This year marks the 60th anniversary of the official description of the thermally dimorphic fungus, Talaromyces marneffei (previously called Penicillium marneffei), which is an important fungal pathogen endemic in Southeast Asia especially to HIV-positive patients, by Professor Gabriel Segretain [1]. This fungus was first isolated in 1955 by Capponi and Sureau from laboratory Chinese bamboo rats (Rhizomys sinensis), which are native in the Central Highlands of Vietnam, at Institut Pasteur de Dalat [2], where it caused fatal spontaneous disseminated infection involving the reticuloendothelial system in three of the rats. The pathogenicity of T. marneffei was further demonstrated by Segretain, Capponi and Sureau in various animal models, including mice, rats, hamsters and guinea pigs [2, 3].

Six decades have passed since the discovery of T. marneffei; however, our understanding on this peculiar fungus is still poor despite its clinical significance. Such a lack of knowledge could be reflected by the fact that up to the end of August 2019, there are only ~730 published articles about T. marneffei indexed in PubMed. This figure significantly lags behind those for Aspergillus fumigatus (~11,480 articles) as well as other thermally dimorphic pathogenic fungi such as Histoplasma capsulatum (~4050 articles) and Coccidioides immitis/C. posadasii (~2020 articles), and it is only slightly more than two times than that for the notorious, recently emerged multidrug-resistant yeast, Candida auris, discovered in 2009 (~300 articles). Mycopathologia, as a journal dedicated to the understanding of human and animal fungal diseases with over 80 years of history [4], has long supported research on T. marneffei. Notably, around 4% of the ~730 T. marneffei articles were published in Mycopathologia [3, 5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34], including one of the initial works by Segretain [3].

Since the first discovery of T. marneffei, its taxonomy had remained stable until early 2010s. The original description by Segretain depicted T. marneffei as a Penicillium-like fungus (and so it was first named as ‘P. marneffei’), and this fungus was classified in Penicillium section Biverticillium following Biourge’s taxonomy or section Asymmetrica subsection Divaricata following Raper’s and Thom’s taxonomy [1]. At room temperature, its morphology was found to partly resemble P. janthinellum and P. citrinum, with bluish light grey–green colonies and diffusing red pigment (Fig. 1a, b). However, at 37 °C ‘P. marneffei’ exhibited a very different morphology. Colonies were bacterial-like, hairless, colourless, matt smooth at first and later pleated cerebriform, with no diffusing pigment. Instead of hyphal filaments, the fungus was composed of separate arthroconidia at this temperature, which were divided by transverse partitioning (fission) [1] but not budding [2]. In mature culture, the cells were almost spherical in shape and this was similar to the morphology of the fungus in vivo [3] (Fig. 1c–e). The affiliation of the fungus to the genus Penicillium, based on phenotypic characteristics, had lasted for over 50 years. In 2011, based on phylogenetic analyses inferred from the RNA polymerase II largest subunit gene (RPB1) and internal transcribed spacer (ITS) region as well as extrolite profiling, ‘P. marneffei’ was transferred to the genus Talaromyces together with other members of Penicillium subgenus Biverticillium. As a result, the fungus attained its present name ‘T. marneffei’ [35]. Such transfer was also supported by phylogenetic analysis based on mitochondrial genomes [36]. Currently T. marneffei is the only member of the genus recognised to cause invasive infections in humans and animals.

Fig. 1
figure 1

Morphology of Talaromyces marneffei. a On Sabouraud glucose agar, after 6 days of incubation at room temperature, colonies are yellowish green with diffusing red pigment. b The conidiophores are usually biverticillate. Oval, smooth-walled conidia are produced in chains. Scale bar = 20 µm. c On Sabouraud glucose agar, after 6 days of incubation at 37 °C, colonies are yeast-like and creamy. Diffusing red pigment is no longer observed. d In Sabouraud glucose broth, after 3 days of incubation at 37 °C with shaking at 250 rpm, the fungus grows as short hyphae composed of undissociated arthroconidia. Scale bar = 20 µm. e In yeast nitrogen base supplemented with 1% mycological peptone [118], after 3 days of incubation at 37 °C with shaking at 250 rpm, arthroconidiation is not observed. Instead, the fungus grows as individual yeast-like cells and divides by fission and this mode of growth resembles the in vivo situation. Scale bar = 20 µm

