Paracoccidioides brasiliensis AND Paracoccidioides lutzii, A SECRET LOVE AFFAIR

SUMMARY

To commemorate Prof. Carlos da Silva Lacaz's centennial anniversary, the authors have written a brief account of a few, out of hundreds, biological, ecological, molecular and phylogenetic studies that led to the arrival of Paracoccidioides lutzii, hidden for more than a century within Paracoccidioides brasiliensis. Lacaz's permanent interest in this fungus, and particularly his conviction on the benefits that research on paracoccidioidomycosis would bring to patients, were pivotal in the development of the field.

INTRODUCTION

In 2006, a seminal paper on the genomic variability of Paracoccidioides brasiliensis was published31 which brought a totally different perspective on the study of this fungal species. By then, Carlos da Silva Lacaz, that iconic figure in Latin American mycological circles, had been dead for four years, so that he could not greet the arrival of Paracoccidioides lutzii as a cryptic species within the genus Paracoccidioides 13 ^, 15 ^, 50 ^, 51, hidden as it was for a century under the guise of P. brasiliensis, ever since Adolpho Lutz described the fungus and its dimorphic nature for the first time in 190828.

As usual in science, P. lutzii arrived as the outcome of decades of experiments on biochemistry, ultrastructure and molecular biology in P. brasiliensis, a very brief account of which we intend to summarize herein.

... AND THE WALLS CAME TUMBLING DOWN

The changing morphology of P. brasiliensis and its direct relationship to pathogenicity (i.e., yeast-like phase as the morphological phase in lesions; mycelial phase as a saprophyte) inspired researchers since the early 1970s to analyze fungal structures directly responsible for the final shape of the cell. In so doing, the San-Blas's group at the Venezuelan Institute for Scientific Research (IVIC) devoted those years to break down yeast and mycelial walls as the most obvious shape-defining structures, in order to analyze their chemical composition, studies already initiated by KANETSUNA et al.23 as a follow-up of CARBONELL's12 ultrastructural analyses.

In fact, the possible role of a-1,3-glucan as a dimorphic determinant24 and virulence factor46 was first reported in this fungus. This polysaccharide is present in the P. brasiliensis yeastlike cell wall but disappears when the fungus changes to its mycelial phase, as it is substituted by a b-1,3-glucan which is almost exclusively present in this morphotype45. As in all pathogenic fungi, chitin was also a major structural component of the cell wall and therefore, subjected to analysis on its role on morphology and pathogenicity42 ^, 43.

Chitin, the b-1,4-linked homopolymer of N-acetylglucosamine, is one of the major components of the fungal cell wall, with important functions in wall integrity as structural component, and involved in morphogenesis and conidiophore development2. In earlier times, its synthesis was proposed as a potential antifungal target by researchers in the field. However, the complexities of its synthesis, where multiple chitin synthases participate at different stages of growth, have cooled down the initial enthusiasm for the search of chitin-inhibitory antibiotics. In Paracoccidioides spp., seven chitin synthases, one for each of the fungal chitin synthase classes15 ^, 33 ^, 34 ^, 57, have been reported.

b-1,3-glucan is also a structural component of the fungal cell walls, but unlike chitin, b-1,3-glucan synthesis appears less complex. In P. brasiliensis and P. lutzii, a single b-1,3-glucan synthase gene has been reported, which is overexpressed in the yeast phase of both species48 ^, 58, a result that contradicts the fact that b-1,3-glucan content is higher in the cell wall of the mycelial phase. However, they agree with earlier in vitro biochemical data, that indicated a higher activity of P. brasiliensis b-1,3-glucan synthase in particulate preparations of yeast-like cells than those of mycelial cultures49. It might indicate a post-transcriptional regulation of the b-1,3-glucan synthesis in Paracoccidioides spp., perhaps requiring the participation of a regulatory subunit of the b-1,3-glucan synthase complex, as already reported in other fungi48. The partial inhibition of yeast growth in P. brasiliensis by echinocandin inhibitors (which target b-1,3-glucan synthases) and their higher growth inhibition in its mycelial phase, are in agreement with the relative content of this polysaccharide in each morphological phase40 ^, 43; it also indicates that despite the role of b-1,3-glucan as a structural component in fungal cell walls, its importance for the maintenance of the Paracoccidioides yeast cell wall is not enough to consider its synthesis as a promising antifungal target.

