|Home | About | Journals | Submit | Contact Us | Français|
Dimorphic yeasts change between unicellular growth and filamentous growth. Many dimorphic yeasts species are pathogenic for humans and plants, being infectious as invasive hypha. We have studied the determinants of the dimorphic switch in the nonpathogenic fission yeast Schizosaccharomyces japonicus, which is evolutionarily close to the well-characterized fission yeast S. pombe. We report that camptothecin, an inhibitor of topoisomerase I, reversibly induced the unicellular to hyphal transition in S. japonicus at low concentrations of camptothecin that did not induce checkpoint arrest and the transition required the DNA checkpoint kinase Chk1. Furthermore, a mutation of chk1 induced hyphal transition without camptothecin. Thus, we identify a second function for Chk1 distinct from its role in checkpoint arrest. Activation of the switch from single cell bipolar growth to monopolar filamentous growth may assist cells to evade the source of DNA damage.
Yeasts and molds are major members of the kingdom Fungi. Molds grow as multicellular filamentous hyphae. On the other hand, yeasts propagate in a unicellular fashion by budding or by binary fission. However, many types of yeast can switch their growth modes, changing from unicellular growth to filamentous branching multicellular hyphae. This hyphal transition can be induced by a wide variety of environmental changes ranging from pH to the nature of the carbon source, and many species of dimorphic yeasts that are pathogenic for humans and plants are infectious in the hyphal form (15, 20).
Hyphal transition is a simple mode of cellular differentiation program that is turned on upon environmental changes. The fungi may differentiate to adapt to the environmental challenges. Especially in the case of Candida albicans strains that infect humans, the hyphal transition may function as an action to resist against attack from macrophages or neutrophils. Hyphae are more difficult to phagocytose (16). It can also eventually kill macrophages if hyphal transition is triggered after ingestion by macrophage (14). Indeed, C. albicans cells that cannot form hyphae are avirulent. However, inducing hyphal growth in pathogenic yeasts is not always readily achievable in the laboratory, and genetic analysis of the hyphal growth phase and transition to this phase is often limited by the lack of appropriate tools. Thus, genetically tractable nonpathogenic dimorphic yeasts are attractive models for investigating invasive hypha.
The nonpathogenic fission yeast Schizosaccharomyces japonicus is evolutionarily close to the well-characterized fission yeast Schizosaccharomyces pombe (5, 24). S. japonicus is dimorphic, transiting between unicellular and hyphal growth, and thus offers itself as an appropriate model to study this differentiation mechanism and the requirements of hyphal growth (25). In S. japonicus, hyphal growth occurs naturally on most solid medium and can occur over a range of nutrient conditions (26). It has been proposed that a gradient of nitrogen in the substrate is necessary to both initiate and direct hyphal growth in S. japonicus (26). In this report we establish conditions to induce hyphal growth in a microchamber in liquid media. In addition, we show that a low dose of the topoisomerase inhibitor camptothecin (CPT) induces hyphal differentiation under rich nutrient conditions and identify a role for the DNA damage checkpoint response in promoting the CPT-dependent transition from unicellular to hyphal growth. Genetic analysis demonstrates that this role of the checkpoint is distinct from checkpoint arrest, and we suggest it may provide an opportunity for S. japonicus to grow away from sources of genotoxic stress.
S. japonicus cells were cultivated as previously described (7). For rich media, YE (yeast extract [5 g], glucose [30 g/liter]) was used. For the induction of nutrient-dependent hypha, ME (malt extract [3 g], agar [20 g/liter]) was used. A final concentration of 2% agar was added for solid media. CPT (Sigma) was used to induce DNA damage-dependent hyphae. For marker selection in YE media, 40 mg of Geneticin/liter was used. For selection on solid Edinburgh minimal medium (EMM), 400 mg of Geneticin/liter was used. For nmt1 promoter induction, EMM was used for induction. For nmt1 repression, EMM was supplemented with 5 μM thiamine.
Strains used in the present study are summarized in Table Table1.1. Transformation of plasmids into yeast cells was performed by electroporation (7). The chk1 (SJAG_01680.2)-, rad9 (SJAG_01605.2)-, and rad51 (SJAG_01410.2)-null mutants were constructed as described previously (7). The majority of each open reading frame was replaced by a kanMX marker. One of the genes encoding histone H3 hht3 (SJAG_00302.2) was tagged with green fluorescent protein (GFP) by using a PCR-based method. The construction of strains was checked by colony PCR. For gene nomenclature, the Schizosaccharomyces group database of the Broad Institute (http://www.broadinstitute.org/annotation/genome/schizosaccharomyces_group/MultiHome.html) was referred to.
Cell density was measured by using a TVS062CA (Advantec). A saturated culture was diluted 10,000-fold in YE or YE containing CPT at time zero, and the optical density at 660 nm was measured every 1 min.
The chk1-hyp mutant was isolated through conditional hyphal screening. Colonies that were fringed with hypha at 30°C, but not at 37°C, were isolated from ethyl methanesulfonate mutagenized cells. Ten colonies were initially isolated; however, only one isolate reproducibly showed temperature-sensitive hyphal induction. We concluded that the mutation site was in the chk1 locus for the following reasons. (i) The genetic distance between the chk1-hyp allele and chk1::kan was <5.5 centimorgans (cM). Spore dissection analysis with NIG6374 showed all nine octads presented parental ditypes. (ii) A total of 480 random kanamycin-sensitive spores were temperature sensitive for hyphal growth (equivalent to 120 parental ditypes from 120 octads). Thus, the distance is <0.4 cM, which corresponds to 2.6 kb in another fission yeast, S. pombe (29). The formula used to determine the genetic distance was previously described (7). (iii) A mutation site was found in a conserved region of the chk1 open reading frame.
The details of the autonomous replicating sequence and vector plasmid pSJK11 will be described elsewhere (K. Aoki, R. Nakajima, and H. Niki, unpublished data). For thiamine-inducible constructs, each specified gene was under the 5′ untranslated region (5′UTR) of thi3/nmt1 (nucleotide [nt] −897 to start codon of SJAG_04866.2). pKF659, pKF660, and pKF724: chk1 (start codon with nt 493 of the 3′UTR), wee1 (start codon to nt 900 of the 3′UTR of SJAG_04280.2), and cds1 (start codon to nt 600 of the 3′UTR of SJAG_04287.2), respectively.
Cells were exponentially grown at 30°C in the YE liquid medium. The culture was applied to a microfluid culture plate [Y2(D); CellASIC Corp.], and the ONIX microfluid perfusion system (CellASIC Corp.) controlled the flow rate at 4 lb/in2. After incubation for 4 h, the medium was automatically exchanged for YE liquid medium containing 0.2 μM CPT. Cells in microfluid culture plate were observed with an inverted microscope (Axiovert 200M; Zeiss, Germany) fully controlled by a PC with Axiovision software v.4.8 (Zeiss), and images were captured by an charge-coupled device camera system (CascadeII:1024; Photometrics). The cells were held at the given temperature by using a thermostatic system for the microscope stage and objective lens. Hyphal induction by nutrient stress was carried out in the liquid EMM at 30°C. First, cells were incubated in the microfluid culture plate [Y2(D)] for 24 h circulating in the liquid EMM at a flow rate of 4 lb/in2. After the medium was exchanged for the liquid EMM without NH4Cl, the cells were incubated for 24 h at a flow rate of 4 lb/in2. After further incubation at a flow rate of 1 lb/in2, hyphal cells elongated toward fresh medium.
When S. japonicus was cultured on solid malt extract medium (where nitrogen is a limiting nutrient), smooth colonies were formed at first (Fig. (Fig.1a).1a). Subsequently, after a further protracted period of incubation (7 days), hyphae were observed to extend from the periphery of the colony (Fig. (Fig.1b).1b). Hyphal cells invaded into the substrate of the solid medium and remained in the agar after washing the surface of the medium (Fig. 1b and c). It has been reported that environmental stimuli can trigger hyphal differentiation, and growth on solid medium plus a nutrient gradient are required (26). However, we were able to use single cell cultivation in a microchamber to induce hyphal growth. Thus, growth on solid media is not prerequisite, and the hyphal transition can occur in liquid medium. The circulation of liquid media in the microfluidic chamber likely forms a nutrient gradient: we observed that growing hyphae elongated toward fresh medium (Fig. (Fig.1d1d and Fig. 2a and b; see also Movies S1 and S2 in the supplemental material).
Filamentous cells showed tip elongation at one end, and dense granules rich in cytoplasm were observed at the apical growing portion (Fig. 1d and e). Large vacuoles developed in the latter half of the cells, a sharp contrast to what is observed in single yeast cells (Fig. (Fig.1e).1e). The rate of cell elongation (0.35 ± 0.034 μm/min) was 5.6-fold faster than observed during yeast growth (0.062 ± 0.004 μm/min). A single large nucleus (labeled by histone H3-mcherry) was normally observed in the hyphae and, after nuclear division, hyphal cells divided by septation, positioned in the middle of the filamentous cell (Fig. (Fig.2a;2a; see also Movie S1 in the supplemental material).
Cells grown on yeast extract agar medium did not generate hyphae under normal conditions, likely because, in contrast to malt extract medium, nitrogen is not limiting (Fig. (Fig.1c).1c). However, we noticed that, when cells were treated with low doses of CPT, hyphal growth on solid YE medium was evident. All colonies on solid YE medium containing CPT were fringed with hyphae, and the hyphal cells invaded into the body of the agar (Fig. (Fig.1c).1c). CPT increases the half-life of the topoisomerase I cleavage complex with DNA, and toxicity is largely a result of the collision of replication forks with TopoI-DNA complex, which generates potentially lethal lesions that require recombination for their repair (9). These DNA lesions activate the checkpoint responses, resulting in cell cycle arrest (28).
Wild-type cells retained their colony-forming ability when treated with 1.0 μM CPT (Fig. (Fig.3a).3a). However, checkpoint defective cells deleted for either the rad9 or chk1 genes (encoding a checkpoint clamp component or a checkpoint kinase, respectively) displayed decreased viability (6, 8, 27). The cell growth of a rad51 mutant was, as expected, highly susceptible to CPT treatment, displaying extreme sensitivity to only 0.1 μM CPT. rad51 encodes the RecA homolog, which is required for the strand invasion step of recombination-dependent repair mechanisms (13). Thus, the genotoxic stress caused by concentrations of CPT below 0.5 μM is mainly repaired by rad51-dependent pathways, whereas rad9-chk1-dependent DNA damage responses were not required until higher doses.
Relatively low concentrations of CPT (0.1 to 0.2 μM), which do not require the chk1 pathway for S. japonicus to survive efficiently, were sufficient for inducing hyphal differentiation (Fig. (Fig.1c).1c). Moreover, when wild-type cells were exposed to the same concentration of CPT in liquid medium, they were also induced to undergo hyphal growth (Fig. (Fig.3b).3b). These concentrations did not affect cell growth as measured by the optical density, even when the chk1 gene was deleted (Fig. 3c and d). The hyphal cells induced by CPT treatment in liquid phase elongated by apical growth and dense granules accumulated at the growing tips (Fig. (Fig.3e3e and Fig. Fig.4a).4a). A single large nucleus appeared in each elongated cell, and large vacuoles were well developed in the latter half of the cells, as seen with hyphal cells induced by nutrient stress.
It has been reported that the induction of hyphal differentiation in response to nutrient stress was restricted to culture temperatures between 18 and 35°C (26). We have examined the effect of temperature to hyphal growth. We found that, in our lab, nutrient stress-induced hypha was still observed at 37°C but was greatly suppressed at 40°C (see Fig. S1 in the supplemental material). Similarly, CPT-induced hyphal differentiation reduced at 40°C, however in lesser extent. The transition from yeast cells to hyphal cells, and vice versa, was monitored in liquid medium using the microchamber and time-lapse imaging (Fig. (Fig.4a,4a, b; see also Movies S3 and S4 in the supplemental material). After the addition of CPT, yeast cells initiated monopolar filamentous growth. The rate of cell elongation (0.22 ± 0.018 μm/min) was 3.6-fold faster than that observed during yeast growth. We also noticed that relatively larger nuclei (labeled by histone H3-GFP) developed in hyphal cells (Fig. (Fig.4a).4a). However, chromosome segregation actively took place during CPT-stimulated hyphal growth, as was also the case for nutrient-stimulated hyphae (Fig. (Fig.2a2a and Fig. Fig.4a;4a; see also Movies S1 and S3 in the supplemental material).
We conclude from our observations that the cell cycle in the hyphal phase progressed in a distinct manner from that seen in the yeast phase. Elimination of CPT from the liquid medium easily recovered yeast growth, which indicates that continued exposure is required to maintain hyphal growth (Fig. (Fig.4b;4b; see also Movie S4 in the supplemental material). We also tested whether the related fission yeast, S. pombe, responded to CPT exposure by initiating hyphal growth. S. pombe cells are more resistant to CPT, growing well even at 5 μM. S. pombe cells did not have induced invasive growth or cell elongation with well-developed vacuoles upon exposure to CPT and did not present hyphae (Fig. 3f and g).
Although H2O2, which may generates DNA damage-like stress, triggers their hyphal transition in pathogenic fungi, a reactive oxygen source such as H2O2 did not induce hypha in S. japonicus in Fig.S2 in the supplemental material. However, for S. japonicus, other DNA-damaging agents such as methyl methanesulfonate (MMS), bleomycin, and gamma rays, or indeed the spontaneous DNA damage stress generated in Δrad51 cells, also resulted the induction of hyphal growth (Fig. (Fig.1c;1c; see Fig. S3 in the supplemental material). Thus, hyphal growth appears to be a “differentiation” event triggered by the presence of sublethal amounts of DNA damage, and this is specific to the S. japonicus species.
DNA damage is sensed by the checkpoint apparatus through the combined action of ATR-ATRIP association with RPA-coated single-stranded DNA (ssDNA) and the loading of the checkpoint clamp at junctions between ssDNA and double-stranded DNA. Subsequently, the ATR kinase phosphorylates and activates Chk1, which implements cell cycle arrest in G2 by regulating Wee1, the Cdk inhibitor, and Cdc25, the Cdk activator. Indeed, checkpoint arrest defect was observed in Δchk1 cells when high concentration of CPT was applied in Fig. S4 in the supplemental material.
To establish the role of the checkpoint pathway on propagating the transition to hyphal growth, deletion mutants of rad9 and chk1 were examined. Both mutants were able to grow on solid YE medium containing 0.2 μM CPT, but neither formed surrounding hyphae, and no evidence of invasive growth into the agar was observed (Fig. (Fig.1c).1c). The deletion mutant of chk1 was also treated with CPT in a microchamber, and no hyphal differentiation was observed (Fig. (Fig.4c;4c; see Movie S5 in the supplemental material). This defect was specific for CPT-induced hyphal differentiation since both mutants maintained their ability for invasive growth on malt extract media (Fig. (Fig.1c1c).
Activation of the Chk1 kinase leads to phosphorylation of target proteins, including those that orchestrate the cell cycle arrest response (2). DNA damage-induced hyphal differentiation appears to be a downstream response to Chk1 activation, and thus we predicted that overproduction of Chk1 protein would trigger hyphal differentiation in the absence of DNA damage. We cloned the chk1 gene of S. japonicus into a vector with an autonomous maintenance sequence and an inducible promoter derived from the S. japonicus nmt1 locus. Expression of the promoter is induced by depletion of thiamine, and expression can thus be regulated by growing cells in EMM either with (off) or without (on) thiamine. Hyphal cell growth was moderately induced in cells harboring the chk1 plasmid compared to the empty vector even on medium containing thiamine, likely due to basal levels of transcription observed for the “off” state. Consistent with this, when transcription was induced, well-developed hyphae surrounded colonies harboring the chk1 plasmid but not the empty vector (Fig. (Fig.5a).5a). These chk1-induced hyphae were typical hyphae that divide and grow monopolarly and developed dense granules at the apical growing portion and vacuoles at latter half of the cell (Fig. (Fig.6;6; see Movie S6 in the supplemental material). Thus, increased chk1 transcription induced hyphal growth.
