Article | Open | Published:

Molecular basis of resistance to the microtubule-depolymerizing antitumor compound plocabulin

Scientific Reportsvolume 8, Article number: 8616 (2018) | Download Citation

Abstract

Plocabulin (PM060184) is a microtubule depolymerizing agent with potent antiproliferative activity undergoing phase II clinical trials for the treatment of solid tumors. Plocabulin shows antifungal activity virtually abolishing growth of the filamentous fungus Aspergillus nidulans. A. nidulans hyphae depend both on mitotic and interphase microtubules, as human cells. Here, we exploited the A. nidulans genetic amenability to gain insight into the mechanism of action of plocabulin. By combining mutations in the two A. nidulans β-tubulin isotypes we obtained a plocabulin-insensitive strain, showing that β-tubulin is the only molecular target of plocabulin in fungal cells. From a genetic screen, we recovered five mutants that show plocabulin resistance but do not carry mutations in β-tubulin. Resistance mutations resulted in amino acid substitutions in (1) two subunits of the eukaryotic translation initiation factor eIF2B activating the General Amino Acid Control, (2) TIM44, an essential component of the inner mitochondrial membrane translocase, (3) two transcription factors of the binuclear zinc cluster family potentially interfering with the uptake or efflux of plocabulin. Given the conservation of some of the identified proteins and their respective cellular functions in the tumor environment, our results pinpoint candidates to be tested as potential biomarkers for determination of drug efficiency.

Introduction

Plocabulin (PM060184, PM184) (Fig. 1a) is a microtubule (MT) depolymerizing agent currently undergoing phase II clinical trials for the treatment of solid tumors. Initially isolated from the sponge Lithoplocamia lithistoides, the compound is currently produced by total synthesis1. Plocabulin binds to β-tubulin, sharing its binding site with maytansine and rhizoxin (the “maytansine-site”)2,3. Plocabulin binding to β-tubulin leads to a decrease of MT dynamic instability and to MT depolymerization in a concentration-dependent manner4. The drug inhibits migration of interphase cells and induces a prometaphasic arrest in mitotic cells, resulting either in the induction of caspase-dependent apoptosis or in the formation of multinucleated, polyploid interphase-like cells4. Plocabulin shows potent antitumor activity in tumor xenograft models, including those tumors overexpressing p-glycoprotein4, and has demonstrated vascular-disrupting activity5. Thus, plocabulin is a promising anticancer drug.

Figure 1
Figure 1

The sole molecular target of plocabulin in A. nidulans is β-tubulin. (a) Chemical structure of plocabulin. Based on the structure of tubulin in complex with plocabulin [ref.2], Asn100 in A. nidulans β-tubulin, BenA, would form a hydrogen (H) bond (dashed line) with plocabulin position 1 and would contribute to hydrophobic interactions (ellipse) with position 27. (b) Growth test of a wt and the benAN100I strain. Consistent with Asn100 forming part of the PM06018 binding site, benAN100I confers resistance to plocabulin, albeit partial. benAN100I shows hypersensitivity to eribulin, a drug with a different binding site on β-tubulin. (c) Alignment of the two A. nidulans β-tubulin isotypes, BenA and TubC, with Bos taurus β-tubulin TUBB2B. (BenA: AN1182, TubC:AN6838; TUBB2B: B. taurus TBB2B_BOVIN, Q6B856). Both β-tubulin isotypes conserve the amino acid residues that form the plocabulin binding site, i.e. Asn99/Asn100/Lys103/Val179/Val180/Phe394/Tyr398. Red box: aa forming hydrogen bonds with plocabulin position 1, Green box: aa forming hydrogen bond with plocabulin position 13, Blue low dash: aa participating in hydrophobic interactions with plocabulin position 272. (d) Strategy used to delete the tubC ORF: The Aspergillus fumigatus pyrG (pyrGAf) complements pyrimidine auxotrophy of pyrG A. nidulans. This allows the selection of double homologous recombination that mediates replacement of tubC by pyrGAf (tubCΔ::pyrGAf) when an appropriated DNA construct is used for transformation. (e) Growth on plocabulin of progeny from a cross between tubCΔ (P1) and benAN100I (P2): ≈42% (n = 24) displayed plocabulin sensitivity, like the wt and tubCΔ strains, ≈21% showed partial resistance like benAN100I and ≈37% showed insensitivity to plocabulin (indicated by orange arrows) unlike either parental. (f) Frequency of the super-resistant class was similar to the expected for the double benAN100I tubCΔ mutant (note that selective plates did not contain pyrimidines, see also text). Genotyping indeed verified that super-resistant strains were benAN100I tubCΔ. (G) The double benAN100I tubCΔ mutant shows super-resistance to 10 and 50 µM plocabulin, indicating that at these concentrations the only direct target of plocabulin relevant to fungal growth is β-tubulin.

The haploid filamentous fungus Aspergillus nidulans is an established genetic model. Unlike the prototypic fungal genetic model Saccharomyces cerevisiae, A. nidulans hyphal cells require MTs not only for enacting mitosis and driving nuclear movement6, but also for accomplishing long-distance transport of endosomes and peroxisomes7,8 and for efficiently delivering secretory carriers to exocytic sites, in collaboration with actin cables9,10. Thus, the roles of MTs in A. nidulans resemble those in mammalian cells11,12,13. This dependence on MT transport is probably imposed by the large intracellular distances (in the order of 0.1 mm) between apex-proximal and distal regions of the tubular hyphal tip cells11. It is thus unsurprising that the field of MT research owes important advances, such as the characterization of the first eukaryotic genes for α, β, and γ-tubulins14,15,16, to genetic studies carried out with this fungus.

We have previously shown that plocabulin exhibits antifungal activity, compromising A. nidulans colony formation at concentrations >5 µM3 (Fig. 1b). Addition of plocabulin to growing hyphae expressing GFP-tagged α-tubulin (TubA-GFP) results in MT depolymerization3 (see also Results), indicating that a common mechanism of action (MoA) underlies both the antifungal (against A. nidulans) and anti-proliferative (against human tumor cell lines) activities displayed by the drug. This suggested that, by using the powerful genetic tools available for A. nidulans, it would be possible to gain insight into the MoA of plocabulin and the cellular potential for drug resistance. As a proof of concept we had previously isolated from a random genetic screen plocabulin-resistant A. nidulans mutants that constituted in vivo evidence for the existence of a novel plocabulin specific binding site on β-tubulin. These strains carried mutations leading to the Asn100Ile substitution in the major β-tubulin BenA [reported in]3, a substitution also shown to confer resistance to rhizoxin17. The substitution did not have any effect on growth on medium without plocabulin, strongly indicating that Asn100Ile would affect a putative plocabulin binding site on β-tubulin, rather than general MT structure or dynamics3. Notably the crystal structures of tubulin bound to plocabulin, rhizoxin and maytansin were consistent with these conclusions, showing that Asn100 is indeed part of the β-tubulin binding site that contacts a common pharmacophore on these compounds2.

Although Asn100 in BenA is critical for plocabulin binding, A. nidulans carrying benAN100I is only partially resistant to the drug3 (Fig. 1b), suggesting that intracellular targets other than BenA would contribute to plocabulin antifungal activity. By combining directed genetic manipulations with an unbiased genetic screen coupled to next generation sequencing (NGS) and genetic mapping, we investigated (1) whether plocabulin has additional intracellular targets other than β-tubulin in A. nidulans and (2) the landscape of genes that have the potential, when mutated, to confer plocabulin resistance on A. nidulans. We show that β-tubulin is the sole direct intracellular target of plocabulin relevant to fungal survival and growth. This negates the hypothetical existence of secondary intracellular targets contributing to drug activity. We then establish a causal relationship between plocabulin partial resistance and mutations in (1) two subunits of the eukaryotic translation initiation factor eIF2B, (2) TIM44, a conserved and essential component of the translocase of the inner mitochondrial membrane, and (3) two DNA binding domain-containing proteins of the zinc binuclear cluster family.

Results

Remaining sensitivity of the benA N100I strain to plocabulin is due to the expression of a second β-tubulin isotype

Asn100Ile in BenA eliminates a hydrogen bond between the carboxamide group of Asn100 and the carbonyl group at position 1 of plocabulin and probably perturbs hydrophobic interactions between the methyl group at position 27 of the drug and the BenA pocket including Asn10022 (Fig. 1a). Thus, BenAN100I directly affects drug binding, thereby conferring resistance to plocabulin at concentrations ranging from 5–50 µM (Fig. 1b,g)3. However, resistance is only partial, as manifested by reduced colony diameter and conidiation (production of asexual spores) displayed by the benAN100I strain growing on plocabulin-containing plates (Fig. 1b). Consistent with the fact that plocabulin and rhizoxin share their binding site on β-tubulin2,17, benAN100I shows cross-resistance to rhizoxin (not shown). In contrast, this mutant is hypersensitive to the MT depolymerizing agents benomyl (see below), and eribulin (Fig. 1b) that bind to different domains in β-tubulin18. These findings suggest that benAN100I substitution changes accessibility to the neighboring benomyl and vinca binding sites. Hypersensitivity to other anti-MT agents combined with rhizoxin resistance could represent a biomarker for plocabulin resistance mutations in β-tubulin. Anti-MT agents with a different binding site might represent an effective therapy in such cases.

Partial resistance of benAN100I to plocabulin might be due to additional cellular target(s). A. nidulans has a second β-tubulin, TubC, sharing ~85% amino acid sequence identity with BenA. While BenA, expressed during vegetative growth and conidiation, is essential, TubC, expressed mainly during conidiation, is dispensable for growth19,20,21,22. Nevertheless, these β-tubulin genes are interchangeable after promoter swapping19. Notably, conidiation on benomyl of strains with partial benomyl resistance-conferring benA mutations improves when tubC is ablated21, indicating that both β-tubulins are benomyl targets and that BenA can form both vegetative and conidiation microtubules when TubC is absent. TubC might also be a target for plocabulin, responsible for at least part of the benAN100I residual plocabulin sensitivity, since the amino acid residues forming the plocabulin binding site are conserved in TubC (Fig. 1c). To test this hypothesis, we deleted tubC in a wt background (Fig. 1d). The resulting tubCΔ was as sensitive to plocabulin as the wt (Fig. 1e,g). We next crossed tubCΔ and benAN100I and analyzed how progeny scored for plocabulin resistance. We identified three phenotypic classes: plocabulin-sensitive (resembling wt or tubCΔ strains), partially resistant (resembling benAN100I) and “super-resistant”, apparently unaffected by plocabulin (Fig. 1e). The “super-resistant” phenotype appeared with a frequency consistent with that expected for the double mutant benAN100I tubCΔ (n = 24, Chi-squared = 0.708, df = 2, two-tailed P > 0.5) (Fig. 1f). Indeed, DNA genotyping demonstrated that this plocabulin super-resistant class corresponds to the double mutant progeny. The benAN100I tubCΔ double mutant was insensitive to plocabulin over a range of concentrations (5–50 µM) (Fig. 1g). We conclude that: (1) plocabulin targets both β-tubulin isotypes in A. nidulans; (2) Asn100Ile abrogates the “maytansin-site” on BenA; (3) Although mutations affecting the plocabulin binding site on the major β-tubulin confer resistance to the drug, the presence and/or expression levels of minor β-tubulin isotypes can further modulate such resistance; (4) The sole plocabulin intracellular target affecting fungal growth and survival at the drug concentrations used is β-tubulin.

Mutations in genes other than tubulins confer resistance to plocabulin

We then sought for non-tubulin mutations that would confer plocabulin resistance in A. nidulans. We plated UV light-mutagenized conidia of A. nidulans on 10 µM plocabulin-containing medium. Resistant colonies were distinguishable over the background of marginally growing colonies (Supplementary Fig. S1). Plocabulin-resistant mutants were classified into five classes, denoted AR, CR, DR, ER and FR, according to their degree of resistance and their growth vigor (Fig. 2a) [a sixth resistant class from the same genetic screen, BR, had been previously attributed to the benAN100I mutation]3. In all five classes, resistance to plocabulin segregated as a single Mendelian trait in genetic crosses, indicating that resistance was caused by mutation in a single locus (Supplementary Fig. S1). Sequencing of tubA and benA ruled out the presence of mutations in the major α- and β- tubulin genes.

Figure 2
Figure 2

Phenotypic profiling of A. nidulans plocabulin resistant mutants. (a) Growth test of AR, FR, ER, CR, DR extra-tubulin mutants selected in a random genetic screen vs. a wt strain and the benAN100I-expressing strain (left panel) and of genetically engineered gcd7-1 gcn3Δ, gcn3Δ, gcn4Δ, benAN100I GCN3-1, benAN100I tubCΔ mutants strains (right panel) on medium with 5 or 10 µM or no (-) plocabulin. Strains were cultured for 2 or 3 days (2d or 3d) at 37 °C, 25 °C or 42 °C, as indicated. (b) Growth of the indicated strains at 37 °C on benomyl (1.4, 2.8 and 5 µM). Colonies were left to grow for the indicated time period. [Strains used: MAD5916 (AR), MAD4751(FR), MAD6098 (ER), MAD6095 (CR), MAD5959 (DR), MAD5321 (wt), MAD4736 (benAN100I), MAD5656 (benAN100I GCN3-1), MAD5877 (benAN100I tubCΔ), MAD6091 (gcn4Δ), MAD5386 (gcn3Δ), MAD5948 (gcd7-1 gcn3Δ), S1 Table].

We checked cross-resistance of the mutants to other anti-MT drugs. All showed resistance to rhizoxin (not shown). Notably, all mutants showed hyper-resistance to benomyl, which contrasts with the hypersensitivity shown by benAN100I (Fig. 2b). ER showed the highest resistance to benomyl and grew similarly on benomyl and plocabulin. AR showed only marginal benomyl resistance, but the highest plocabulin resistance.

Effect of resistance mutations on microtubule stability and mitotic progression in the presence of plocabulin

To test whether AR, CR, DR, ER and FR mutations modify the stability of interphase MTs and/or the mitotic defects caused by the drug, we introduced by genetic crosses GFP-TubA to label MTs and HhoA (H1 histone)-mCherry to label chromatin in the five mutant strains (see in ref.23 strain LO1915 for a description of these markers and ref.24 for the first simultaneous visualization of DNA and spindles in A. nidulans). The presence of GFP-tagged α-tubulin did not affect resistance to plocabulin (Supplementary Fig. S2).

Conidia were inoculated into inverted microscopy chambers containing liquid medium and incubated at 34 °C. Uninucleate A. nidulans conidiospores sediment at the bottom glass, break dormancy, establish a polarity axis and undergo growth by apical extension to form germlings that gradually give rise to hyphae (Fig. 3a). At 8 hours post-inoculation three rounds of mitosis have produced 8 nuclei and the first septum has been formed25,26,27. For both the wt and the mutant strains, hyphal length was in the order of hundreds of micrometers and hyphae contained multiple nuclei and well-discernible interphase microtubules at 15 hours post-inoculation (Fig. 3b). Including plocabulin in the growth medium prevented formation of long hyphae in the wt strain, which produced instead germlings with very short germ tubes [≈ 22 ± 12 µm (mean ± S.D.) N = 77)] and abnormal morphology (Fig. 3b). Most of these germlings had one or two nuclei (Fig. 3b), often containing elongated lumps of chromatin, suggesting defective chromosome segregation and mitotic blockade. Consistently, septae were rarely observed. All resistant mutants progressed beyond the wt in the presence of plocabulin (Fig. 3b): ER, CR and FR doubled in length the wt [≈45 ± 18 µm, ≈54 ± 20 µm and ≈55 ± 35 µm (mean ± S.D.), N = 23, 14 and 17; in unpaired t-test with Welch’s correction, P = <0.0001, <0.0001 and 0.0014, respectively] (measurements in FR were an underestimation because this mutant hyperbranched). AR hyphae grew up to hundreds of µm and also displayed numerous branches. Mitosis was at least partially restored in all resistant mutants, as indicated by the generation of septated hyphae that contained several nuclei per cell (Fig. 3b). Nevertheless, GFP-TubA was largely observed in a cytoplasmic haze, indicating extensive interphase MT depolymerization both in the wt and resistant cells, which suggested that the mitotic spindle rather than interphase MTs must be the growth-limiting physiological target of plocabulin in the wt.

Figure 3
Figure 3

wt and mutant strains’ growth, nuclei and microtubules in the presence of plocabulin. (a) A. nidulans asexual spores (conidia) are dormant and uninucleate. Conidia break dormancy soon after their inoculation in liquid medium in microscopy chambers and initiate isotropic growth (i), followed by the establishment of a polarity axis (ii), i.e. the selection of the plasma membrane spot towards which biosynthetic material will be directed leading to germ tube emergence, which takes place around the time when the first mitosis occurs (≈4 h at 37 °C) (iii). Growth continues exclusively by apical extension (iv). The first septum is normally laid after the third mitosis has taken place25 (v). As short germlings are being transformed into hyphae (with several multinucleated cellular compartments separated by septae) (vi), their apical extension rate increases, shifting from solely actin-dependent to both actin- and microtubule-dependent transport of exocytic vesicles. (b) After ≈15 h, wt spores have given rise to hundreds of µm-long hyphae with well discernible HhoA (histone 1)-mCherry-labelled nuclei and GFP-TubA (α tubulin)-labelled microtubules traversing the hyphal tip cell (upper left panel). Resistance mutations do not show any major effect on this growth pattern (not shown). After 15 h in 10 µM plocabulin, wt spores only form short germlings, that contain 1 to 2 nuclei and have no septae (upper middle panel), suggesting a block in mitosis. In contrast, resistant mutants form long hyphae that contain multiple nuclei and septae [upper right and lower panels; note that for the resistant strains only the hyphal tip cells are shown, while for the wt strain the entire germling can be seen]. Both in the wt and the mutants, interphase microtubules are largely depolymerized. (c) A large proportion of the FR population displays abnormally high numbers of nuclei and septae per compartment. C: conidiospore; A: apex; T: tip cell; S: septum; The dotted line in (a) indicates that only part of the hypha is depicted, green arrows indicate growth direction and rate, red circles represent nuclei. Single wavelength fluorescent microscopy images are inverted contrast maximal projections of z-stacks. All images are at the same magnification. Bar = 5 µm.

To discern differences, if any, in the stability of interphase MTs, we added 1 µM plocabulin [semi-restrictive in plates]3 to fast-growing hyphae pre-cultured without the drug. In the wt interphase MTs were rapidly depolymerized at both 27 °C and 34 °C (Fig. 4), resulting in a sharp increase in cytoplasmic GFP-α-tubulin fluorescence. GFP-TubA was also detected in abnormal spindles in metaphase-like nuclei, but these spindles did not elongate, indicating a blockade in metaphase (Supplementary Movie S1)24,28,29. In the AR mutant MTs were clearly resistant to the drug at 27 °C, but some depolymerization occurred at the tips such that MTs did not reach the apex (Fig. 4). Mitosis in the AR proceeded beyond metaphase (not shown). The FR mutant showed a similar phenotype, but MTs appeared less resistant to plocabulin than in AR (Fig. 4). In FR mutant hyphae nuclei blocked in prophase/metaphase or proceeding through anaphase were both observed (Supplementary Movie S2). In the CR mutant essentially no depolymerization of interphase MTs was observed (this mutant is cryosensitive and therefore it was tested at 34 °C) (Fig. 4). In the ER mutant, addition of plocabulin at 27 °C resulted in rapid MT depolymerization like in the wt, but MTs started recovering after 30–60 min, unlike the wt. Because ER resistance is slightly cryosensitive (Fig. 2a), we also tested ER at 34 °C. At this temperature, MTs depolymerized upon drug addition, but this response was not completely homogeneous in the population, indicating marginal resistance. Overall, our results showed that AR, CR, and FR displayed increased interphase MT stability in the presence of plocabulin, whereas ER had a weak effect. The increased interphase MT resistance in AR and FR, as well as the higher resistance displayed by AR MTs compared to FR MTs, correlated with growth on plocabulin-containing plates displayed by these two mutants (Fig. 2a). Notably, the CR mutant displayed the most stable interphase MTs, although on plates it displayed weak resistance, resembling that of the ER strain.

Figure 4
Figure 4

Effect of plocabulin on interphase microtubules of fast-growing plocabulin-resistant hyphae. Conidia from the wt and the indicated mutants were left to form fast-growing hyphae (15-18 h) (“Before”). Hyphae were then shifted to 1 µM plocabulin-containing medium and were observed 15 to 30 min after the shift. Note that α-tubulin becomes cytoplasmic in the wt and in the ER mutant, with the exception of some microtubule remnants (arrows), possibly reflecting bundles of microtubules stabilized by lateral interactions. Free tubulin is excluded from the nucleoplasm of interphase nuclei (n). MTs are largely resistant to plocabulin in the CR mutant. Interphase MTs are partially resistant in posterior areas (P) of the AR and FR hyphae, but depolymerize close to the hyphal apices (A). Images are sum projections of z-stacks modified with the unsharp mask filter of Metamorph. Bar = 5 µm.

Mutations in eIF2B that constitutively activate the general amino acid control pathway protect cells from plocabulin

CR and DR mutations conferred moderate-to-low resistance to 5–20 µM plocabulin (Fig. 2a). At 37 °C, CR and DR strains displayed reduced growth and formed sparse mycelium (Fig. 2a). Both were cryosensitive (Fig. 2a). Growth defects were more marked in CR. In both mutants reduced growth and cryosensitivity were linked to plocabulin resistance (0 recombinants out of 48 progeny).

We identified the CR mutation by combining genetic mapping with whole genome sequencing (WGS). We constructed a diploid strain combining CR with a tester strain carrying diagnostic markers in each chromosome and, subsequently, promoted chromosomal re-assortment by haploidization30. By phenotypic analysis of the haploid segregants we assigned the CR mutation to chromosome VIII (mitotic recombination is very rare in A. nidulans). We set up meiotic crosses between CR and strains carrying chromosome VIII markers. Linkage analysis suggested the CR mutation was located between the sE and nirA genes, closely linked to sE (Fig. 5a,b). This is a ≈230 kb-long DNA track containing more than 70 candidate open reading frames (ORFs). Sequencing the CR sE-nirA interval by WGS, we detected two adjacent transitions within the AN0167 ORF (g.493 C > T + g.494 T > C, resulting in the amino acid substitution Ser149Phe), a single nt deletion within an intron of AN0149, a single nt deletion in an intergenic region and multiple nt deletions within the first intron of AN10019 (Supplementary Table S3). The double transition in AN0167 lied at 9 kbp from sE, within the range of physical distance expected from recombination frequencies, unlike the non-coding changes, (Fig. 5c) and resulted in a non-conservative amino acid substitution in the encoded protein, strongly indicating that this mutation was causative of the CR plocabulin resistance.

Figure 5
Figure 5

Plocabulin resistance in the CR mutant is conferred by Ser149Phe amino acid substitution in the α subunit of eIF2B, Gcn3. (a) Haploidization analysis located the CR mutation on chromosome VIII. Using chromosome VIII markers in meiotic crosses, we detected linkage of the CR mutation with ureB3 (P = 0.018, recombination frequency = 36%), palB16 (P < 0.0001, recombination frequency = 25%) and chaA2 (P < 0.0001, recombination frequency = 19%), suggesting that CR was located centromere-distal to chaA. With a subsequent cross we detected 18, 2 and 17% recombination between the CR mutation and pantoA, sE and nirA respectively (progeny n = 96), indicating that the mutation lied between pantoA and nirA, closely linked to sE. (b) To determine whether CR is centromere distal or proximal relative to sE, we selected progeny in which recombination between the CR and sE loci had occurred. In 81% (n = 59) C/sE recombinants pantoA and sE had not recombined, indicating that sE lies between pantoA and CR. In agreement, 65% of the C/sE recombinants showed recombination between sE and nirA, suggesting that CR was between the sE [trxA73] and nirA. P1, P2: parental 1 or 2. (c) Sequencing of the CR genome (WGS) revealed that the only polymorphism between sE and nirA close enough to sE (<60 kbp) to account for the 2 cM sE/CR genetic distance was a double transition within the AN0167 ORF (GCN3-1) leading to Asn149Phe in the encoded protein, Gcn3. (d) We reconstructed the GCN3-1 mutation into an otherwise wt strain (GCN3-1 pyrG+) and confirmed that GCN3-1 suffices to confer the CR-characteristic plocabulin resistance and cryosensitivity. pyrG+ control transformants that had not incorporated GCN3-1 were neither cryosensitive, nor resistant to plocabulin. + and −: medium with/without plocabulin.

To formally demonstrate this contention, we reconstructed the mutation in a wt genetic background. A linear DNA fragment including the CR mutation was used to replace wt genomic AN0167 by homologous recombination31, after co-transformation of a pyrG89 (pyrimidine auxotrophy) strain with mutant AN0167 and wt pyrG DNA fragments to facilitate the selection of transformants (Supplementary Fig. S3). We obtained transformants that showed moderate-to-low resistance to plocabulin, displayed the CR characteristic growth defect, were cryosensitive and had replaced wt AN0167 by the mutant gene (Fig. 5d), confirming that g.493 C > T + g.494 T > C leading to Ser149Phe in AN0167 was the cause of these phenotypes.

AN0167 is homologue of yeast GCN3. We denoted the CR mutant allele GCN3-1 (see below). Gcn3 is a subunit of eIF2B, an oligomeric guanine nucleotide exchange factor (GEF) that activates translation initiation factor eIF2, a heterotrimer consisting of subunits α, β and γ. eIF2B promotes recycling of inactive eIF2-GDP released after translation initiation into active eIF2-GTP capable of initiating a new round of translation by delivering methionyl-tRNA to the ribosome32. Upon stress such as amino acid starvation, phosphorylation of eIF2α at Ser51 converts eIF2 into a competitive inhibitor of eIF2B, reducing the number of translation initiation complexes and leading to a reduction in general translation33,34. This, in turn, elicits global changes in gene expression by facilitating translation of certain mRNAs, such as those of transcription factors Gcn4 (in yeast) and ATF4 (in mammals) that carry short upstream ORFs hindering translation of the main ORF in non-stress conditions33,34. This stress response is conserved between fungi and mammals [General Amino Acid Control (GAAC) in yeast, Cross-Pathway Control (CPC) in Aspergillus and Neurospora or Integrative Stress Response (ISR) in mammals].

eIF2B is a heterodecamer containing two sets of α, β, γ, δ and ε subunits (yeast Gcn3, Gcd7, Gcd1, Gcd2, Gcd6, respectively), where α, β and δ subunits form the regulatory and γ and ε the catalytic sub-complex35,36,37. Tight binding of P-eIF2α into a pocket formed by the regulatory subunits obstructs interaction of eIF2γ (the guanine nucleotide binding subunit in eIF2) with an Asn-Phe motif in eIF2Bε, impeding GDP/GTP exchange on eIF238,39. Genetic studies have shown that GCN3 encodes a positive regulator of GAAC. Recessive gcn3 mutations confer sensitivity to amino acid starvation (Gcn phenotype, general control non-inducible). GCD genes encode negative-acting subunits, with gcd partial loss-of-function mutations leading to constitutive GAAC activation and increased resistance to amino acid starvation (Gcd phenotype, general control derepressed)33,40,41.

To test whether A. nidulans GCN3-1 is a loss-of-function mutation, we compared it to a gcn3∆ mutant lacking the complete AN0167 ORF [yeast GCN3 is not essential]40,42 (Fig. 6a). A. nidulans gcn3Δ grew like the wt, was not cryosensitive and did not display plocabulin resistance, unlike GCN3-1, suggesting that GCN3-1 is not a loss-of-function mutation (Fig. 2a,d). We next tested GCN3-1 and gcn3Δ sensitivity to 3-amino-1,2,4-triazol (3-AT), a competitive inhibitor of histidine biosynthesis that causes His starvation, inducing the GAAC pathway. Both the wt and GCN3-1 were resistant to 2.5 mM 3-AT, whereas gcn3Δ was hypersensitive (Fig. 6b), suggesting that at this 3-AT concentration the GAAC pathway was activated in a gcn3-dependent way and was necessary for survival in the wt. Doubling 3-AT concentration to 5 mM abolished wt but not GCN3-1 growth (Fig. 6b), strongly indicating that GCN3-1 is a gain-of-function allele resulting in hyper-activation of GAAC.

Figure 6
Figure 6

Impact of GCN3-1 (CR) and gcd7-1 (DR) on eIF2B and the General Amino Acid Control pathway. (a) Strategy used to delete gcn3, gcn4 and gcnE ORFs by replacing them with pyrGAf (Fig. 1D legend). (b) Growth of the wt (MAD5321) and CR (GCN3-1) (MAD6095), DR (gcd7-1) (MAD4740) and gcn3Δ (MAD5386) mutants on 3-aminotriazol, an inhibitor of histidine biosynthesis that induces the GAAC. (c) Surface representation of the S. pombe eIF2B oligomer [oriented as in Fig. 2, ref.35]. Different subunits are indicated with Greek letters. Regulatory subunits α, β and δ form a central cavity thought to accommodate eIF2α. Several residues that are cross-linked to eIF2α in the eIF2B::eIF2α complex are highlighted in orange. Yellow labelling indicates subunit β Tyr377 whose hydroxyl group forms a hydrogen bond with ND2 of Asn460 in δ (see enlargement). β-Tyr377 is the equivalent of Tyr417 in A. nidulans. gcd7-1 (DR) results in Tyr417Asn substitution, thereby disrupting the βTyr377 - δAsn460 hydrogen bond and in all likelihood affecting the position of δ Asn460 that contributes to the base of eIF2α binding cavity. (d) gcn3Δ has a slight effect on the plocabulin resistance conferred by gcd7-1, but restores normal (wt) resistance to 3-AT, possibly counteracting the effect of gcd7-1 on the GAAC pathway (see text) (gcn3Δ gcd7-1 is MAD5948, gcd7-1 is MAD5950, S1 Table). (e) wt and gcn4Δ growing on sub-lethal concentration of plocabulin. gcn4Δ results in hypersensitivity to plocabulin. (f) Deletion of the histone acetyltransferase gcnE impairs growth, but it does not modify the plocabulin resistance conferred by GCN3-1. (g) Both single gcnEΔ and double gcnEΔ GCN3-1 mutant strains show resistance to 5 and 7.5 mM 3-AT, where wt growth is inhibited. (h) benAN100I and GCN3-1 have an additive effect on plocabulin resistance.

As the DR mutant bore phenotypic similarities to CR (GCN3-1), we also tested its response to 3-AT. DR was resistant to 5 mM 3-AT as was GCN3-1 (Fig. 6b), suggesting that it also results in hyper-activation of GAAC. Prompted by this observation, we sequenced the genes for all regulatory subunits of eIF2B (Gcn3/AN0167/eIF2Bα, Gcd7/AN1344/eIF2Bβ, Gcd2/AN6864/eIF2Bδ) in the DR strain and, indeed, detected a g. 1408 T > A substitution in AN1344, resulting in Tyr417Asn in the A. nidulans Gcd7 homologue. Genetic crosses showed that the mutant gcd7 allele, gcd7-1, is linked to plocabulin resistance.

Since the DR mutant strain displays 3-AT resistance, i.e. a Gcd phenotype, the causative Tyr417Asn substitution must lead to Gcd7 loss of function. In yeast, substitution in Gcd7/eIF2Bβ of the equivalent Phe360 (see Supplementary Fig. S4 alignment of A. nidulans, S. cerevisiae, Homo sapiens and Schizosaccharomyces pombe eIF2B α and β) and of the neighboring amino acid residues Pro358 and Ser359 by Ala is lethal43 and reduces co-immunoprecipitation of eIF2 with eIF2B, suggesting that a critical function of Gcd7 is stabilizing the eIF2::eIF2B interaction43. Indeed, the atomic structure of S. pombe eIF2B complex showed that Gcd7 Tyr377 (equivalent to Tyr41λn A. nidulans and Phe360 in S. cerevisiae) is located in the interface between regulatory subunits, within the central cavity that mediates binding to eIF2/eIF2α-P35 (Fig. 6c).

The observation that Gcd mutations in different eIF2B subunits conferred plocabulin resistance suggested that resistance was mediated by the upregulation of the GAAC/CPC pathway. We asked whether GCN3-1 and gcd7-1 would have a synergistic effect. However, from the cross between GCN3-1 and gcd7-1 we recovered three phenotypic classes in 1:1:1 ratio, corresponding to wt and to the single GCN3-1 and gcd7-1 mutants, which indicated that ascospores of the double mutant were unviable (see Materials and Methods for details). In yeast, the combination of constitutively de-repressing GCN3 and gcd2 alleles also leads to negative synthetic effects44. Synthetic lethality would be consistent with both GCN3-1 and gcd7-1 diminishing eIF2B GEF function by affecting independent molecular interactions within the eIF2B::eIF2 supercomplex.

We next asked whether gcn3Δ would modify gcd7-1 resistance to plocabulin. gcn3Δ improved growth of gcd7-1 at 37 °C and suppressed its cryosensitivity (Figs 2a and 6d). The double mutant was instead thermosensitive at 42 °C (Fig. 6d) and, unlike the single mutants, showed normal sensitivity to 3-AT (Fig. 6d). 3-AT normosensitivity may be explained as an intermediate phenotype between hypersensitivity of gcn3Δ and hyper-resistance of gcd7-1. Therefore, by reducing or eliminating the GAAC stress response, gcn3Δ may counteract GAAC de-repression mediated by gcd7-1. At the same time, general translation levels would increase in the double mutant compared to the single gcd7-1 mutant, which may explain growth restoration at both optimal and lower temperatures. Synthetic thermosensitivity, on the other hand, may result from increased instability of the double mutant eIF2B. Interestingly, resistance to plocabulin is only weakly decreased in the double gcn3Δ gcd7-1 mutant with respect to single gcd7-1 (Figs 2a and 6d), indicating that 3-AT normosensitivity is not diagnostic for plocabulin sensitivity.

We next tested whether components of the GAAC pathway acting downstream of eIF2B are required for plocabulin resistance in eIF2B mutants. We first deleted gcn4/cpcA45. gcn4Δ and wt strains were inoculated on medium containing 2.5 (Fig. 6e), 5 or 10 µM (Fig. 2a) plocabulin and incubated until the wt (whose growth was strongly inhibited) formed a microcolony. gcn4Δ showed hypersensitivity, which was very clear in the 2.5 µM sub-lethal plocabulin concentration (Fig. 6e), suggesting that Gcn4 counteracts plocabulin toxicity. To test whether GCN3-1 resistance also requires Gcn4, we set a genetic cross between GCN3-1 and gcn4Δ strains. However, we did not recover double mutant progeny, despite that GCN3-1 and gcn4Δ parental mutant and recombinant wt progeny appeared at the expected ratios, indicating that the double mutant progeny is unviable. Gcn4-dependent gene expression would then be essential for survival when general translation is reduced by GCN3-1.

We next ablated Gcn5/GcnE, a SAGA complex histone acetyltransferase46, necessary in S. cerevisiae for normal transcriptional activation by Gcn447. A. nidulans SAGA complex mediates induction of some gene clusters but represses others48,49. GcnE is dispensable for growth, but necessary for conidiation49,50. Indeed, deletion of the entire gcnE ORF greatly impaired but did not abolish hyphal growth, whereas it abolished conidiation (Fig. 6f). gcnEΔ was slightly hypersensitive to 5 µM plocabulin (Fig. 6f), possibly reflecting the mutant’s growth defect. By meiotic crossing, we obtained GCN3-1 gcnEΔ double mutants, which grew similarly to gcnEΔ (Fig. 6f) and were additionally cryosensitive, like GCN3-1 (not shown). Although to a lesser extent than GCN3-1 single mutant, the double mutant was still resistant to plocabulin (Fig. 6f), indicating that resistance in GCN3-1 is largely GcnE independent. Unexpectedly, gcnEΔ (and GCN3-1 gcnEΔ) mutants were resistant to 5 mM 3-AT (Fig. 6g), indicating diverged roles of A. nidulans GcnE and yeast Gcn5.

Since interphase MTs of the CR mutant displayed the strongest resistance to plocabulin (Fig. 4), we asked whether benAN100I and GCN3-1 would have a synthetic effect. Notably, benAN100I and GCN3-1 showed a strong additive effect on plocabulin resistance, the double mutant being almost insensitive to the drug (Figs 2a and 6h). As the residual plocabulin sensitivity of benAN100I depends on TubC (see above), it is possible that GCN3-1 modifies the role that TubC plays in the susceptibility to plocabulin by counteracting, directly or indirectly, MT depolymerization or by altering the TubC/BenA expression profile.

Substitution of a TIM44 amino acid residue critical for mitochondrial protein import reduces growth-inhibitory effect of plocabulin

The ER mutation conferred moderate-to-low resistance to 5–50 µM plocabulin at 37 °C (Fig. 2a). ER resistance was cryosensitive, whereas ER growth was slightly thermosensitive and the ER strain was defective in conidiation (Fig. 2a). These phenotypes were linked to resistance (no recombinants were obtained out of 48 progeny).

To identify the causative mutation, we assigned ER to chromosome VIII by haploidization analysis. We then identified nucleotide polymorphisms on chromosome VIII by WGS. ≈250 polymorphisms were rejected because they also appeared in the parental strain used for mutagenesis or in the AR and FR mutants (AR and FR map to different linkage groups). Five candidate polymorphisms were ER-specific (indicated with lower case roman numerals in Supplementary Table S4).

Meiotic crossing detected linkage of ER to choC. Among the ER-specific polymorphisms, two (iii and iv, Fig. 7a) were physically linked to choC. We pooled PCR-generated DNA fragments of these two polymorphic loci from nine ER (resistant) progeny of an outcross between an ER strain (iii, iv) and the wt (III, IV) and sequenced them. Only the mutant nucleotide peaks were present in the sequencing chromatographs, indicating co-segregation of polymorphisms iii and iv with resistance, and implying that one (or potentially both) of them was causative. We next exploited linkage between loci iii and iv and landmarks in the region (choC-46 kb-III-220 kb-IV) to identify the causative mutation. We performed a three-point-cross between choC + iii iv and choC- III IV parental strains and selected for progeny with recombinant plocabulin resistance and choline auxotrophy traits (i.e. choline-independent, ES and choline-dependent, ER progeny) (Fig. 7b). Should the causative mutation have been iv, progeny arising from a single cross-over between choC and the ER-specific mutation that were plocabulin sensitive would have had a higher probability of carrying iii and IV (crossover 1, Fig. 7b) than III and IV (crossover 2). On the contrary, should the causative mutation have been iii, the only way to get plocabulin sensitive strains would have been the rare crossover 2 between choC and iii, leading to progeny carrying III and IV. Sequencing of three recombinant plocabulin sensitive progeny demonstrated that they carry iii and IV (i.e. it had arisen from crossover 1) (Fig. 7c), showing that polymorphism iii does not confer plocabulin resistance and indicating that polymorphism iv is instead causative. Sequencing of one plocabulin-resistant recombinant showed that it carried III and iv (i.e. the reciprocal progeny of crossover 1) (Fig. 7c), indicating that polymorphism iv alone is sufficient to confer ER resistance to plocabulin.

Figure 7
Figure 7

ER plocabulin resistance is due to TIM44R263C affecting mitochondrial protein import. (a) Position on chromosome VIII of polymorphisms i through v identified in the ER genome. iii and iv are linked to choC, as is plocabulin resistance. (b) iii and iv are closely linked and both co-segregated with resistance. A three-point-cross between a choline prototroph (choC + ) ER and a choline auxotroph (choC-) wt (ES, plocabulin sensitive) was used to determine the resistance-causative polymorphism. We selected progeny with recombined characteristics (choline auxotrophs/plocabulin resistant or choline prototrophs/plocabulin sensitive), i.e. recombination events within the choC/IV 260 kbp region. As distance between III and IV (220 kbp) is much larger than between choC and III, recombinants originating from a crossover in the III to IV interval (crossover 1) are more frequent, but they would only appear if they satisfied the selection criteria, i.e. if polymorphism iv were conferring resistance. If instead polymorphism iii were required for plocabulin resistance, then selected progeny would originate from a crossover within the choC/III region (crossover 2). (c) Three out of three choline prototroph, plocabulin sensitive progeny had originated from crossover (1), showing that polymorphism iii is not sufficient for resistance. A choline auxotroph, plocabulin resistant strain had originated from crossover (1), showing that polymorphism iv suffices to confer resistance. (d) By co-transformation of a pyrimidine prototroph strain with pyrG and a DNA fragment carrying polymorphism iv and by selection for pyrimidine prototrophy on plocabulin, we obtained 21 colonies, of which one (arrow, upper panel) showed plocabulin resistance after purification (arrow, lower panel). This transformant displayed an ER-like growth defect and had incorporated polymorphism iv into AN1281/tim44.

Polymorphism iv is a missense mutation in AN1281. We reconstructed it by genetic engineering in a wt background, using a mutant linear DNA AN1281 fragment containing iv to replace the wt genomic DNA, co-selecting for a transformation marker (pyrG) and plocabulin resistance (Fig. 7d). We were able to recover one pyrimidine prototroph transformant that was plocabulin-resistant, which we showed to carry the correct gene replacement. This strain displayed all ER characteristic phenotypes: plocabulin resistance, thermosensitivity and deficient conidiation (Fig. 7e), thus establishing that polymorphism iv confers resistance.

AN1281 codes for a conserved and essential component of the Presequence translocase-Associated Motor (PAM) complex, Tim44. PAM is recruited to the inner mitochondrial membrane translocation channel, TIM23, to complete translocation of mitochondrial matrix proteins51. Tim44 binds the imported protein precursor as it emerges from the TIM23 channel and recruits Hsp70 (yeast Ssc1)-ATP. Hsp70-mediated ATP hydrolysis, stimulated by PAM18, leads to tight binding of Hsp70 to the precursor protein and to subsequent release of the Hsp70::precursor protein complex from Tim44 into the matrix. Another molecule of Hsp70-ATP can then associate with Tim44 at the translocon exit to initiate a new cycle of ATP hydrolysis >precursor binding >matrix release, such that the precursor protein is being transported in a rachet-like fashion [reviewed in]51. Tim44 has an N-terminal domain contacting Tim23, Hsp70 and PAM16-PAM18 and a C-terminal domain contacting Tim17 and the translocating precursor proteins; it has been proposed that rearrangements of the two domains, possibly triggered by incoming protein traffic, drive protein translocation52.

The ER mutation leads to the Arg263Cys substitution within the N-terminal domain of AN1281/Tim44 (Supplementary Fig. S5). The equivalent amino acid residue in S. cerevisiae Tim44, Arg180, is critical for the polypeptide substrate-induced destabilization of the Ssc1(Hsp70)-Tim44 complex and for the association of Tim44 with the TIM23 translocase53. Arg180Ala-Tim44 in yeast results in growth defects53, in line with the conidiation and growth defect displayed by Arg263Cys-Tim44 A. nidulans. Given the conservation of Tim44 homologues, Arg263Cys may have similar effects on A. nidulans Tim44 to those caused by Arg180Ala on the S. cerevisiae protein, suggesting that an altered mitochondrial protein import would confer resistance to plocabulin.

Substitutions in the middle homology region of two binuclear zinc cluster transcription factors confer resistance to plocabulin

The AR mutant conferred the highest resistance to 5–50 µM plocabulin at 37 °C among the mutants obtained from the genetic screen (Fig. 2a). AR showed a slight conidiation defect at 37 °C, exacerbated at 25 °C. AR resistance to plocabulin was partially thermosensitive (Fig. 2a).

Using WGS and comparing the mutant genome with that of the strain used for mutagenesis, we identified 21 polymorphisms in AR, potentially resulting in missense or truncating mutations. By haploidization analysis, we assigned the plocabulin resistance mutation to chromosome II, which contained three of these mutations, affecting AN3969, AN4084 and AN8292 (Supplementary Table S5).

We crossed the AR with a wt strain and checked if any of the candidate mutations co-segregated with plocabulin resistance. To streamline this analysis, we PCR-amplified each of the three candidate genetic loci from 12 plocabulin-resistant progeny, mixed equal quantities of the 12 individual PCR products for each locus and sequenced the resulting mixtures. Overlapping wt and mutant nucleotide peaks were identified in the positions of the AN4084 and AN8292 polymorphisms in the corresponding sequencing chromatograms (Fig. 8a). On the contrary, a single peak was observed at the position of the AN3969 polymorphism and this peak corresponded to the nucleotide identified in the AR mutant (Fig. 8a). To buttress this result, we then sequenced a pool of AN3969 DNAs from 12 plocabulin-sensitive progeny. A single peak in the chromatogram (Fig. 8a) indicated that sensitive strains were also homogeneous with regard to that nucleotide position, and showed that the identified nucleotide in this case corresponded to the wt AN3969 allele. These results strongly indicated the AN3969 polymorphism was causative of the AR plocabulin resistance.

Figure 8
Figure 8

AR and FR resistance is conferred by mutations in the middle homology region of Zn2Cys6 binuclear zinc cluster transcription factors. (a) Sequencing chromatographs of PCR-amplified fragments of the indicated genes from a wt and an AR strain and from pools of 12 plocabulin resistant or sensitive progeny from a cross between the wt and the AR. Note, in the mix from resistant strains, the double peaks for genes AN8292 and AN4084 in the position (indicated by an asterisk) of polymorphisms found in the AR strain (MAD4738) by WGS, indicating that these polymorphisms do not co-segregate with resistance. Note the single peak observed instead for polymorphism in AN3969 from resistant or sensitive progeny pools, with the nucleotide in the resistant population corresponding to the polymorphism found in the AR while that from the sensitive population to the wt sequence. (b) Strategy to reconstruct polymorphism AN3969 in a wt recipient strain (left) and direct selection for plocabulin-resistance after transformation (right). Colonies that have incorporated the mutation are large and produce green conidia (yellow arrows) on plocabulin. (c) Transformants were indistinguishable from the original AR strain, confirming that AN3969 polymorhism confers the AR phenotype. (d) AN3969 gene model based on amplification and sequencing of cDNA with the DNA primers indicated in the scheme (S2 Table). The start codon is uncertain. Exon coordinates on chromosome II: (exon 1) 2.476.638-2.476.570, (2) 2.476.510-2.476.487, (3) 2.476.423-2.476.323, (4) 2.476.259-2.475.754, (5) 2.475.693-2.475.359, (6) 2.475.261-2.474.800, (7) 2.474.719-2.474.478, (8) 2.474.423-2.474.204. (e) AN3969/PloA architecture: Tyr235Asp confers plocabulin resistance. The zinc binuclear cluster domain [Zn(II)2Cys6], followed by a coiled-coil domain, as well as the MHR regulatory region, are characteristics of this family of transcription factors. (f) Strategy for deletion of the AN3969/ploA ORF and sensitivity to plocabulin displayed by AN3969Δ (Δ). (g) Anti-GFP western blot of C-terminally tagged AN3969wt and AN3969Y235D with GFP. (h) AN8345/PloF domain organization. (i) Strategy for reconstructing AN8345 polymorphism by in locus integration of a DNA fragment with the mutation, followed by pyrGAf. Double homologous recombination results in pyrimidine prototrophy when a crossover in the 3′UTR region is combined with either crossover 1 (incorporating the polymorphism in the genome) or crossover 2 (without polymorphism incorporation). (j) Two types of transformants were obtained by transformation with the PloF DNA fragment: Type I had incorporated the mutation, grew compact and displayed plocabulin resistance, although less than the original FR mutant. Type II that had not incorporated the polymorphism, grew as the wt and was sensitive to plocabulin. This showed that Ile505Asn in AN8345 confers plocabulin resistance.

To demonstrate this contention, we re-introduced the AN3969 mutation into a wt strain by transformation, directly selecting transformants on plocabulin. Large conidiating colonies could be singled out over the background of non-transformed cells (Fig. 8b). After purifying these colonies, we verified that they had incorporated the relevant AN3969 mutation, and confirmed that their growth with or without the drug was indistinguishable from that of the original AR strain (Fig. 8c). This showed that AN3969 polymorphism confers the AR plocabulin resistance. We denoted AN3969, thus far uncharacterized in fungi, ploA (plocabulin resistant class A)

PloA is predicted to include a fungal transcription factor (TF) regulatory middle homology region (MHR), but, according to the genomic database (www.aspgd.org), it would lack the N-terminal binuclear zinc cluster DNA-binding domain normally associated with the MHR54. By PCR amplification from a total cDNA pool, we found that AN3969 encodes an mRNA much larger than predicted. Figure 8d shows a corrected gene structure for ploA, as deduced from cDNA sequencing. Once corrected, ploA includes coding information for a canonical fungal-specific GAL4-like Zn(II)2Cys6 binuclear cluster DNA-binding domain55. PloA is predicted to have coiled-coil domains in addition to at least one transmembrane domain towards its C-terminus (Fig. 8e). The Zn(II)2Cys6 binuclear cluster protein family is composed of fungal-specific TFs playing multiple regulatory roles including in multi-drug resistance (yeast PDR1/3), sterol sensing and regulation of sterol biosynthetic genes (ECM22/UPC2), mitosis (kinetochore protein CEP3) and chromatin remodeling56. Their activity is modulated by small molecules such as chemicals, nutrients and metabolites57. According to the most probable PloA N-terminus based on cDNA sequencing (Fig. 8d), class A mutation c. 703 T > G results in Tyr235Asp substitution within the putative regulatory PloA MHR.

We deleted ploA to test whether loss of function would confer plocabulin resistance. ploA∆ grew normally and did not show resistance to plocabulin (Fig. 8f), suggesting that Tyr235Asp leads to gain- or modification-of-function. This conclusion is in line with previous reports indicating that the MHR domain plays an inhibitory role in the transcriptional activity of zinc cluster proteins56.

To gain insight into the function of PloA we investigated its localization after GFP tagging. If the protein fusion were functional, GFP-tagged Tyr235Asp-PloA should confer plocabulin resistance. Instead N-terminal GFP tagging abolished resistance. C-terminally tagged Tyr223Asp-AN3969 permitted partial plocabulin resistance (not shown). Anti-GFP western blotting revealed the presence of a faint band showing a slightly slower mobility than the expected from the fusion protein size (Fig. 8g). Hardly any fluorescence was detectable with the wt PloA-GFP protein, while a minority of Tyr223Asp-PloA-GFP hyphae showed clustered foci of fluorescence with a subcellular distribution that would be consistent with nuclear association (not shown).

A second zinc cluster family protein appeared to be responsible for the moderately resistant FR mutant. Plocabulin resistance in FR was partially thermosensitive (Fig. 2a). Chromosomal allocation localized the FR mutation on chromosome V, where three potentially causative polymorphisms in genes AN8345, AN12394 and AN5786 were identified by WGS (Supplementary Table S6). A genetic cross between the FR and a wt strain, followed by sequencing of candidate loci in plocabulin resistant and sensitive progeny showed that polymorphism AN8345 co-segregated with resistance, unlike polymorphisms in AN12394 and AN5786, suggesting that a mutant AN8345 conferred FR resistance.

According to the A. nidulans genome sequence annotation, the FR AN8345 polymorphism g. 1885 T > A lies in the middle of an abnormally large, 138 bp-long seventh intron. Inspection of RNAseq data unambiguously showed that the intron acceptor site lies upstream from the predicted site, producing a more typical 66 bp intron. Corrected accordingly, AN8345 product has 24 additional amino acid residues, encoded by the beginning of exon 8, which would be part of the AN8345 MHR domain. In this corrected gene model, the FR polymorphism would fall in exon 8, resulting in substitution Ile505Asn within the MHR (Fig. 8h).

We were unable to reconstruct the FR mutation by transformation and direct selection on plocabulin or by co-transformation with pyrG DNA and selection for pyrimidine prototrophy, like in the cases of PloA/AR, Tim44/ER and Gcn3/CR. Therefore, we constructed a linear DNA molecule consisting of a C-terminal AN8345 fragment containing the FR mutation, A. fumigatus pyrG and AN8345 3′UTR. The presence of pyrGAf permits selection of the desired gene replacement by a double crossover in the AN8345 locus (Fig. 8i). Depending on the position of the N-proximal crossover the FR mutation is incorporated in a proportion of gene replacement events. We obtained two types of transformants with regard to colony growth: wt, which were sensitive to plocabulin, and compact, which showed substantial plocabulin resistance, albeit weaker that the original FR mutant (Fig. 8j). We verified by sequencing that plocabulin-resistant strains had incorporated the FR mutation, while plocabulin-sensitive strains had instead the wt AN8345 allele. Both lower levels of resistance and compact growth may result from insertion of pyrG immediately downstream the AN8345 ORF, which removes the ‘natural’ 3′UTR from AN8345 transcript and may interfere with the promoter of the next gene, AN12339. Thus, with the caveat that we could only partially recapitulate the resistance displayed by FR, the above experiments strongly indicate that AN8345 polymorphism is causative of the FR plocabulin resistance.

We denoted the uncharacterized AN8345 gene as ploF. ploF codes for a protein containing a Zn(II)Cys6 binuclear cluster and the associated MHR (Fig. 8h). Interestingly, the divergently transcribed gene AN8344, located upstream from AN8345 (≈ 400 bp separate the two start codons) codes for a homologue of PDR15, a plasma membrane multidrug transporter of the ATP binding cassette (ABC)-family that is regulated by zinc cluster TFs in response to stress58. We observed a similar arrangement of an ABC transporter and a zinc cluster protein in AN8345 homologues from Aspergillus oryzae and Aspergillus niger, suggesting that the two genes are under common regulation, which would implicate the FR Zn(II)Cys6 binuclear cluster TF in multidrug resistance responses including the expression of multidrug efflux pumps.

Discussion

Fungi are a sister taxonomic group to metazoa. Therefore, many antitumor compounds have concomitant antifungal activity, which is also the case for plocabulin. Given the importance of MTs for filamentous fungal cells, and the many precedents in MT discovery in filamentous fungi, we reasoned that mutants of A. nidulans resistant to plocabulin would not only provide information relevant to its MoA, but they would also help define cellular pathways by which tumor cells might be able to overcome the anti-proliferative activity of plocabulin. Using the Aspergillus genetic toolbox, we established causative relationship between mutations in several genes and resistance to plocabulin.

A major achievement was the generation of a completely resistant mutant by combining a binding site-disrupting mutation in the major β-tubulin with deletion of the minor β-tubulin gene, such that all cellular β-tubulin was of the “resistant-type”, BenAN100I. This positive result showed that, up to the maximum concentration used, plocabulin has a unique molecular target relevant to fungal growth, β-tubulin, and underlined the fact that in fungi the available tubulin isotypes co-determine the response of microtubules to the drug. Knowing the spectrum of their molecular targets is crucial to comprehend and regulate the drug effects at the cellular and organismal level. One case in support of this statement is the shift in the perception of the MoA of anti-MT compounds59,60. Although it has long been thought that anti-MT compounds kill cells by blocking mitosis, human tumors divide rather slowly and the primary toxicity associated with the administration of anti-MT agents, neurotoxicity, concerns non-dividing cells60. This is considered as evidence that anti-MT compounds induce cell death independently of mitosis; it has been proposed that they attack interphase microtubules instead60. However the need to confidently rule out the existence of secondary drug targets becomes most apparent. Ruling out secondary targets was important to assess the mechanistic basis of resistance conferred by the non-tubulin A. nidulans mutations: if only one plocabulin target is assumed, these mutations must affect drug availability, microtubule dynamics or the consequences of depolymerizing microtubules.

Another mutant that was nearly insensitive to the drug, although at the expense of debilitated growth, was the double benAN100I GCN3-1 mutant. benAN100I GCN3-1 resistance to plocabulin suggested that GCN3-1 diminishes the sensitizing effect that the minor β-tubulin isotype, TubC, contributes to the response to plocabulin. GCN3-1 is the mutant with the most plocabulin-resistant interphase MTs, despite that GCN3-1 colony formation on plocabulin was more debilitated than in other mutants. That drug-resistant interphase MTs do not directly translate into resistant growth suggests that MT resistance in GCN3-1 cannot be solely attributed to a reduction in intracellular drug concentration, as this would restore growth proportionately. Instead, GCN3-1 may have an effect directly on MTs. A plausible explanation would be that GCN3-1 triggers an increase in BenA expression that would predictably increase the drug concentration threshold needed to overrun tubulin and that may cause growth defects related to abnormal MT stability. Such hypothetical increase in BenA expression, if combined with an inefficient plocabulin binding site on BenA, as we demonstrate to be the case for benAN100I, might give rise to complete plocabulin resistance.

The finding that mutations in subunits α and β of eIF2B that display a Gcd phenotype (constitutive GAAC activation) confer plocabulin resistance is noteworthy. Constitutive mutations in GCN3 have been previously isolated in S. cerevisiae44 and most of them appear to interfere with interactions between the N- and the C-terminal domains of eIF2Bα61. Ser149Phe substitution in GCN3-1 is a novel gain of function mutation in eIF2Bα. The conserved human eIF2Bα-Ser131 is positioned at the center of a positively charged inter-domain pocket61. eIF2Bβ Tyr417 that confers plocabulin resistance when substituted by Asn in gcd7-1 appears to be located at the interface of eIF2B subunits β and δ, within the central cavity accommodating eIF2α35. Its interaction with Asn460 of subunit δ might be important for maintaining the bedrock of this central cavity (see Fig. 6c). The isolation of GAAC pathway-activating mutations in two different eIF2B subunits strongly argues for a causative role of GAAC activation and/or general translation reduction in plocabulin resistance. However, the double gcd7-1 gcn3Δ mutant nears the plocabulin resistance of gcd7-1, although it presents normal sensitivity to 3-AT, unlike its constituents, gcd7-1 (hyper-resistant) and gcn3Δ (hyper-sensitive). Would this indicate that GAAC activation and plocabulin resistance are separable? It is likely that, despite the apparent normosensitivity to 3-AT, the GAAC would still be constitutively activated in the double mutant background, possibly because eIF2α would not be optimally accommodated by eIF2B due to the presence of Gcd7(Tyr417Asn). Loss of Gcn3 from the Gcd7(Tyr417Asn)-containing eIF2B may however lead to lower constitutive activation, which would explain the remediation in the double mutant of the growth defect and cryosensitivity displayed by the single gcd7-1 mutant. Upon stimulus, there would be no eIF2α-P-mediated induction of the GAAC in the gcd7-1 gcn3Δ mutant, perhaps because enhanced interactions between the regulatory eIF2B subunit and eIF2α-P, thought to prevent catalytically productive eIF2B/eIF2 interaction33, depend on Gcn3. Thus, although the apparent response to starvation of the double mutant resembles the wt response, resistance to 2.5 mM 3-AT would be independent of eIF2α phosphorylation, unlike in the wt. It is plausible that this low, but constitutive GAAC activation seen in the double gcn3Δ gcd7-1 mutant causes plocabulin resistance under non-starvation conditions. Whether other GAAC-activating eIF2B mutations would confer plocabulin resistance merits further examination. It is worth noting that features of the solid tumor microenvironment such as hypoxia and nutrient deprivation have been associated with ATF4 overexpression regulated by the PERK/GCN2-eIF2α pathways and eIF2α phosphorylation has been shown to promote tumor growth62,63. It would thus appear that the ISR pathway may be constitutively activated in some tumors and it would be important to understand if and when this activation may be relevant for the response in therapeutic treatments with anti-MT agents.

We also identified Arg263Cys-Tim44 as causative of plocabulin resistance. Judging by the effect of analogous mutations in yeast homologues, this substitution most likely affects a critical function of TIM44 in pre-protein translocation into the mitochondrial matrix. In agreement, we found that TIM44R263C expression in a strain co-expressing GFP-tagged Tom20 (an outer mitochondrial membrane translocase component) leads to a strong synthetic negative effect on growth (not shown). Identifying that mitochondrial function may affect plocabulin resistance could be important, as the efficiency of anti-MT agents has been linked to the apoptotic mitochondrial pathway59,64. The MT-mitochondria relationship appears also important for chemotherapy-induced peripheral neuropathy, as atypical mitochondria have been suspected to underlie neuron dysfunctions after treatment with MT targeting agents64.

Some of the non-tubulin mutations resulting in plocabulin resistance confer similar cross-resistance levels to benomyl, another anti-MT agent, suggesting they might target cellular “escape doors” downstream from the imposed MT depolymerization problem. These mutations would be particularly relevant in therapeutic treatments because they might provide resistance to a broad range of anti-microtubule agents. In fact, it is plausible that non-tubulin resistance mutations would be the primary therapeutic concern, as the presence of several β-tubulin isotypes in human cells would minimize the effect of mutations altering the drug-binding site of one particular β-tubulin isotype.

Finally, the finding that two TFs of the binuclear zinc cluster family are involved in plocabulin resistance was unsurprising, given the importance of these factors in multidrug resistance. The mutations obtained fall within the MHR region, whose deletion/mutation renders this type of TFs constitutively active56. Of note, at least the AN3969/ploA-associated mutation leads to gain of function. Interestingly, the two affected TFs display different cross-resistance to benomyl, suggesting that the pathways implicated are not overlapping. PloA is predicted to have one TMD towards its C-terminus, similar to Candida albicans Upc2, a binuclear cluster zinc family TF regulating ergosterol biosynthetic genes56. Microscopic studies monitoring the uptake of a dimethylaminocoumarin-plocabulin analogue by fluorescence microscopy suggested that Tyr223Asp in AN3969 might prevent drug uptake (not shown). It would be interesting to investigate whether this effect involves an alteration of plasma membrane lipid content. On the other hand, AN8345/ploF, whose chromosomal location suggests co-regulation with a PDR15 homologue, may act by regulating expression of drug efflux pumps.

In conclusion, our genetic analyses in A. nidulans have provided important insights relevant to the mechanism of action of plocabulin. They have also provided important resistance-associated hit proteins involved in metabolic processes that would potentially affect treatment outcomes. Given that, as in cell lines65, resistance hits involving mutations in the primary target were also identified by our screen, the presence of different tubulin isotypes was shown to be important also for fungal cells and fungal cell fitness has been shown to rest on both mitotic and interphase MTs, we have grounds to believe that it would be reasonable to pursue those novel hits to assess their importance in tumor environments.

Materials and Methods

Aspergillus nidulans growth media and techniques

Aspergillus standard complete (CM) or minimal medium (MM) (Cove 1966) was used for growth tests. Plocabulin and benomyl resistance were tested on CM, 3-aminotriazol resistance was tested on MM with 1% glucose, 10 mM ammonium tartrate and the appropriate vitamins. Transformations were as described66. Gene targeting strategies and analysis of transformants were as described in31,67,68.


Microscopy

Microscopy was carried out with a Leica DMI6000B, as described10. MAD4659 (S1 Table) was crossed to obtain resistant progeny expressing gfp::tubA and hhoA::mcherry. For shift experiments plocabulin was added to pre-warmed fresh medium used to substitute the plocabulin-free medium in which hyphae had been growning overnight. GFP and mCherry channels were acquired successively by rapid change of the microscope filters driven by the multidimensional acquisition function of Metamorph (Molecular Devices).


A genetic screen for plocabulin resistant mutants

We mutagenized conidiospores of strain MAD2 (S1 Table) by UV irradiation on plate using a UV stratalinker 1800. 106 to 108 viable spores per petri dish were plated on medium containing plocabulin 10 µM and left to grow in the dark. After 5 days, we picked colonies that had grown faster than a dense background of micro-colonies and streaked them on medium with plocabulin to clean from non-resistant cells. Colonies were further purified by plating spore dilutions on medium without plocabulin (to avoid further plocabulin-induced mutagenesis). Spore dilutions and plating were repeated until colonies appeared homogeneous. This gave rise to strains MAD4647 through MAD4653 (S1 Table). These were backcrossed with MAD3688 to test for Mendelian segregation of resistance and their tubA and benA genes were sequenced using DNA primers 1 to 8 (S2 Table) to test for the presence of mutations in the major tubulins.


Chromosomal allocation

Diploid formation between MAD4629, a master strain carrying diagnostic markers in each chromosome, and MAD4744 (CR), MAD4738 (AR), MAD4747 (ER), MAD4751 (FR), as well as haploidization analysis were as previously described30,69.


Whole genome sequencing

Genomic DNA for WGS was isolated using standard techniques. Genome sequencing and analysis of SNPs were carried out by Life Sequencing S.L. (www.lifesequencing.com) (Illumina HiSeq. 100PE, 50–100x, MAD4650) or Otogenetics corporation (www.otogenetics.com) [Illumina HiSeq. 2500, PE 100–125, 100x, MAD4738(AR), MAD4747 (ER), MAD4751 (FR) and MAD2 (wt)]. To pick candidate SNPs to cause resistance in each strain, we selected all missense, frameshift or 5′UTR mutations that were not present in the genomes of the strains with resistance mutations that were classified to different linkage groups. We visualized WGS data with the Integrative Genomics Viewer70,71.


Construction of tubCΔ benA N100I

For tubC ORF deletion, an appropriate DNA fragment was constructed (fusion PCR, primers 41 to 46) (Fig. 1d) and used to transform MAD5319. tubCΔ (confirmed by Southern blot) strain MAD5755 was crossed with benAN100I MAD4736 to acquire double mutants.


CR identification

To map the causative CR-resistance mutation, we crossed MAD4650, MAD4744, and MAD4745 (CR) with MAD4852, MAD4853, MAD4855, MAD873 and MAD1426 carrying appropriate markers on chromosome VIII. GCN3-1 was followed in genetic crosses by Sanger sequencing or XbaI digestion (GCN3-1 interrupts a XbaI site) of a DNA fragment amplified with primers 9 and 10 (S2 Table). We reconstructed GCN3-1 co-transforming MAD5319 with the pyrG of A. nidulans (amplified with primers 82 and 83) and a fragment of gcn3 carrying GCN3-1 (amplified with primers 9 and 10 from MAD4744) and selecting for pyrimidine prototrophy and GCN3-1 growth/plocabulin resistance.


DR identification

Sequencing of eIF2B regulatory subunits genes GCN3/eIF2Bα (AN0167), GCD7/eIF2Bβ (AN1344) and GCD2/eIF2Bδ (AN6864) was carried out using primers 9, 10, and 19–23 (S2 Table).


Construction of mutants in the GAAC pathway

GCN3-1 benA N100I

MAD4744 was crossed to MAD4757.

Construction of gcn3Δ

Primers 11–16 were used to construct the linear fragment 5utrgcn3::pyrGAf::3utrgcn3 (Fig. 6a)67 that was used to transform MAD1739. gcn3Δ was verified by Southern blot analysis.

The double mutant gcn3-1 gcd7-1 is unviable

We crossed GCN3-1 (MAD5660) and gcd7-1 (MAD4740) and recovered three phenotypic classes (1:1:1 ratio) that were wt-, GCN3-1- and gcd7-1-like (see Supplementary Fig. S6 for details). We genotyped five wt–like, six GCN3-1-like and seven gcd7-1-like strains and confirmed that they carried one or the other or no mutation, but we did not recover any double mutant. Because gcn3/gcd7 loci recombine freely (suggested by recovery of the wt strains at the expected 1/3 ratio), this indicated that the double GCN3-1 gcd7-1 mutant must be unviable.

gcn3Δ gcd7-1

We crossed MAD4740 (gcd7-1) with MAD5386 (gcn3Δ)

gcn4Δ

The appropriate construct for the deletion of gcn4 (AN3675) (Fig. 6a) was constructed by fusion PCR using primers 68–73 and was used to transform MAD5319. Correct genome integration was verified by Southern blot analysis. gcn4Δ displayed compact growth. A genetic cross between a pyrG+; gcn4+ (MAD5660) and pyrG89; gcn4Δ::pyrGAf (MAD5991) strain yielded gcn4Δ strains that were compact and gcn4Δ strains that were indistinguishable from wt. 3 compact gcn4Δ strains were pyrG89, while 3 non-compact gcn4Δ were pyrG+, strongly suggesting that compact growth was related to pyrG89 (pyrimidine auxotrophy).This indicated that pyrGAf only partially complements pyrG89 when integrated in the gcn4 locus.

gcnEΔ

The appropriate construct for the deletion of gcnE (AN3621) (Fig. 6a) was constructed using primers 74–79 and was used to transform MAD5319. Transformants were genotyped using primers 80 and 81. gcnEΔ gcn3-1 strains were obtained by crossing MAD6085 with MAD5660.


AR identification

To test which of the candidate mutations (S5 Table) co-segregated with resistance, MAD4738 (AR) was crossed to MAD3685 and, from the genome of progeny, DNA fragments were amplified that included the candidate polymorphisms’ regions. Primers 30 and 31 were used for AN4084 amplification, 32 and 33 for AN3969 and 34 and 36 for AN8292. PCR fragments from 12 plocabulin-resistant or 12 plocabulin-sensitive mutants were mixed and the mixtures sequenced. Two peaks in the chromatograph at the relevant polymorphism position were indicating that this polymorphism was not homogeneous in the population and, thus, it did not co-segregate with resistance, while single peaks from both the resistant and sensitive strains indicated that the polymorphism co-segregated with resistance (Fig. 8a).

AR reconstruction

A DNA fragment with the AN3969 mutation was amplified from MAD4738 with primers 32 and 33 and was used to transform MAD5321. Transformants were selected on 5 µM plocabulin.

AN3969 cDNA amplification

From total A. nidulans cDNA, we managed to amplify overlapping AN3969 fragments with primers 33–36, 37–38 and 39–40 (Fig. 8a and S2 Table). Sequencing of these fragments verified the introns in the genome annotation, as well as the stop codon, but showed that a much larger transcript, extending towards the 5′prime, is expressed. This transcript contained a Zn2Cys6 binuclear cluster DNA binding motif.

PloA::GFP

A construct to tag AN3969/ploA with GFP at its C-terminus (ploApart::gfp::pyrGAf:: 3utrploA) was constructed by fusion PCR using primers 39 and 57, 25 and 58 and 59 and 60 (gfp::pyrGAf was amplified from p1439). The construct was used to transform MAD5736 (wt) and MAD6079 (AR), pyrG89 nkuAΔ::bar strains (i.e. appropriate for genome integration by homologous recombination of DNA and for selection of the transformants by pyrimidine auxotrophy) coming from the cross between MAD5914 and MAD5736. PCR genotyping of the transformants was with primers 39 and 61. To assess the size of the hybrid protein and its expression levels, total protein was extracted by the alkaline lysis method72. Samples were resolved in a 7.5% SDS-polyacrylamide gel and transferred to a nitrocellulose filter that was reacted with α-GFP antibody (1:5000; Roche). Peroxidase-coupled goat-anti mouse IgG (H + L) (1:5000; Jackson ImmunoResearch) was used as secondary antibody. Bands were detected using ECL (GE Pharmaceuticals, RPN2106).

AN3969Δ

To delete AN3969, primers 24 to 29 were used in fusion PCR (Fig. 8).


ER identification

Linkage assessment of ER to choC and further analysis of the two linked polymorphisms iii and iv used progeny from a cross between MAD4747 (ER) and MAD5660 (choC). Genes affected by iii and iv were amplified and sequenced using primers 47 and 48 (iv) or 49 and 50 (iii). We reconstructed ER by transforming MAD5319 with pyrGAnid and a fragment of AN1281 carrying the resistance mutation (amplified using primers 48 and 84) and selecting for pyrimidine prototrophy and plocabulin resistance.


FR identification

To test how candidate polymorphisms in FR strains segregate relative to plocabulin resistance, we crossed MAD4751 (FR) to MAD3685. Polymorphisms were sequenced in PCR pools from resistant or sensitive progeny, using primers 51 and 52 (AN5786), 53 and 54 (AN8345), and 55 and 56 (AN12394). Reconstruction of AN8345/ploF polymorphism co-segregating with resistance was achieved by transformation of strain MAD5736 with the appropriate construct (Fig. 8i) carrying part of the mutant AN8345 ORF, followed by the pyrGAf and 3′UTR of AN8345. Primers 62 to 67 were used to amplify and fuse the appropriate DNA fragments.


Availability of materials and data

Strains, materials, data and protocols will be made available to readers upon well-founded and reasonable request. Requests for Plocabulin must be addressed to Pharmamar SA, a for-profit company, and can be obtained only through a Material Transfer Agreement with Pharmamar, that will implement a detailed analysis of the scientific proposal.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    Martin, M. J. et al. Isolation and first total synthesis of PM050489 and PM060184, two new marine anticancer compounds. Journal of the American Chemical Society 135, 10164–10171, https://doi.org/10.1021/ja404578u (2013).

  2. 2.

    Prota, A. E. et al. A new tubulin-binding site and pharmacophore for microtubule-destabilizing anticancer drugs. Proceedings of the National Academy of Sciences of the United States of America 111, 13817–13821, https://doi.org/10.1073/pnas.1408124111 (2014).

  3. 3.

    Pera, B. et al. New Interfacial Microtubule Inhibitors of Marine Origin, PM050489/PM060184, with Potent Antitumor Activity and a Distinct Mechanism. ACS Chem Biol 8, 2084–2094, https://doi.org/10.1021/cb400461j (2013).

  4. 4.

    Martinez-Diez, M. et al. PM060184, a new tubulin binding agent with potent antitumor activity including P-glycoprotein over-expressing tumors. Biochemical pharmacology 88, 291–302, https://doi.org/10.1016/j.bcp.2014.01.026 (2014).

  5. 5.

    Galmarini, C. M., Guillen, M. J., Martinez-Leal, J. F., Martinez-Diez, M. & Aviles, P. Abstract 3066: Anti-angiogenic properties of PM060184. Cancer research 76, 3066–3066, https://doi.org/10.1158/1538-7445.am2016-3066 (2016).

  6. 6.

    Oakley, B. R. & Morris, N. R. Nuclear movement is beta–tubulin-dependent in Aspergillus nidulans. Cell 19, 255–262 (1980).

  7. 7.

    Abenza, J. F., Pantazopoulou, A., Rodriguez, J. M., Galindo, A. & Penalva, M. A. Long-distance movement of Aspergillus nidulans early endosomes on microtubule tracks. Traffic (Copenhagen, Denmark) 10, 57–75, https://doi.org/10.1111/j.1600-0854.2008.00848.x (2009).

  8. 8.

    Salogiannis, J., Egan, M. J. & Reck-Peterson, S. L. Peroxisomes move by hitchhiking on early endosomes using the novel linker protein PxdA. The Journal of cell biology 212, 289–296, https://doi.org/10.1083/jcb.201512020 (2016).

  9. 9.

    Penalva, M. A., Zhang, J., Xiang, X. & Pantazopoulou, A. Transport of fungal RAB11 secretory vesicles involves myosin-5, dynein/dynactin/p25 and kinesin-1 and is independent of kinesin-3. Molecular biology of the cell, https://doi.org/10.1091/mbc.E16-08-0566 (2017).

  10. 10.

    Pantazopoulou, A., Pinar, M., Xiang, X. & Penalva, M. A. Maturation of late Golgi cisternae into RabE(RAB11) exocytic post-Golgi carriers visualized in vivo. Molecular biology of the cell 25, 2428–2443, https://doi.org/10.1091/mbc.E14-02-0710 (2014).

  11. 11.

    Penalva, M. A. et al. Searching for gold beyond mitosis: Mining intracellular membrane traffic in Aspergillus nidulans. Cellular logistics 2, 2–14, https://doi.org/10.4161/cl.19304 (2012).

  12. 12.

    Egan, M. J., McClintock, M. A. & Reck-Peterson, S. L. Microtubule-based transport in filamentous fungi. Current opinion in microbiology 15, 637–645, https://doi.org/10.1016/j.mib.2012.10.003 (2012).

  13. 13.

    Salogiannis, J. & Reck-Peterson, S. L. Hitchhiking: A Non-Canonical Mode of Microtubule-Based Transport. Trends in cell biology 27, 141–150, https://doi.org/10.1016/j.tcb.2016.09.005 (2017).

  14. 14.

    Morris, N. R., Lai, M. H. & Oakley, C. E. Identification of a gene for alpha-tubulin in Aspergillus nidulans. Cell 16, 437–442 (1979).

  15. 15.

    Sheir-Neiss, G., Lai, M. H. & Morris, N. R. Identification of a gene for beta-tubulin in Aspergillus nidulans. Cell 15, 639–647 (1978).

  16. 16.

    Oakley, C. E. & Oakley, B. R. Identification of gamma-tubulin, a new member of the tubulin superfamily encoded by mipA gene of Aspergillus nidulans. Nature 338, 662–664, https://doi.org/10.1038/338662a0 (1989).

  17. 17.

    Takahashi, M., Matsumoto, S., Iwasaki, S. & Yahara, I. Molecular basis for determining the sensitivity of eucaryotes to the antimitotic drug rhizoxin. Molecular & general genetics: MGG 222, 169–175 (1990).

  18. 18.

    Doodhi, H. et al. Termination of Protofilament Elongation by Eribulin Induces Lattice Defects that Promote Microtubule Catastrophes. Current biology: CB 26, 1713–1721, https://doi.org/10.1016/j.cub.2016.04.053 (2016).

  19. 19.

    May, G. S. The highly divergent beta-tubulins of Aspergillus nidulans are functionally interchangeable. The Journal of cell biology 109, 2267–2274 (1989).

  20. 20.

    Oakley, B. R. Tubulins in Aspergillus nidulans. Fungal genetics and biology: FG & B 41, 420–427, https://doi.org/10.1016/j.fgb.2003.11.013 (2004).

  21. 21.

    Weatherbee, J. A., May, G. S., Gambino, J. & Morris, N. R. Involvement of a particular species of beta-tubulin (beta 3) in conidial development in Aspergillus nidulans. The Journal of cell biology 101, 706–711 (1985).

  22. 22.

    May, G. S. & Morris, N. R. Developmental regulation of a conidiation specific beta-tubulin in Aspergillus nidulans. Developmental biology 128, 406–414 (1988).

  23. 23.

    Xiong, Y. & Oakley, B. R. In vivo analysis of the functions of gamma-tubulin-complex proteins. Journal of cell science 122, 4218–4227, https://doi.org/10.1242/jcs.059196 (2009).

  24. 24.

    Su, W., Li, S., Oakley, B. R. & Xiang, X. Dual-Color imaging of nuclear division and mitotic spindle elongation in live cells of Aspergillus nidulans. Eukaryotic cell 3, 553–556 (2004).

  25. 25.

    Harris, S. D., Morrell, J. L. & Hamer, J. E. Identification and characterization of Aspergillus nidulans mutants defective in cytokinesis. Genetics 136, 517–532 (1994).

  26. 26.

    Bergen, L. G. & Morris, N. R. Kinetics of the nuclear division cycle of Aspergillus nidulans. Journal of bacteriology 156, 155–160 (1983).

  27. 27.

    Wolkow, T. D., Harris, S. D. & Hamer, J. E. Cytokinesis in Aspergillus nidulans is controlled by cell size, nuclear positioning and mitosis. Journal of cell science 109(Pt 8), 2179–2188 (1996).

  28. 28.

    Dorter, I. & Momany, M. Fungal Cell Cycle: A Unicellular versus Multicellular Comparison. Microbiology spectrum 4, https://doi.org/10.1128/microbiolspec.FUNK-0025-2016 (2016).

  29. 29.

    Ovechkina, Y. et al. Spindle formation in Aspergillus is coupled to tubulin movement into the nucleus. Molecular biology of the cell 14, 2192–2200, https://doi.org/10.1091/mbc.E02-10-0641 (2003).

  30. 30.

    Todd, R. B., Davis, M. A. & Hynes, M. J. Genetic manipulation of Aspergillus nidulans: heterokaryons and diploids for dominance, complementation and haploidization analyses. Nature protocols 2, 822–830, https://doi.org/10.1038/nprot.2007.113 (2007).

  31. 31.

    Nayak, T. et al. A versatile and efficient gene-targeting system for Aspergillus nidulans. Genetics 172, 1557–1566, https://doi.org/10.1534/genetics.105.052563 (2006).

  32. 32.

    Hinnebusch, A. G. & Lorsch, J. R. The mechanism of eukaryotic translation initiation: new insights and challenges. Cold Spring Harbor perspectives in biology 4, https://doi.org/10.1101/cshperspect.a011544 (2012).

  33. 33.

    Hinnebusch, A. G. Translational regulation of GCN4 and the general amino acid control of yeast. Annual review of microbiology 59, 407–450, https://doi.org/10.1146/annurev.micro.59.031805.133833 (2005).

  34. 34.

    Klann, E. & Dever, T. E. Biochemical mechanisms for translational regulation in synaptic plasticity. Nature reviews. Neuroscience 5, 931–942, https://doi.org/10.1038/nrn1557 (2004).

  35. 35.

    Kashiwagi, K. et al. Crystal structure of eukaryotic translation initiation factor 2B. Nature 531, 122–125, https://doi.org/10.1038/nature16991 (2016).

  36. 36.

    Yang, W. & Hinnebusch, A. G. Identification of a regulatory subcomplex in the guanine nucleotide exchange factor eIF2B that mediates inhibition by phosphorylated eIF2. Molecular and cellular biology 16, 6603–6616 (1996).

  37. 37.

    Pavitt, G. D., Ramaiah, K. V., Kimball, S. R. & Hinnebusch, A. G. eIF2 independently binds two distinct eIF2B subcomplexes that catalyze and regulate guanine-nucleotide exchange. Genes & development 12, 514–526 (1998).

  38. 38.

    Kashiwagi, K., Ito, T. & Yokoyama, S. Crystal structure of eIF2B and insights into eIF2-eIF2B interactions. The FEBS journal, https://doi.org/10.1111/febs.13896 (2016).

  39. 39.

    Krishnamoorthy, T., Pavitt, G. D., Zhang, F., Dever, T. E. & Hinnebusch, A. G. Tight binding of the phosphorylated alpha subunit of initiation factor 2 (eIF2alpha) to the regulatory subunits of guanine nucleotide exchange factor eIF2B is required for inhibition of translation initiation. Molecular and cellular biology 21, 5018–5030, https://doi.org/10.1128/mcb.21.15.5018-5030.2001 (2001).

  40. 40.

    Harashima, S., Hannig, E. M. & Hinnebusch, A. G. Interactions between positive and negative regulators of GCN4 controlling gene expression and entry into the yeast cell cycle. Genetics 117, 409–419 (1987).

  41. 41.

    Hinnebusch, A. G. & Fink, G. R. Positive regulation in the general amino acid control of Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America 80, 5374–5378 (1983).

  42. 42.

    Elsby, R. et al. The alpha subunit of eukaryotic initiation factor 2B (eIF2B) is required for eIF2-mediated translational suppression of vesicular stomatitis virus. Journal of virology 85, 9716–9725, https://doi.org/10.1128/jvi.05146-11 (2011).

  43. 43.

    Dev, K. et al. The beta/Gcd7 subunit of eukaryotic translation initiation factor 2B (eIF2B), a guanine nucleotide exchange factor, is crucial for binding eIF2 in vivo. Molecular and cellular biology 30, 5218–5233, https://doi.org/10.1128/mcb.00265-10 (2010).

  44. 44.

    Hannig, E. M., Williams, N. P., Wek, R. C. & Hinnebusch, A. G. The translational activator GCN3 functions downstream from GCN1 and GCN2 in the regulatory pathway that couples GCN4 expression to amino acid availability in Saccharomyces cerevisiae. Genetics 126, 549–562 (1990).

  45. 45.

    Hoffmann, B., Valerius, O., Andermann, M. & Braus, G. H. Transcriptional autoregulation and inhibition of mRNA translation of amino acid regulator gene cpcA of filamentous fungus Aspergillus nidulans. Molecular biology of the cell 12, 2846–2857 (2001).

  46. 46.

    Georgakopoulos, P., Lockington, R. A. & Kelly, J. M. The Spt-Ada-Gcn5 Acetyltransferase (SAGA) complex in Aspergillus nidulans. PloS one 8, e65221, https://doi.org/10.1371/journal.pone.0065221 (2013).

  47. 47.

    Georgakopoulos, T. & Thireos, G. Two distinct yeast transcriptional activators require the function of the GCN5 protein to promote normal levels of transcription. The EMBO journal 11, 4145–4152 (1992).

  48. 48.

    Nutzmann, H. W. et al. Bacteria-induced natural product formation in the fungus Aspergillus nidulans requires Saga/Ada-mediated histone acetylation. Proceedings of the National Academy of Sciences of the United States of America 108, 14282–14287, https://doi.org/10.1073/pnas.1103523108 (2011).

  49. 49.

    Reyes-Dominguez, Y. et al. Nucleosome positioning and histone H3 acetylation are independent processes in the Aspergillus nidulans prnD-prnB bidirectional promoter. Eukaryotic cell 7, 656–663, https://doi.org/10.1128/ec.00184-07 (2008).

  50. 50.

    Canovas, D. et al. The histone acetyltransferase GcnE (GCN5) plays a central role in the regulation of Aspergillus asexual development. Genetics 197, 1175–1189, https://doi.org/10.1534/genetics.114.165688 (2014).

  51. 51.

    Schulz, C., Schendzielorz, A. & Rehling, P. Unlocking the presequence import pathway. Trends in cell biology 25, 265–275, https://doi.org/10.1016/j.tcb.2014.12.001 (2015).

  52. 52.

    Banerjee, R., Gladkova, C., Mapa, K., Witte, G. & Mokranjac, D. Protein translocation channel of mitochondrial inner membrane and matrix-exposed import motor communicate via two-domain coupling protein. eLife 4, e11897, https://doi.org/10.7554/eLife.11897 (2015).

  53. 53.

    Schiller, D., Cheng, Y. C., Liu, Q., Walter, W. & Craig, E. A. Residues of Tim44 involved in both association with the translocon of the inner mitochondrial membrane and regulation of mitochondrial Hsp70 tethering. Molecular and cellular biology 28, 4424–4433, https://doi.org/10.1128/mcb.00007-08 (2008).

  54. 54.

    Schjerling, P. & Holmberg, S. Comparative amino acid sequence analysis of the C6 zinc cluster family of transcriptional regulators. Nucleic Acids Res 24, 4599–4607 (1996).

  55. 55.

    Todd, R. B. & Andrianopoulos, A. Evolution of a fungal regulatory gene family: the Zn(II)2Cys6 binuclear cluster DNA binding motif. Fungal genetics and biology: FG & B 21, 388–405, https://doi.org/10.1006/fgbi.1997.0993 (1997).

  56. 56.

    MacPherson, S., Larochelle, M. & Turcotte, B. A fungal family of transcriptional regulators: the zinc cluster proteins. Microbiology and molecular biology reviews: MMBR 70, 583–604, https://doi.org/10.1128/mmbr.00015-06 (2006).

  57. 57.

    Naar, A. M. & Thakur, J. K. Nuclear receptor-like transcription factors in fungi. Genes & development 23, 419–432, https://doi.org/10.1101/gad.1743009 (2009).

  58. 58.

    Wolfger, H., Mamnun, Y. M. & Kuchler, K. The yeast Pdr15p ATP-binding cassette (ABC) protein is a general stress response factor implicated in cellular detoxification. The Journal of biological chemistry 279, 11593–11599, https://doi.org/10.1074/jbc.M311282200 (2004).

  59. 59.

    Furst, R. & Vollmar, A. M. A new perspective on old drugs: non-mitotic actions of tubulin-binding drugs play a major role in cancer treatment. Die Pharmazie 68, 478–483 (2013).

  60. 60.

    Komlodi-Pasztor, E., Sackett, D., Wilkerson, J. & Fojo, T. Mitosis is not a key target of microtubule agents in patient tumors. Nat Rev Clin Oncol 8, 244–250 (2011. http://www.nature.com/nrclinonc/journal/v8/n4/suppinfo/nrclinonc.2010.228_S1.html

  61. 61.

    Hiyama, T. B., Ito, T., Imataka, H. & