line defence against invading pathogens, including
C. neoformans.
The phenotypic defects in the
Δsre1 mutants are similar to the observed defects in the
Δcir1, Δcuf1 and
Δcft1 mutants, which further confirmed the interrelationship of the molecular oxygen, iron, and copper to facilitate the virulence traits of
C. neoformans [
43]. In fact, the
Sre1 transcript was downregulated in the
Δcir1 mutant, and the
Δsre1 mutant will not grow under iron-limiting conditions or in the presence of an iron chelator. However, adding ferritin to the iron-chelating media could restore the growth but not as much as the
wt or reconstituted mutant [
60,
69].
Unsaturated acyl-CoA is a component of the membrane bilipid layer. The synthesis of this fatty acid around the endoplasmic reticulum is controlled by the availability of molecular oxygen. There is evidence of upregulation of the Ole1 transcription factor due to hypoxia in
C. neoformans, but this regulation seemed to be Sre1-independent [
68]. The
Ole1 encodes the endoplasmic reticulum membrane-bound Δ
9-fatty acid desaturase in the yeast. This oxygen-dependent enzyme, in the presence of NADH-Cyt b
5 reductase, incorporates a double-bond into the saturated fatty acyl-CoA through the diiron-oxo catalytic centre of the enzyme. Invariably, this enzymatic reaction is orchestrated by the ubiquitin/proteasome-mediated proteolytic activation of Spt23p and Mga2p and subsequent translocation of these dual-activators into the nucleus for the induction of
Ole1 transcript formation [
75,
76]. This activation is influenced by the low level of oxygen, low level of unsaturated fatty acids (or high level of saturated fatty acid), metal chelators, and carbon source. Explicitly, any event leading to iron depletion, such as transition metals like cobalt and nickel that compete with iron for heme, and metal chelators, which withhold iron from binding the heme, will reduce the oxygen-binding capacity of the heme. These events will induce the
Ole1 expression even under normoxic conditions.
C. neoformans can withstand excessive oxygen by expressing two loci,
Oxy1 and
Oxy2, presumably linking hyperoxia-sensitivity to melanin production. The
Δoxy1Δoxy2 mutants are avirulent
albino colonies with defective melanin production [
77] and greater sensitivity to oxygen than
Δoxy1 mutant [
78]. Emery et al. observed a lack of correlation between the
oxy phenotype and melanin production, and it was concluded that the degree of melanin production is a defective predictor of hyperoxia resistance [
78]. The reason is that
oxy2 defective loci in the
Oxy1 genetic background were characterised by defective melanin production and hyperoxia sensitivity, which means that the
Oxy2 locus is more important in conferring hyperoxia resistance. Though the antioxidant effect of melanin in
C. neoformans might not be completely ruled out but Emery et al. predicted that the mechanism of defence against exogenous/endogenous neutral/reactive oxidant associated with hyperoxia stress in fungal aerobiosis solely relies on the interplay of
Oxy1 and
Oxy2 loci in the genetic make of
C. neoformans [
78].
Besides,
Δoxy2 mutants are defective in ferric/cupric reductase and high-affinity copper uptake. Even after reconstitution of the mutant to restore melanin production, most of the mutants retained hyperoxia sensitivity, interpreted as non-allelic, non-interfered, and unlinked loci between the
Mel (for melanin locus) and
Oxy loci (for hyperoxia locus). Nonetheless, the importance of melanin to withstand reactive oxidants cannot be overestimated. The interconnection between the
Oxy and
Mel loci seems elusive. Melanin and polymeric dopamine are effective cellular anionic antioxidants produced by
C. neoformans against hypochlorite (produced by myeloperoxidase as found in inflammatory cells such as neutrophils) [
79], superoxide (O
2‾) and other reactive oxidants. However, activity against neutral oxidants such as H
2O
2 is weak, especially in the presence of copper and oxygen [
80], which by Fenton reaction generate reactive OH
‾ (hydroxyl ion), OH
• (hydroxyl radical), and HOO
• (hydroperoxyl radical) nucleophile that predisposes phagocytosed
C. neoformans to oxidative arrest. Sometimes, perhaps, neutral oxidants can cross the cell wall and bypass the effect of melanin, which is why
C. neoformans are highly susceptible to peroxides such as H
2O
2,
t-BOOH, and COOH (for details on the phenotypic responses of different mutants of
Cryptococcus against various quantitative external factors, check
Supplementary 3). Whatever the case,
Oxy1 and
Oxy2 expression may be needed to resist hyperoxia-based melanin oxidation imposed by the endogenous oxidant from the leucocytes and macrophages.
4.8. Sugar, Lipid, and Proteins (Amino Acids) Metabolisms
C. neoformans possess arrays of metabolic enzymes that help re-wire carbon metabolites to favour the sugar-conjugate products and metabolic intermediates such as acetyl-CoA used for capsule and melanin biosynthesis. In addition, the
in vitro and
in vivo analysis showed numerous membrane transporters specific for each nutrient within their environment.
C. neoformans secretes hydrolytic enzymes such as peptidases, lipases, and glycosidases to release amino acids, fatty acids, and sugars, respectively. The plethora of enzymes involved in these metabolic activities are influenced by cAMP signalling pathways. The cAMP-deficient mutants are poor in sensing glucose (via Hxtp – a sugar/hexose transporter and G-protein coupled receptor, like Gpr1p) and amino acids (via the Gpr4p) because these metabolites provide precursors for growth, mating, cell differentiation, and virulence factors which are constantly under the regulation of cAMP-dependent signalling genes [
86,
87]. The
Δgpr4 mutants are very poor in sensing certain amino acids, such as Met, which triggers receptor internalisation of Gpr4p and subsequent activation of the cAMP/Pka pathway [
18]. However, their response to glucose remained unchanged, unlike the poor glucose response from
Δhxt mutants. Despite a good response to glucose, the
Δgpr4 mutants still displayed impaired capsule formation and mating defect, both of which are recoverable in the presence of exogenous cAMP [
86].
In a DNA microarray analysis, nearly 20% of total gene expression is dedicated to the transport and metabolism of carbohydrates and amino acids metabolism [
88]. The expression of
Gpp2 encoding glycerol-3-phosphate phosphatase controls various metabolic systems as well as osmotic and cold shock responses. Deletion of this gene induced various permeases/transporters (Aapp, Mup1p, Mup3p, Ena1p, Nha1p, Plb1p, Hxt1p, Stl1p, Uga4p), significant facilitator transporter super-family genes, anti-oxidative enzymes (Cat2p, Ccp1p, and Gst1p), and membrane transporters (for sulphite, myo-inositol, and pantothenate). However, various genes involved in the redox process, the glycolytic pathway, and the TCA cycle are downregulated in this mutant [
89]. More important is the induction of genes involved in sulphur-containing amino acid biosynthesis, such as
Cys3,
Sul1,
Soa1,
Bds1, Met17, and
Jlp1 in this mutant. The expression of
Gpp2 modulates the regulation of
S-containing amino acids by facilitating the proteolytic degradation of Cys3p and subsequently reducing the sulphur assimilation [
89,
90].
Hitherto, the deletion of
Aap4 and
Aap5 allowed eugenol resistance in
C. neoformans. However, deletion of
Gpp2 alters the biosynthetic regulation of Met and Cys due to the perpetual activation of amino acid permease protein (Aapp), which probably explains the increased sensitivity of the
Δgpp2 mutants to eugenol. This dire situation could not be rescued by exogenous proline; however, it is uncertain whether glycerol can suppress this defect or not [
89]. Reasonably,
Δhog1 mutants might as well be sensitive to eugenol because Aap is also induced in
hog1 mutants [
91]. Thus, osmotic stress response by glycerol in the yeast is essential, but Cys can feed the glutathione cycle to generate GSH for alternative osmotic stress response in a case where glycerol production is truncated in a mutant such as
Δgpp2. This seemingly inversely proportional metabolic network was investigated further in the
C. neoformans wt to be true because 1 M NaCl can downregulate
Cys3 and
Sul1 expression to some extent [
89].
Furthermore, the activity of the Gpp2p is under the regulation of Ca
2+ via the Cna1/Cnb1 cascade event. Just as in
A. nidulans and
N. crasssa, deletion of
Cys3 in
C. neoformans has been shown to reduce the expression of
Sul1 and both expressions are grossly repressed in the presence of
S-containing amino acids (e.g., Met or Cys) in SD (synthetic dextrose) media more than the YPD [
90]. This shows that media with low
S-containing amino acids will likely induce
de novo sulphur amino acid synthesis. Again, Cys seems better as a sulphur source than Met because methionine could be formed from cysteine in a transulphurylation reaction [
90]. Further investigation on
Δcys3 mutants showed that the mutants displayed a
wt level of urease activity and capsule size yet attenuated for virulence in
Galleria mellonella. Bad still, the experimental condition with Cys or Met in the culture media failed to support the melanin and phospholipase assay of the fungus [
90].
The proteomic analysis that showed a concurrent induction of
Met3, Tef1, Gpd1, Ccp1, Rps0, Rps1, Tif1, Sub2, Tm8, Gpp2, Cna1, and
Cnb1 has led to the proposal of a metabolic relationship between sulphur-containing amino acid biosynthesis, glycerol phosphatase (encoded by
Gpp2), and calcineurin pathway (Cna1/Cnb1), which was further confirmed by protein-protein interactions [
90]. Expression of Cna1/Cnb1 motivates and maintains Cys3p protein levels for nuclear localisation and subsequently increases the Cys3p-target genes such as
Sul1, Met2, Met3, Met10, Str1, and
Sam1 to promote sulphur, purine, Gly, Ser, Thr, Asp, Asn, Cys, and Met metabolism while Gpp2p orchestrates inactivation of Cys3p by proteolytic degradation. A proposed mechanism is that inorganic sulphur induces Cna1/Cnb1 activation to deactivate Gpp2p to promote Cys3p and activate downstream genes involved in S-containing amino acid biosynthesis. However, the presence of organic sulphur such as Met and Cys deactivates Cna1/Cnb1, and this activates Gpp2p to deactivate Cys3p and repress the downstream genes involved in
S-containing amino acid biosynthesis [
90].
Among the genes induced in
Δcys3 mutant are putative
Aro3 and
Aro4 involved in aromatic amino acids biosynthesis;
Jlp1 in sulfonate metabolism to provide SO
42- for Met5/10p; sulphiredoxin in oxidative stress response; and
Clr6 – encoding class I histone deacetylase (a transcriptional epigenetic repressor of
Mat2 – Mat3 region and centromere for precise heterochromatin packaging during chromosomal segregation in fission yeast) [
92], but the repression of
Clr6 in the
C. neoformans wt promotes capsule formation and biosynthesis of various amino acids. With Maier et al.,
- ❖
Clr6, along with Hog1, Mbs1, Ste12α, and Tup1, is classified as a cluster of genes induced against reactive oxidants but repressor of respiratory activities and also a repressor of sugar, amino acid, and ionic transporters that are needed for growth,
- ❖
the cluster of Ada2, Clr3, Clr4, Clr5, Ecm2201, Fkh101, Gat201, Hap3, and Rim101 induced ionic transporters but repressed chitin synthesis,
- ❖
the cluster of Cac1, Cep3, Cir1, Clr2, Fap1, Fhl1, Fkh2, Gat1, Mcm1, Pdr802, Sp1, Swi6, and Usv101 are involved in the mitochondrial respiratory process and activation of proteins/enzymes involved in gluconeogenesis and ionic transporters for osmoregulation; however, cAMP signalling is repressed, and
- ❖
the cluster of
Clr1, Hap5, Nrg1, Pkr1, and
Ssn801 enhanced ribosomal biogenesis, response to oxidative stress, amino acid biosynthesis, cAMP signalling but repressed respiratory activities, ionic transporters, gluconeogenesis, and osmoregulatory process [
93].
Transcriptome analysis
Δpka1 and
Δpkr1 mutants have been characterised with higher-level transcripts of genes involved in glycolysis/carbohydrate metabolism, TCA, amino acid metabolism, iron uptake, cell wall glycan formation, translational ribosomal proteins, translational initiation and proteins elongation, heat shock, protein and vesicular trafficking, UPR, cytoskeletal proteins, membrane transporters/permeases, cell surface and extracellular transmembrane proteins, oxidative and nitrosative stress response proteins, mitochondria membrane proteins, virulence factors, and phospholipid metabolism [
94].
In addition, there is a characteristic increase in transporters/permeases such as sugar transporter, hexose transporter, glucose transporter, and maltose permease (all for sugar assimilation); ammonium permease and amino acid transporter (to assimilate nitrogen and amino acids, respectively); peroxisomal fatty acid transporter and ATP carrier (to assimilate phospholipase products), and many other residential permeases and ion-affinity coupled transporters, which are used for importing nutrients and micro-needed ions from the extracellular digestion. Therefore, through the interplay of Pka1 and Pkr1 expressions, cAMP plays one of the major central roles in coupling environmental sensing to metabolism, virulence, cell differentiation, mating, and growth.
C. neoformans is sensitive to glucose limitation, and phagolysosome does this by constantly maintaining acidic pH to minimise glucose availability to pathogens. To circumvent this,
C. neoformans embarked on gluconeogenesis to supply sugars for metabolisms leading to titanisation and capsule formation. The interdependence of peroxisome, glyoxysome and cytoplasm to break down assimilated fatty acids (
β-oxidation) to generate acetyl-CoA is essential to drive the glyoxylate cycle to generate oxaloacetate used in gluconeogenesis. Contrary to
C. albicans, which showed up-regulated genes for key enzymes of gluconeogenesis [
95], phagocytosed
C. neoformans showed no upregulation of such enzymes [
16]. However, transcriptional response and analysis of
C. neoformans ingested by macrophage and amoeba showed that key enzymes like phosphoenolpyruvate carboxykinase and fructose-1,6-bisphosphatase were respectively upregulated [
13]. The
Icl1 (encoding isocitrate lyase),
Mls1 (encoding malate synthase in glyoxylate cycle), and genes for
β-oxidation enzymes are upregulated in
C. neoformans recovered from murine macrophage [
16]; however,
Mdh2 (encoding malate dehydrogenase) and genes for ergosterol biosynthesis were downregulated in the phagocytosed
C. neoformans [
13]. The findings of Rude et al. showed that though the Icl1p and the glyoxylate shunt are important for ATP production during fungal infection, still
Δicl1 mutants failed to display any apparent virulent defect in animal studies and produced a
wt level of capsule and melanin without any defect in macrophage survival or phagocytic index [
96].
In all these reactions, reducing equivalents such as NADPH
+ are highly required for redox processes. Hypervirulence-associated protein 1 (encoded by
Hva1) has been identified to be associated with NADPH to regulate
C. neoformans metabolisms [
97]. Derengowski et al. have discovered a multiple-fold increase in the transcript level of
Hva1 in amoeba-phagocytosed cryptococcal cells [
13] and the absence of this Hva1p unexpectedly increased the level of phosphoenolpyruvate kinase but definitely impaired the TCA due to reduced activity of 2-ketoglutarate dehydrogenase complex activity. This condition probably attenuates mitochondria oxidative reaction but favours sudden energy increased from preponderant floating of cytoplasmic NADPH via alternative pathways to produce ATP. This metabolic energy production shift favours cell proliferation and growth, hence the hypervirulence trait of
Δhva1 mutants in the murine model but not moths or worms [
97]. This means the
Δhva1 virulence is temperature dependent. The exogenous supply of NADPH
+ further enhanced the higher tissue burden of this mutant in the spleen and brain than in the lungs [
97]. Paradoxically, an extensive comparison of the virulence factor expression, capsule size and structure, melanin production, GXM content, phospholipase and urease activity, growth and doubling time under stressful conditions, survival and fungal burden in macrophage, animal model, moths, and worms revealed no significant difference between
Δhva1 and
Δhva1+Hva1 mutants [
97]. This displayed the importance of regulated metabolic systems in
C. neoformans virulence.
To further appreciate the involvement of metabolic systems and their regulatory enzymes in the virulence, cell wall integrity, and stress-resistance phenotypic traits in
C. neoformans, the lack of
Pgi1 expression (encoding phosphoglucose isomerase/glucose-6-phosphate isomerase), though showed no observable difference in the cell morphology but resulted in reduced capsule biosynthesis, impaired cell wall integrity, fragile cell membrane, osmotic stress hypersensitivity (due to impaired Hog1 pathway), and failure to utilise mannose and fructose [
98]. Insertional mutation T-DNA at the promoter site of
Pgi1 reduces the activity of Pgip, and in the presence of a 2% glucose supplement, this mutant tagged as LZM19 (
Δpgi1) is able to produce melanin due to the derepression of
Lac1. Further work showed that an exogenous supply of cAMP could restore capsule deficiency in the LZM19 strain, but this repressed the
Lac1 expression [
98].