Full size image

Given the high mortality rate of T. marneffei infection in untreated patients, especially in immunocompromised patients, it is important to identify the possible sources of talaromycosis. In particular, the fact that HIV patients who have travel histories to endemic areas could also become infected by T. marneffei implies that short-term exposure to the source(s) of the fungus is sufficient to trigger infection and that the infection source(s) should originate from an environment contacted by tourists [37]. Studies beginning from the 1980s identified a number of bamboo rats species, including Chinese bamboo rats, hoary bamboo rats (R. pruinosus), large bamboo rats (R. sumatrensis) and lesser bamboo rats (Cannomys badius), as the natural carriers of T. marneffei in southern China, central and northern Thailand as well as India [9, 19, 38,39,40,41,42,43,44,45]. The fungus could also be recovered from the faeces of bamboo rats as well as soils from their burrows [40, 43, 45, 46]. Apart from bamboo rats and their associated soil samples, recently the fungus has been detected from nasal swabs of outdoor dogs in Chiang Mai, Thailand, as well, although fungal culture was not successful [37]. A case-control study in 1997 demonstrated that exposure to or consumption of bamboo rats did not constitute a risk factor for talaromycosis [47]. There was also no significant association between exposure to bamboo thickets or forests, where the rodents reside, to T. marneffei infection [47]. Instead, it was shown that recent occupational exposure to animals or plants, probably involving soil, was associated with talaromycosis [47]. Indeed, a previous environmental study detected T. marneffei from soil samples collected outside the habitats of bamboo rats, including a bat cave, an elephant camp and surroundings of a Buddhist Temple using molecular methods [48]. Unfortunately, viable culture of the fungus could still not be obtained from these positive soil samples [48]. These findings suggested that an environmental reservoir, likely soil, should exist for T. marneffei and humans and bamboo rats, and possibly dogs, may acquire the infection from this common environmental source. Interestingly, although earlier studies in Thailand and Vietnam showed that the incidence of T. marneffei infection increased during rainy season [49, 50], a later study in Vietnam found that humidity, but not rainfall, temperature nor wind, was the environmental predictor of T. marneffei hospital admission [51]. It was suspected that humidity may facilitate the environmental reservoir of T. marneffei to expand, favour fungal growth or aid in conidia release into the environment [51]. Further efforts are in need to determine the environmental source(s) of this fungus and to connect this/these environmental reservoir(s) with the animal hosts and humans such that the route of infection acquisition can be identified or confirmed.

Temperature-dependent morphological dimorphism is one of the characteristics of T. marneffei. Indeed, this fungus is the only known Talaromyces species exhibiting such a feature. Although the change in morphology has been well studied during phase transition [52], the underlying molecular mechanism still warrants much further investigation. From 2000 onwards, a number of genes have been identified to be involved in morphogenesis and during phase transition, including abaA [53], brlA [54], cflA [55], cflB [56], drkA [57], gasA [58], gasC [59], hgrA [60], madsA [61, 62], myoB [63], pakA [64], pakB [65], PmHHK1 [66], rasA [67], rfxA [68], rttA [69], sadA [70], sakA [71, 72], slnA [57], sskA [73], stuA [74], tupA [54] and yakA [75], by loss-of-function studies. The majority of these genes are related to transcriptional regulation [53, 54, 60, 62, 68, 74] or G-protein signalling [55, 56, 58, 59, 67] or encode kinases [57, 64,65,66, 71, 72, 75]; the functions of these genes were recently thoroughly reviewed by Andrianopoulos and his colleagues [76, 77]. Other than these genes, differentially expressed gene analyses by suppression subtractive hybridisation [78], microarray [79, 80] or RNA-sequencing [61, 81] identified groups of genes that are signature to phase transition. In addition to protein-coding genes, small RNA-sequencing also revealed 24 microRNA-like small RNA (milRNA) candidates which were more abundantly expressed during the hyphal stage than the yeast phase [82]. However, further functional studies are needed to authenticate the participations of these genes in dimorphic switching. Moreover, how T. marneffei detects the temperature stimulus and transduces the environmental signals to effect morphological change is still unclear. Elucidation of the molecular mechanism for this could help reveal potential drug targets for stopping dimorphic switching to the yeast phase in vivo so as to prevent the fungus from evading host immune response by hiding intracellularly.