We also tested the ability of the cds1 (Chk2) and wee1 genes to induce hyphae. Wee1 is a mitotic inhibitor that is a downstream target of Chk1, whereas Cds1 is a second mitotic kinase that, in S. pombe, is specific to intra-S-phase checkpoint signaling. Neither was able to induce the hyphal transition or cell elongation (Fig. (Fig.5a;5a; see Fig. S5 in the supplemental material) (4, 18, 19).
To further examine the mechanism underlying the CPT-dependent hyphal growth transition, we performed a screen for mutants able to initiate hyphal growth in the absence of DNA damage in a temperature-dependent manner (see Materials and Methods). A single stable mutant was identified that presented hyphae on rich media without CPT when shifted to temperatures below 33°C (Fig. (Fig.5b,5b, c, and d and Fig. Fig.7;7; see Movie S7 in the supplemental material). Genetic mapping and sequence analysis indicated that this mutation lay within the chk1 gene, further confirming the contribution of chk1 in hyphal differentiation. The chk1-hyp mutation was a single base substitution in the portion of the gene encoding the C-terminal domain of Chk1 resulting in a missense pro470leu change in a residue conserved among several yeasts (Fig. (Fig.5e5e).
We have demonstrated that CPT induces a yeast-hypha transition in S. Japonicus. CPT was efficient in inducing hyphal growth compared to other DNA damage agents, such as MMS and bleomycin, and can be used to initiate hyphal growth at concentrations where it does not interfere with the growth of the culture. Thus, the level of DNA damage sufficient to induce hypha does not require checkpoint arrest for its efficient repair. Furthermore, the transition from yeast growth to hyphal growth is reversible, and thus this experimental system can be used to dissect the yeast-to-hypha or the hypha-to-yeast transition. Our findings also have implications for hyphal differentiation in S. japonicus and potentially other fungal species, since they reveal a mechanism for dimorphism. It should be noted that C. albicans can differentiate to hypha inside macrophages where they receive the reactive oxygen attack (14, 16). We also demonstrated that S. pombe, which is phylogenetically related to S. japonicus, even being in the same genus, did not show the DNA damage-induced dimorphic switch, even though it is well established that it can undergo this transition in response to nutrient stress (1, 5). Thus, although very similar, these two yeast species have distinct differences regarding their DNA damage response and their growth profiles, suggesting there are likely other differences in cellular function.
The pathogenic fungus C. albicans is polymorphic, and many environmental stimuli can trigger the transition to hyphae (3, 15). It was recently reported that DNA replicative stress could induce filamentous growth in C. albicans (21). However, the filamentous growth always associates with a slow S phase, and whether the filamentous growth can be induced without an associated checkpoint arrest remains unclear (21). In Saccharomyces cerevisiae, the slowing down of DNA replication by hydroxyurea or MMS treatment can induce filamentous growth, and checkpoint proteins Mec1, Rad53, and Swe1 are required. However, in this organism checkpoint activation per se is not sufficient to induce this differentiation event since bleomycin treatment does not induce filamentous growth (10). Moreover, it is not clear whether the filamentous growing cells seen in Saccharomyces cerevisiae in response to S-phase inhibition are true hyphal cells. We show here that in S. japonicus several cellular features, including invasive growth, apical elongation, extreme filamentous growth, and vacuolation, all of which are specific to hyphal cells, can be observed in differentiated growing cells under DNA damage stress.