The mammalian homologue of AMPK protein is Snf1p in the yeast, which is critical to cellular energy homeostasis through oxidative phosphorylation by facilitating sugar and fatty acid uptake. In a normal state, Snf1 is turned off by Reg1-Glc7 protein phosphatase 1 (PP1) [
99]. However, high glucose concentration activates Cac1p to produce cAMP, which interacts to activate Pka to repress the H
2C
2 zinc-finger nuclear-translocated proteins Msn2p and Msn4p that are programmed to induce genes containing SRE-sequence promoters such as
Hxt1, Hog1, Ras1 and others, which are generally regarded as NSR, OSR, OMSR, ESR, GSR, HMSR, and CSR. In addition, the cAMP/Pka also inhibits the redundant upstream Snf1p-activation kinases, Sak1, Tos3, and Elm1, to put Snf1 under repression in high glucose content.
High glucose deactivating Snf1 leads to the activation of Mig1p by dephosphorylation to enhance the association of Mig1p with Suc2p. The Mig1p is a universal gene repressor in the yeasts [
100], which binds the promoters of target genes, including the
Lac1 gene, to repress melanin biosynthesis as a response to glucose repression. In low glucose levels, however, the Cac1 activity is reduced to lower the cAMP level, which invariably elevates the Sak1 activity to activate Snf1p [
101]. The activated Snf1p further inhibits the Cac1 to promote nuclear translocation of Msn2p to induce
hxt7 and other similar stress response genes, including the
Lac1 gene.
Canonically, despite glucose availability, mutation of
Pgi1 expression may have limited glucose catabolism and subsequently altered the ATP homeostasis. This condition favours Snf1p activation leading to the deactivation of Mig1p and subsequently derepresses
Lac1 for melanin production. Therefore, the involvement of the Snf1-Mig1 regulatory pathway through nutrient homeostasis cannot be ruled out in the pathogenesis and virulence of
C. neoformans. Unprecedentedly, the involvement of
Snf1 expression in antifungal, oxidative, and osmotic stress responses is less significant. Though Hu et al. claimed to have observed increasing sensitivity of
Δsnf1 mutants to AmpB relative to the
wt response; however, the
Δsnf1 mutants apparently showed no significant growth defect to AmpB and FCZ, but a very slight growth defect in the presence of RPM [
102].
Generally, phenotypic defects of
Δsnf1 mutants become conspicuous with increasing temperature. For example, melanin formation decreases above 30
oC and ultimately becomes a non-melanised mutant at 37
oC [
102]. Surprisingly, though the transcript level of
Snf1 is practically increased in a glucose-limited or acetate medium, but there is no evidence of increased transcript in the cryptococcal cells recovered from the infected lungs. Besides,
Δsnf1 mutants in a glucose-limited medium displayed an increased marginal level of
Ady2, Acs1, Ato2, and
Pck1 transcripts; a marginal decrease in the expression levels of
Ctr4 and
Cgl1; and a significant reduction in the
Hxt1 and
Cft1 transcripts [
102]. This means the latter two transcription factors solely depend on the
Snf1 expression, especially in the glucose-limited medium. In the acetate medium, however, deletion of
Snf1 showed no effect on the expression of all these genes except the
Hxt1, where an elevated transcript is observed. The additional report showed that
Δsnf1 mutants displayed a
wt Mls1 transcript irrespective of the media used [
102]. The activation of Hxt1p in
Δsnf1 mutants cultivated in the acetate medium showed the characteristic of glucose-deprived cryptococcal cells and readiness to assimilate sugars as soon as it is available.
In addition to
Lac1 regulation,
Snf1 expression also affects the SOD activity. A drastic reduction of
Lac1 in
Δsnf1 mutants cultivated in the glucose-limited medium indicates the
Lac1 expression being regulated by Snf1p. Similarly, the
Sod1 transcript is reduced in
Δsnf1 mutants in a glucose-limited medium but not in the acetate medium. Irrespective of the medium, deletion of
Snf1 failed to significantly affect the expression of
Fhb1, Tps1, Glr1, and
Skn7, but
Lac2 and
Ssa1 transcripts are significantly increased in acetate media [
102]. This means that in the glucose-limiting medium, Snf1p orchestrates the expression of
Lac1 to produce melanin and the expression of
Sod1 against oxidative stress. Strategically,
Lac2 expression may be preferable for melanin synthesis under alternative carbon sources coupled with oxidative and nitrosative stress, which might have been addressed by the concomitant expression of
Tsa1 under the same nutrient-limiting medium; after all,
Δlac1 mutants still cause infection in a murine model [
102,
103]. Unfortunately,
Δlac1 and
Δlac1Δlac2 mutants are totally
albino. In contrast,
Δlac2 mutants are melanised in catecholamine-enriched media [
104]. However, the elevated
Lac2 transcripts more than the
Lac1 in the presence of alternative carbon sources such as acetate [
102] is reminiscent that
Lac2 may probably be involved in melanin production when unusual carbon sources weaken
Lac1 involvement. Therefore, by inference, catecholamine (such as
L-DOPA and norepinephrine) metabolism into eumelanin may seem more complex than already revealed, especially in the presence of alternative carbon sources.
The role of
Acs1 expression encoding acetyl-CoA synthetase has also been investigated in the cryptococcal cells recovered from the lungs of infected animals. Evidence showed an elevated expression of the
Acs1 gene in
C. neoformans during infection, coupled with concurrent elevated expression of different membrane transporters and stress response genes, which is indicative of a nutrient-limited environment and the route for alternative carbon sources as facilitated by the expression of
Snf1 [
102]. Deleting
Acs1 yielded moderately attenuated mutants that cannot utilise alternative carbon sources such as acetate.
Sometimes
C. neoformans can utilise other carbon sources, including alcohol. Inexpediently, mutants lacking
Hap3 and
Hap5 showed weaker growth in sucrose, acetate, and ethanol and are more sensitive to 0.01% SDS. In contrast, the
ΔhapX mutant and the
wt showed similar growth in the glucose, sucrose, acetate, and ethanol, even in 0.01% SDS. Although
Hap genes are iron-dependent regulatory genes, surprisingly, supplementing the media with hemin failed to improve the weak growth in
hap mutants [
40]. Because of this, iron-enrich media specifically improved gene expression encoding several iron-dependent enzymes in the glycolytic pathway, TCA cycle, functional oxidative pathways, and mitochondrial redox processes, but gene expression targeting the pentose phosphate pathway was scarcely tagged in the SAGE analysis [
46]. Apart from these elevated key enzymes, other accessory proteins such as ubiquinone, prohibitin (a membrane-bound chaperone stabilising the mitochondria proteins), and ATPase are equally pre-eminent [
46]. The expression of TCA-targeting enzymes such as aconitase and succinate dehydrogenase further confirmed the importance of the heme and Fe-S redox proteins to the upkeep of
C. neoformans metabolic systems. In fact, aconitase identified in
C. neoformans seems to have paralogous encoding genes to probably enhance its extracurricular activities such as
mRNA translation/stability – a function known for Irp1p. It could probably be that
C. neoformans regulates excessive iron by converting this Irp1p to cytosolic aconitase, which fails to stabilise
mRNA but promotes Irp2 ubiquitination and proteasomal degradation [
105].
Further, into the regulatory roles of Hap3p and HapXp, these two transcription factors negatively modulate the expression of the orthologue
Lys4 in the LIM. This was confirmed by low detection of
Lys4 in
ΔhapX and
Δhap3 mutants cultured in the presence of 100 μM FeCl
3, but in the absence of FeCl
3, the expression of
Lys4 is well induced in either of these mutants [
106]. Relative to the
wt, the expression of
Lys4 remained unchanged in
Δcir1 mutants in either iron enrich media (IEM) or LIM. Like
Lys4,
Leu1 expression was also similarly regulated by Hap3p and HapXp, either in the LIM or IEM. Low iron supports the expression of
Leu1 and
Lys4 in
Δhap3 and
ΔhapX mutants, but only a high level of iron favours the expression of
Leu1 in
Δcir1 mutant [
107]. This means that Cir1p positively regulates this cytoplasmic located Leu1p in
Δhap3 and
ΔhapX mutants cultured in LIM but regulates negatively in IEM. There is speculation, however, that
Leu1 expression may be more under the regulation of Hap3p than HapXp expression because while
ΔhapX mutant produced a comparable
Leu1 transcript with the
wt,
Δhap3 mutant produced a significantly more transcript of
Leu1 under the same IEM conditions [
107]. Like
Lys4 expression,
Leu1, with a putative ―CCAAT―Hap binding element at the promoter site for repression during the iron-limiting condition, is also very important in Leu biosynthesis; however, the uptake of these amino acids is generally under the influence of NCR.
Being a eukaryote,
C. neoformans compartmentalise and adopt differential regulation of transcription factors to facilitate catabolism (in the mitochondria, glyoxysome, and peroxisome) or anabolism (in the cytoplasm) depending on the inducing factor and the availability of specific nutrients. For example, IEM elevates enzymes (including fatty acid synthase and acetyl-CoA carboxylase) in the cytoplasmic lipogenesis and promotes extracellular digestion of lipids (lipolysis) which is usually commenced with the release of phospholipase B (encoded by
Plb1). Analysis showed that
C. neoformans possessed genes for lipase (encoded by
Tgl2 among at least three encoding genes) and enoyl-CoA hydratase [
46,
108]. However, there is no specific evidence of the functional extracellular lipase expressed in this fungus due to the lack of a signal peptide in these proteins, unlike the lipase secreted by
C. albicans [
108,
109].
The lack of trypsin, chymotrypsin,
β-galactosidase,
β-glucuronidase, and α-mannosidase by the species of this fungus showed that certain proteins and
β-
D-sugars might not be suitable in the culture media for
C. neoformans [
109]. However, many species of cryptococcal cells do possess esterase lipase (active against triacylglycerol with C4 – C8 acyl chain), α-galactosidase, α-glucosidase, β-glucosidase, N-acetyl-β-glucosaminidase, α-fucosidase, galactokinase, D-lactate dehydrogenase [
46,
109], which means that various α-D-sugars, amino sugars, and disaccharides are suitable carbon sources for cryptococcal cells. Furthermore, since IEM elevates sterol biosynthesis, it can be speculated that fatty acid/sterol-formulated media could motivate sterol and phospholipid biosynthesis because of the possibility of expressing lipolytic and sterol biosynthetic enzymes under the iron influence.
In addition, lipid biosynthesis is an essential biochemical pathway leading to cell membrane integrity. The Mga2 is an orthologue transcription activating factor for the component of fatty acid synthesis and regulates
in vitro normal growth at a range of temperatures. Therefore, mutation of
Mga2 makes
C. neoformans thermosensitive and hypersensitive to FCZ with a concomitant reduction in the expression of Fas1, Acc1, Rpl11, Rps7, Leu1, Pho, Pdh, eEF3, and other orthologue proteins. However, expression of Lys2, Hxt, Cat, Pak1, Chs, Mdr1, ABC-family, Vps, Hsp78, Hsp104, calnexin, Rho-GAP domain protein, exo-1,3-
β-glucanase, and other relevant proteins are induced in this mutant [
110]. Ambiguously, it is unclear how the presence of Mga2 regulates polarisation and filamentation growth in
C. neoformans from the work of Kraus et al. because
Δmga2 mutants are characterised with elevated expression of the homologues of small GTPase effectors, which control the polarisation growth, and the Pak1, which controls the filamentation growth. Therefore, it is possible that during this morphogenesis, Mga2 expression is repressed at ≤30
oC for GTPases and Pak1 expressions that will promote polarisation during growth, but this, however, will contradict the involvement of Mga2 in thermotolerance and morphogenesis as reported by Kraus et al. [
110]. Besides, the dependency of
fas1 and
acc1 expressions on Mga1 for lipid and sterol biosynthesis will literally promote polarisation growth, mating, and haploid fruiting. Therefore, the background expression of Mga1 may have promoted polarisation and cell growth in
Δmga2 mutant.
Still, under the influence of iron, secretory vesicles (such as exocyst complex encoded by
Sec gene), enzymes (such as glutamate dehydrogenase,
p-aminobutyrate aminotransferase, proteasomes, ubiquitin ligase, RNA-processing enzymes, and
mRNA splicing proteins), and transporters/permeases (for allantoate, urea, ammonia, and purine-cytosine) involving in nitrogen and amino acid metabolism including protein trafficking have been observed to be elevated under the IEM [
46]. Not only these, iron influences several of the amino acid metabolism (anabolism and catabolism). In
S. cerevisiae, iron probably orchestrates the binding of α-isopropylmalate to homodimeric DNA-binding protein Leu3 (a repressor that, together with Gcn4, regulates amino acid biosynthesis) to regulate branched-chain amino acid biosynthesis, Aapp, nitrogen and carbohydrate metabolisms. The expression of Lys4p for Lys biosynthesis and Leu1p for Leu biosynthesis is iron-dependent because of the presence of conserved aconitase repeating motifs that bind Fe-S; and probably other amino acids biosynthesis proteins such as Met2p, Met3p, and Met6p for Met; Ilv2p for Ile and Val. Mutants generated from each of these transcription factors are avirulent in the MIMC.
Iron homeostasis in
C. neoformans consists of arrays of gene clusters harmonising the acquisition and utilisation of iron, and this could be similar to the iron regulon of
S. cerevisiae under the influence of Aft1 and Aft2 transcription factors. With NCR, certain nitrogen sources are not assimilated, perhaps for the reason of energy-costly catabolism. Deletion of
Gat1, among other GATA transcription factors, promotes poor to no growth of
C. neoformans in the majority of different nitrogen sources except in the presence of Pro and, to some extent, Arg, while single deletion of other GATA family genes (
Δgat201, Δgat204, Δbwc2 or
Δcir1) failed to shut down totally the fungal growth in all the popular nitrogen sources [
111]. This means that various amino acid permeases may still be activated to assimilate other nitrogen sources under later conditions, contrary to
Δgat1 mutants where the
Amt and majority of the
Aap gene expressions are impaired. However, Pro and Arg may probably not be under NCR control.
To affirm this, the Pro medium induces
Put1 gene expression for proline oxidase in the
wt under the influence of NCR as mediated by
Gat1 expression in the presence of NH
4+. In addition, the expression of
Put5 (a paralogue of
Put1) and
Put2 genes, which are involved in the downstream catabolism of Pro to Glu, remained unchanged in
Δgat1 mutants and
wt [
111]. This means that though Gat1 mediates NCR, Pro metabolism is Gat1-independent. The ability to form Glu from Arg may also explain the moderate growth of
Δgat1 mutants in this medium. Gln and His can also form Glu directly. However, because the cytoplasmic content of Gln is always high as a form of nitrogen donor in most metabolic reactions and the metabolism of His is also more complex, this may explain why
Δgat1 mutants had poor growth in a medium with each of these two amino acids as nitrogen sources.
The growth defect of
Δgat1 strain H99 mutant in urea, urate, or creatinine-containing media with reduced expression of genes involved in nitrogen metabolism (such as
Gdh1, Amt1, and
Amt2) is a serendipitous observation that contradicts the ecological niche of
C. neoformans. Similar growth survival observed with the
wt and
Δgat1 mutants cultivated in pigeon guano media is an indication that though this ecological niche is rich in urea, urate, and creatine, which may probably be selective by the NCR but at the same time, the survival of the
Δgat1 mutant in the same environment poses an investigation into the unique nitrogen sources suitable for such a mutant. Perhaps, such media may be rich in Pro or Arg as inferred from Lee et al. observation [
111]. Notwithstanding, limited nitrogen sources in the environmental pigeon guano may have repressed the
Gat1 for the mating process that promotes propagule/basidiospore formation of the filamentous stage of this opportunistic human infectious fungus. Lee et al. had discovered that mating is greatly enhanced through the robust filament and basidia formation either in the unilateral or bilateral crossing of
Δgat1 mutants [
111]. In the same way, deletion of
Gat1 drastically reduces capsule formation, enhances melanin formation at physiological temperature, and promotes extreme thermotolerance but shows no significant effect on urease activity. Unexpectedly, these conditions seem to moderately reduce the pathogenic efficiency of
Δgat1 mutants compared to the
wt [
111]. This shows that the key virulence factor of
C. neoformans is firmly under the influence of NCR.
Keeping NCR under repression by Gat1p is necessary to facilitate the utilisation of other nitrogen sources. Lee et al. discovered that capsule formation increased considerably in the strain H99 when cultivated in a medium containing Asn, urea, urate, or creatinine but not Gln, Pro or Ala-containing medium. This capsule formation is prosaically affected by the presence of NH
4+ in each of the media [
111]. This shows that NH
4+ represses capsule formation by activating NCR even in capsule-induced nitrogen sources, but, on the contrary, the presence of NH
4+ induces melanin formation. To buttress this, Lee et al. observed melanin formation in each of the aforementioned media when prepared along with NH
4+ except in the Pro medium, where a slight melanisation was observed even when NH
4+ is present [
111]. Without NH
4+, creatinine and urate failed to support melanin formation, and only minimal melanisation was observed in the Ala medium. However, melanin was significantly induced in urea, Asn, or Gln-containing media [
111]. This shows that NCR differentially regulates virulence accessory factors and that certain amino acids can override the NCR effect in the dire need for virulence. So, while NH
4+ represses capsule formation in the presence of different nitrogen sources, melanin is significantly induced.
It is shown that among the numerous amino acid permeases,
Aap2, Aap4, Aap5, Aap6, and
Aap8 are more differentially expressed in the amino acid fortified SD media compared to the rich medium (YPD) [
112,
113], which means that NCR is highly induced in YPD compared to the SD. Without NH
4+ (a preferred nitrogen source), global amino acid control is triggered by nitrogen starvation leading to preponderant production of
t-RNA for amino acid biosynthesis in the yeast. For example, nitrogen starvation induces Trp biosynthetic enzymes encoded by
Trp2, Trp3, Trp4, and
Trp5 [
112]; also, under repressive NCR,
Aap2 and
Aap3 are induced in the presence of Lys-containing medium [
106]. However, amino acid supplements in SD do repress the general amino acid control (GAAC) but promote the APC-superfamily membrane transporters/permeases (APC, AAAP, AGCS, CCC, HAAAP, BCCT, SSS, NSS, NCS1, NCS2, and SulP families [
114,
115,
116]).
Trp supplement alone in the SD could increase
Aap1 and
Aap2 transcripts by 3.5 and 3.8-fold; similarly, adding Trp+Met+His could improve
Aap4, Aap2, Aap8, and
Aap5 transcripts by 3.3, 4.2, 7.5, and 21-fold respectively [
112]. The transcription analysis further showed that NCR could repress
Aap2, Aap5, and
Aap8 in the presence of NH
4+ but not
Aap4, which means
Aap4 is sensitive to any nitrogen sources; however,
Aap1 and
Aap7 remained minimally expressed [
112]. Though the NCR selectively represses Aap, carbon catabolite repression (CCR) appears to repress all the
Aap expression in the presence of glucose (preferred carbon source) compared to galactose (alternative carbon source) [
112].
The
Δaap1Δaap2 mutants are thermosensitive (at 37
oC in synthetic media fortified with amino acids), hypocapsulated, and hypovirulent in the
G. mellonella infection model [
117]. Mutants such as
Δaap1, Δaap2, Δaap6, Δaap8, and Δaap1Δaap2 failed to display any significant difference in growth when compared to the
wt at 30 or 37
oC in SD fortified with amino acids/NH
4+ or in YPD; however, growth reduction was observed in liquid culture of
Δaap1 containing Met or Pro at 30 or 37
oC [
117]. Also, a liquid culture of
Δaap1Δaap2 mutant containing Gln or Arg displayed growth reduction at 30
oC, and
Δaap8 mutant displayed growth reduction in media containing Met, Glu, and Trp.
Generally,
Δaap1Δaap2 mutants appeared to show defective growth in 60% of all culturable amino acid media (AAM), which means that these amino acid permeases are essential for growth at 37
oC on amino acids media, just like Aap4 and Aap5. However,
Aap1, Aap2, Aap6, and Aap8 appeared to be non-essential for virulence because mutants of these genes, including
Δaap1Δaap2 mutant, failed to show any significant difference in the mating, filamentation, melanin expression, PLB and urease activities when compared to the
wt [
117]. Apart from being non-essential for virulence, none of the mutants seemed to be affected by oxidative stress agents, alkaline conditions, osmotic and salt stress agents, and cell wall stress agents at 30 or 37
oC. Surprisingly, unlike the single mutants
(Δaap1, Δaap2, Δaap4, Δaap5, Δmup1, and
Δmup3), the double mutants
Δaap1Δaap2, Δaap4Δaap5, and
Δmup1Δmup3 displayed reduced capsule size at 37
oC when compared to the
wt. However, the capsules are similar at 30
oC, and the virulence of
Δaap1Δaap2 and
Δaap8 mutants only is attenuated in
G. mellonella [
117].
From Martho et al.,
Δaap2, Δaap4, Δaap5, Δmup1, Δmup3, and
Δmup1Δmup3 showed no sign of attenuated virulence in
G. mellonella, but
Δaap4Δaap5 mutants are hypovirulent at 30 or 37
oC [
118]. Furthermore, virulence investigation in the murine model showed that mice inoculated with the
wt and
Δaap4 and
Δaap5 died within a month, while the mice inoculated with
Δaap4Δaap5 mutants and PBS-treated strain survived with viable recovery colonies from their lungs and livers. In addition,
Δaap2,
Δaap4, Δaap5, and
Δaap4Δaap5 mutants displayed
wt phenotypic sensitivity to AmpB but are hypersensitive to FCZ, and each of the mutants is hyper-resistant to eugenol due to Aap inactivation. This invariably shows that Aap4 and Aap5 are important permeases for thermotolerance, antifungal resistance, and oxidative stress response for tissue invasion, survival, and virulence in the animal model. By speculation, they are a potential target for therapeutic application. Notwithstanding, the compensatory essence of permeases/transporters cannot be ruled out. By harmonising several works on amino acid permeases – Aap, Mup, and Dao:
- ❖
all mutants of aap and mup are not affected in rich media at high temperatures except Δaap4Δaap5,
- ❖
are not affected in AAM at 37oC and in capsule formation except Δaap1Δaap2, Δaap4Δaap5, and Δmup1Δmup3,
- ❖
are not sensitive to any stress agent except Δaap4Δaap5,
- ❖
are virulent in G. mellonella except for Δaap1Δaap2 and Δaap8 (hypovirulent), Δaap4Δaap5 (avirulent as well as in murine model),
- ❖
the Aap1, Aap2, Aap3, Aap4, and Aap5 are very similar in the sequence homology, and each of the Aaps in the double mutants appeared redundant in their functions,
- ❖
the Aap3 and Aap7 expressions are below detection in YPD or SD; however, Lys-containing medium induced Aap2 and Aap3 expressions under alleviated NCR,
- ❖
irrespective of the nitrogen sources, the expression of Aap6 remained relatively unchanged,
- ❖
the Aap8 responds to amino acid-supplemented media only, but the highest expression is usually found in Aap2, Aap4, and Aap5,
- ❖
Mup1 and Mup3 expressions are under the NCR regulation as Aap2 and Aap5 but can be induced by His, Trp, and Met when NRC is shut down,
- ❖
only Mup1 can be induced by S-containing amino acids,
- ❖
galactose induces the expression of all the Aap and Mup genes,
- ❖
expression of Aap6 and Aap8 is temperature-independent,
- ❖
Aap4, Aap5, and Mup1 induction increased from 30 to 37oC in SD medium,
- ❖
Aap2, Aap4, Aap5, and Mup3 expressions are repressed further from 30 to 37oC except for Mup1, which is induced further,
- ❖
irrespective of the growing media, there is no significant change in the growth of Δaap2, Δaap4, Δaap5, Δmup1, Δmup3 and Δmup1Δmup3 mutants at 30 or 37oC,
- ❖
the significant growth defect of Δaap4Δaap5 mutants at especially 37oC in YPD or SD showed that the two permeases (or at least one of them) are important for thermotolerance,
- ❖
the use of amino acids as nitrogen sources impaired the growth in Δaap4Δaap5 mutants,
- ❖
relative to NH4+ at 30oC, Val, Ile, and Met-containing SD media poorly support the growth of C. neoformans, but Leu, Ser, Lys, and Phe are better nitrogen sources,
- ❖
Gly, Asp, Asn, Glu, Gln, Arg, Trp, and Pro are highly competitive with NH4+ in culturable AAM,
- ❖
at 37oC, Val and Met poorly support the growth of C. neoformans in SD media, but Gly, Leu, Ile, Ser, Trp, and Phe are good nitrogen sources, while Asp, Asn, Glu, Gln, Arg, Lys, and Pro are better nitrogen sources,
- ❖
stereospecifically, var. gattii metabolises D-amino acids because of the more active expression of Dao1, Dao2, and Dao3 genes (encoding D-amino acid oxidase) but less metabolisable for var. neoformans, which prefers L-amino acids as nitrogen sources due to the inefficient evolutionary expression of Dao gene,
- ❖
in all, growth is denser in L-amino acids containing media compared to corresponding D-amino acids media,
- ❖
pathologically, Δdao mutants of C. neoformans are virulent, but Δdao mutants of C. gattii are attenuated,
- ❖
positive correlation of Gat1 expression that represses NCR is confirmed to aid the expression of Aap when the preferred nitrogen source is absent/limited but uncertain with Dao expression,
- ❖
L-Tyr failed to dissolve at permissive pH for C. neoformans to grow,
- ❖
Aap2, Aap3, Mup1, and Mup3 may be global amino acid permeases/transporters because of the most significant growth defect of their corresponding mutants, especially in their double mutant states, from 30 to 37oC and their ability to be induced by various amino acids,
- ❖
in addition to being a global permease and redundant transcription factors, Aap4 and Aap5 promote thermotolerance and response to oxidative stress, and the growth of the double mutant is significantly impacted from 30 to 37oC in a single amino acid medium or in the presence of ≥5 mM H2O2,
- ❖
unlike the Δaap4 and Δaap5 mutants, the growth defect of Δaap4Δaap5 mutants at 37oC appeared to be restored as pH increased gradually into the alkaline state or when supplemented with 0.75 M NaCl (this condition generates H+ via Na+/H+ antiporter that drives other amino acid permeases to compensate for the deletion of aap4 and aap5),
- ❖
by functions, Aap2, Aap3, Mup1, and Mup5 are regarded as minor permeases while Aap4 and Aap5 are major permeases [
117,
118,
119].
It is very surprising, therefore, to have observed the relevance of the Ras1 pathway in the control of amino acid permeases leading to thermotolerance. The
Δras1 mutants had shown growth reduction compared to the
wt in various amino acid fortified cultures at 30 or 37
oC except for media with Ala, Cys, His, Tyr, and Thr that failed to support the growth of
Δras1 mutants [
117]. The involvement of Ras1 in the assimilation of amino acids is a unique feature of this transcription factor apart from the popular thermotolerance that it is known for. This means that amino acid uptake may be part of the function of Ras1 to facilitate thermotolerance. Alternative carbon and nitrogen sources such as galactose and Pro may also induce
aap expression. It is not surprising, thus, to observe growth reduction in
Δras1 mutants cultured in either SD or synthetic galactose (SG) fortified with NH
4+ or Pro at either 30 or 37
oC [
117]. This shows the importance of Ras1p in the nutritional balance of
C. neoformans against thermal stress. The controlling effect of Ras1p on the expression of
Aap was further investigated by quantifying the expression level of each
Aap transcript in
Δras1 mutants. All
Aap and
Mup transcripts were repressed in
Δras1 mutants in SD fortified with Trp, His, and Met at 30 and 37
oC except
Aap5, which showed a
wt transcript level at 37
oC [
117]. The exception of
Aap5 may suggest a different regulatory mechanism in addition to or different from Ras1 expression.