The advancement of various omics technologies in the twenty-first century has allowed more in-depth and sophisticated characterisation of the mycology of T. marneffei. The first draft genome of the fungus, based on the original ex-type strain ATCC 18224T isolated by Capponi and Sureau, was sequenced using the Sanger strategy. It was around 28 Mb in length and was released online in 2007 [83]. A few years later, using a similar approach the draft genome for a second strain (PM1), isolated in Hong Kong, was published [84]. Subsequent second- and third-generation sequencing technologies further improved this draft genome [61]. Earlier this year, the genome of T. marneffei, based on a Chinese clinical strain TM4, was completed with the help of optical mapping; it was revealed that the T. marneffei genome consisted of eight chromosomes [85]. The availability of the whole genome sequences of T. marneffei has boosted further research on this fungus, especially in the past decade; knowledge generated from the fungal genomes as well as additional transcriptomic, proteomic and metabolomic studies was discussed in detail in a recent review [86]. One example was that the pigmentation phenomenon observed for T. marneffei could now be explained, with the molecular identities of the various pigments uncovered [87,88,89,90].

While there is accumulating information about the basic mycology of T. marneffei, it is equally important to understand what makes this special fungus a successful pathogen to humans and animals. Thermal dimorphism, as discussed above, is one of the more well-studied properties of the fungus which aids in causing infection. Apart from this, a number of virulence factors of T. marneffei have been identified. These virulence mechanisms include the production of melanin [87, 91, 92], mitorubrinol and mitorubrinic acid (yellow pigment) [88], aspartyl protease [93], catalase–peroxidase [94], laccases [95] and superoxide dismutase [96], induction of the glyoxylate cycle at host’s body temperature [97,98,99], utilisation of non-preferred nitrogen sources in host’s environment [100], utilisation of the methylcitrate cycle for detoxification of propionyl-CoA [101] as well as sequestration of host’s proinflammatory lipids [102,103,104,105]. Interestingly, a study last year found that T. marneffei infected with the mycovirus Talaromyces marneffei partitivirus-1 exhibited hypervirulence in a murine model [106], although the underlying mechanism still awaits additional investigation. In addition to the intrinsic virulent properties of T. marneffei, how it interacts with the human/animal hosts may also contribute to pathogenicity. Although the infection route for talaromycosis is still unclear as discussed above, it is generally believed that patients acquire pathogenic conidia through inhalation. After entering the host’s airway, conidia bind to the extracellular matrices [107,108,109,110], which is mediated by glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [110], and adhere to the host bronchoalveolar epithelium. There, the fungus interacts with epithelial cells [111, 112] and is phagocytosed by pulmonary macrophages. Our limited understanding on macrophage immunity against T. marneffei and the fungal adaptation response was summarised in three review articles [52, 77, 113]. Yet, the ability of T. marneffei to survive inside macrophages might also help them evade host immunity [52, 77]. In addition to macrophage response, CD4+ T cells were also shown to be key mediators in anti-T. marneffei response [114, 115]. It is of note that there have been an increasing number of talaromycosis cases in patients with non-HIV-related immunodeficiency in recent years [116] and whether the pathologies in these patients are the same or similar to those observed in HIV-positive patients requires further research. A number of experimental infection models, including nematode (Caenorhabditis elegans), greater wax moth (Galleria mellonella), zebrafish (Danio rerio), mouse (Mus musculus) and human primary monocyte-derived macrophages, have been established for T. marneffei, and their pros and cons were reviewed in a recent article by Weerasinghe et al. [77]. More studies are needed to verify how representative these infection models are for the in vivo conditions during T. marneffei infections in humans.

Following 60 years of effort, the mist of T. marneffei has somehow faded away. We have gained some knowledge on its taxonomy, source(s), underlying mechanism of thermal dimorphism, molecular genetics, virulence and pathogenesis. However, the jigsaw puzzle is still far from completion. Continuous collective endeavour is needed to uncover the T. marneffei mystery so that this fungus, currently recognised as one of the ten most feared fungi in the world [117], no longer poses threats to humanity.