Besides its role on cell wall maintenance, b-1,3-glucan has been shown to be an immunostimulatory molecule, inducing TNFa production by macrophages and recognized by the pattern-recognition receptor (PRR) dectin-1, a C-type lectin expressed on the surface of many mammal cell types, including macrophages, monocytes, neutrophils and T-cells8 ^, 37 ^, 62. The drastic reduction of b-1,3-glucan in the cell wall of the pathogenic yeast phase of Paracoccidioides, and its replacement by a-1,3-glucan as an outermost layer, could work as an evolved mechanism to avoid host recognition by the fungus.

Alpha-1,3-glucan, also present as the outermost layer in Histoplasma capsulatum, has been shown to block innate immune recognition by host cells in this fungus37, reinforcing SAN-BLAS et al. 46 original proposal of this polysaccharide as a fungal virulence factor, and its emergence as a potential antifungal target. In P. brasiliensis, a-1,3-glucan is essentially a linear polysaccharide, with less than 3% of a-1,4-linked glucose branches, occasionally attached as single units to the a-1,3-backbone. A sole a-1,3-glucan synthase, Ags1, has been identified as responsible for the polysaccharide synthesis in Paracoccidioides 48. Also one a-1,3-glucanase, Agn1, has been identified in the Paracoccidioides spp. genomes as the only hydrolase gene present in them60.

A second element with a probable role in the synthesis of a-1,3-glucan has been identified. We refer to P. brasiliensis Amy1, a cytoplasmic a-1,4-amylase homologous to the H. capsulatum Amy1 (involved in the synthesis of a-1,3-glucan in the latter), that complements an H. capsulatum Amy1 mutant11. The enzyme showed a relatively higher hydrolyzing activity on amylopeptin than in starch, producing oligosaccharides 4 to 5 glucose residues long. The role of P. brasiliensis Amy1 in the synthesis of a-1,3-glucan is still unclear, although the authors suggest that it could be related to the generation of oligosaccharides that might act as primers for the biosynthesis of this polysaccharide by Ags1p11.

The absence of a-1,3-glucan from the natural fungal host, its role in virulence and the apparent simplicity of the mechanisms of synthesis and hydrolysis of this polysaccharide (one synthase, one hydrolase), make both processes attractive targets for the development of specific drugs. Blocking the mechanism of synthesis (by inhibiting either Ags1 or Amy1 activities), or increasing its degradation (by stimulating Ang1 activity), might result in the depression of fungal virulence, and trigger the natural immune response of the infected organism against the fungus11.

Despite its role in shaping the cell, the fungal wall is a dynamic structure. Rather than a rigid structure shaped only by polysaccharides such as chitin and b-1-3-glucan, the cell wall is a plastic and permeable structure, where those structural polysaccharides serve as scaffolds for proteins involved in signaling, cell-cell interactions, host attachment, as well as lipids related to the fungal secretion system, playing roles in the maintenance and modulation of the wall30, all of which gives the fungal cell the ability to adapt to environmental changes.

Contrary to the abundant information on cell wall polysaccharides in Paracoccidioides spp., data on the composition of cell wall proteins and lipids are scarce, mostly going back to the earlier and general data of KANETSUNA and SAN-BLAS. Recently, the PUCCIA's group in São Paulo, have started a thorough work on lipid composition of P. brasiliensis cell wall, comparing isolates Pb3 and Pb18, which belong to the PS2 and S1 species, respectively, isolates whose infection profiles in murine models are different27 ^, 36. They found 49 phospholipid species in Pb3 and 38 in Pb18, phosphatidylcholine and phosphatidylethanolamine being the most abundant phospholipidic molecules in both isolates. Brassicasterol was the most abundant sterol in the cell walls of both isolates, a compound already reported as the main sterol in the cytoplasmic membrane of P. brasiliensis yeast phase61.