In particular, the development of vacuoles and the enlargement of nucleus are unique changes in the subcellular structure in hyphal cells. Hyphal elongation requires a rapid increase in cell volume, and vacuoles that are filled with water would help to increase cell mass quickly. Indeed, it seemed likely that elongation of vacuoles coupled with elongation of cell tips (Fig. (Fig.2a;2a; see Movie S1 in the supplemental material). Many hyphae have relatively large nucleus. Compared to yeast cells, the area of the nucleus was 1.7 to 3.6 arbitrary units (average, 2.2 [n = 7]) fold large in hyphae. Nuclear contents, which are estimated by fluorescence intensity of histone H3-GFP, also increased in proportion to the area of the nucleus. Although we have not examined actual DNA content, it may be possible that cells become polyploid as they differentiate to hyphae. It is known that nuclear size of S. pombe increases as the cell size increases irrespective of the DNA content (17); thus, further study of cell cycle regulation in hypha is needed to elucidate enlargement of nucleus.
Mechanistically, the signal that results in the dimorphic switch in response to DNA damage is dependent on the checkpoint DNA damage sensing pathway and the checkpoint kinase Chk1. In the absence of genotoxic stress, we demonstrate that untimely activation of Chk1 is sufficient to induce hyphal differentiation in S. japonicus. Furthermore, we isolated chk-hyp, a mutation that activates the switch to hyphal growth in the absence of DNA damage, most likely acting as a gain-of-function mutation. The C-terminal domain of Chk1 in which the mutation was occurred is well conserved from yeast to humans and is proposed to act as an autoinhibitory domain (11, 12). In S. pombe, a Glu472Asp mutation within the domain leads to a gain-of-function phenotype when expressed at multiple copies (12), suggesting that the pro470leu mutation of S. japonicus may also be a gain-of-function mutation that partially activates Chk1 kinase activity. Interestingly, the S. pombe a Glu472Asp mutation does not activate checkpoint arrest or affect cell cycle progression when present at a single copy. However, the mutation of a single copy can activate transition of hyphal growth in S. japonicus. This contrast indicates that the magnitude of the Chk1 kinase activity might be different in two different pathways: cell cycle arrest and hyphal transition. Thus, the similar mutation in S. pombe does not induce checkpoint arrest, and S. japonicus chk1-hyp mutants continue cellular growth, suggesting that unlocking the autoinhibition of Chk1 is sufficient for hyphal differentiation but that further activation of the Chk1 pathway is needed for checkpoint arrest.
Although we have not been able to construct wee1- or cdc25-defective strains, we observed that the overproduction of chk1 but not of wee1 could induce hyphae. Thus, our results suggest that Chk1 may activate a different target other than cell cycle proteins to promote hyphal differentiation of S. japonicus cells (Fig. (Fig.8).8). However, we cannot completely exclude the possibility that these Chk1's cell cycle targets are not involved in induction of chk1-dependent hyphal growth. Alternatively, other functions of wee1 or cdc25 may contribute to hyphal transition, since Wee1 is involved in nutrient sensing. To better understand targets under the activated Chk1, we succeeded in isolating mutants that are defective in hyphal differentiation caused by CPT, but not by nutrient stress (unpublished data). Further analysis of the mutants may reveal the pathways of hyphal differentiation.
Interestingly, Chk1 is a key factor in the induction of cellular differentiation in higher eukaryotes. In fruit flies, chk1/grapes was isolated as a maternal effect gene that is required to turn on the differentiation program at the mid-blastula transition (22, 23). In yeasts, it has been proposed that hyphal differentiation is one of several nutrient-induced cellular adaptations that promote survival in the continuously changing nutritional environment and provides motility to this otherwise nonmotile organism (25, 26). Our data lead us to propose that some yeasts also use hyphal differentiation to take evasive action in response to DNA damage-inducing stress. In S. japonicus this Chk1 signaling pathway is used not only to promote cell cycle arrest in response to checkpoint activation at high doses of DNA damage but also to promote hyphal differentiation at low doses, offering the opportunity to grow away from localized sources of such stress. This could be a conserved developmental program utilized to avoid environmental stress in other yeasts, including pathogenic organisms.
We thank Nanayo Ishihara for technical assistance and all members of the Niki lab for helpful comments and suggestions.
This study was supported by Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
Published ahead of print on 5 March 2010.
†Supplemental material for this article may be found at http://mcb.asm.org/.