Strategies for studies on gene function have proven cumbersome in Paracoccidioides spp., and although antisense RNA technology has recently emerged to overcome problems with gene knockout, the multinucleated nature of Paracoccidioides cells makes it difficult to apply it as a universal technique, since it only appears to work in a few strains. Progress towards development of molecular tools for gene function studies that work in all Paracoccidioides cryptic species are still needed, in order to provide information for definitely determine the roles of genes associated with regulation, synthesis and hydrolysis of the cell wall on growth, morphogenesis and pathogenesis of these important human pathogens. Not all has been said about their cell walls and therefore, we must expect further developments to arrive in due time.

WHEN ONE BECAME TWO... OR MORE

The observation that distinct isolates of P. brasiliensis were variable in mycological, antigenic and virulence aspects did not pass unnoticed through the years17 ^, 20 ^, 25 ^, 29 ^, 47. The first molecular typing studies, such as RAPD analysis, had already indicated that the genetic variability among P. brasiliensis isolates might be beyond the species level10 ^, 19 ^, 21. After applying the Phylogenetic Species Concept (PSC) in studies of Multi-Locus Sequence Type (MLST)31, it became clear that instead of a unique species, P. brasiliensis might contain several different cryptic species, a fact that was confirmed by additional studies13 ^, 32 ^, 50 ^, 54. Initially, three phylogenetic species were recognized: S1, a paraphyletic and recombining species; PS2, a monophyletic and recombining species; and PS3, a monophyletic and clonal species31 ^, 32. Once additional isolates, mainly from the Brazilian Central-Western Amazonia, were subjected to the same13 ^, 50 or different molecular analyses54 ^, 55, a fourth more diverging phylogenetic species (Pb01-like) emerged, later designated as P.lutzii 51 ^, 52.

Genomic analysis of P. brasiliensis and P. lutzii pointed to significant differences. While the genome sizes of the two cryptic species (S1/Pb18 and PS2/Pb03) of P. brasiliensis are similar (30.0 Mb and 29.1 Mb respectively), the P. lutzii genome is nearly 3 Mb larger (32.9 Mb). The total number of initial predicted genes varies between 7,875 in P. brasiliensis to 9,132 in P. lutzii. The three genomes are highly syntenic, though S1/Pb18 and PS2/Pb03 of P. brasiliensis, share a significantly higher percentage of sequence similarity (96%) to each other in comparison to P. lutzii (90%). Transposons may be one of the reasons for the enlargement of P. lutzii genome, as they constitute 16% of it, twice as much as that of P. brasiliensis S1/Pb18 and PS2/Pb03 genomes (8-9%)15.

Differences between P. lutzii and P. brasiliensis (S1, PS2 and PS3 species) were also detected in their proteomic yeast profiles by means of 2D electrophoresis and mass spectrometry. Out of 343 protein spots, 267 were differentially expressed, and 193 proteins were identified. Glycolysis/gluconeogenesis and alcohol fermentation-related proteins were more abundant in P. lutzii, indicating a higher use of anaerobic pathways for energy production. Antigenic proteins such as GP43 and 27-kDa, were less abundant in P. lutzii and PS2 genotype of P. brasiliensis 35. The P. lutzii GP43 orthologue is poorly expressed and contains few epitopes in common with the P. brasiliensis immunodominant antigen GP4326, contributing to serological diagnostic difficulties in patients infected with P. lutzii 6.

Before P. lutzii was actually recognized as a new species, HAHN et al.19 observed differences in pathology and response to treatment in paracoccidioidomycosis patients from the Central-Western region of Brazil where most P. lutzii isolates have been reported so far, as compared with other regions in which P. brasiliensis predominates (see below). This is a field that deserves urgent attention, in order to preserve the life of patients afflicted by this severe and frequently lethal disease.

FAMILIES ARE MESSY

Morphological and molecular findings indicate that both Paracoccidioides species might be classified as Ascomycota, order Onygenales, family Ajellomycetaceae59, in a natural group that also includes other vertebrate-associated fungi of the genera Histoplasma, Blastomyces, Emmonsia and Lacazia 22 ^, 44. This fungal family presents some common mycological and ecological features, such as dimorphism, arthroconidia production and occurrence in restricted geographic areas, and affinities for animal product derivatives or remnants like feces and uric acid9.

TWO HOMES, TWO WORLDS IN TENSION

The ecological niche of Paracoccidioides spp. is not completely understood18 ^, 38. P. brasiliensis has been isolated sporadically from soil and related materials (animal feces and dog food mixture with soil), and frequently from armadillos5. Several other domestic and wild mammals appear to be infected by the fungus, as determined by intradermic test, serology and molecular procedures7 ^, 16 ^, 39. The fungus has been demonstrated by molecular tools in soil samples from burrows of the nine-banded-armadillo, Dasypus novemcinctus 4 ^, 14 ^, 53 ^, 56, and also in aerosol samples, confirming the spread of the fungus by inhalation1.

The recent genome survey of the Paracoccidioides species proved that this pathogen has the metabolic apparatus to degrade cellulosic plant material in the soil35, making soil a probable saprobe habitat for Paracoccidioides. But more important is to elucidate its environmental requirements, tolerance and interaction with other species, in a multidimensional space. Paracoccidioides spp. and other Ajellomycetacean fungi, except Lacazia loboi, typically present at least two ecological niches, because of their dual saprobic and parasitic life style. During the saprobic mycelial phase these pathogens interact with different conditions and resources, such as periodic changes in temperature and humidity, competition with other microorganisms, and probable predation by amoebae and nematodes, commonly found in soil; during the parasitic yeast phase they are exposed to the mammal tissues, with temperature increase, hormone influences and response to the immune system.

As noticed in previous paragraphs, there are much more studies about the parasitic niche of Paracoccidioides than in its saprobic phase, reason for which BUSTAMANTE et al. 9 or TERÇARIOLI et al. 53, among others, evaluated the growth and conidia production of P. brasiliensis on soil in order to understand how the pathogen interacts with some abiotic factors, such as soil texture and availability of water. P. brasiliensis grows well both in clay and sand soil, provided they are saturated with water, a limiting factor for the fungal growth. It was also observed that soil containing high amount of exchangeable aluminium (H+Al) and low base saturation inhibit growth of the fungus; at the same time, it turned out that armadillos infected with P. brasiliensis were more easily found in soils low in H+Al, sandy and with medium to low concentrations of organic matter3. Most isolates were able to produce arthroconidia though a lower conidia production in PS2 genotype was observed in comparison with S1 isolates, a fact that could explain the unbalanced proportion of 1:9 (PS2:S1) isolates found in the same endemic area of Botucatu, SP, Brazil55.

Additional ecological studies are pending in order to detect and differentiate the cryptic species in soil, a survey that is already under way in several regions of the Brazilian geography, i.e., ARANTES et al.'s1 detection of P. lutzii in a hyperendemic area for P. brasiliensis by means of nested PCR amplicons from the ITS region, that clearly differentiates both species50. Such studies will help strengthen phylogeographic information on Paracoccidioides, operational to design a possible evolutionary scenario for the speciation process of this genus4.

So far, phylogeographic inferences revealed simultaneous geographic expansions of S1 isolates55, one of which is now considered a S1 subspecies, represented by Venezuelan isolates41. Another expansion gave rise to the Colombian PS3 species, geographically isolated in the Andes, once they emerged around eight million years ago. As there are no clear geographic barriers in the Brazilian Shield, there are no obvious geological allopatric events to explain the speciation processes that gave rise to S1, PS2, and P. lutzii. Probably this is a kind of microallopatry, where the ecological differences in the same geographic area may play an important role in the isolation and divergence of the species.

CLOSING REMARKS

When Carlos da Silva Lacaz was born in 1915, seven years had already passed since P. brasiliensis made its début in the annals of the Latin American medical mycology28. The fungus turned out to be the agent of one of the most relevant fungal diseases of the region, in terms of frequency and devastating consequences both for patient and society. The efforts and funds invested in state-of-art research on the subject have paid in the huge amount of information now available on biology, immunology, ecology, phylogeny and much more, leading to the design of better treatment procedures for the benefit of paracoccidioidomycosis patients, either from P. brasiliensis or P. lutzii. However, the puzzle is far from been solved, and improved protocols still await further fundamental and applied research to comply with one apothegm of the Hyppocratic Oath: ἐπὶ δηλήσει δὲ καὶ ἀδικίῃ εἴρξειν (abstain from doing harm), popularized in a Latin version as "Primum non nocere", (first of all, do not harm), a maxim that Lacaz always had in mind to act accordingly.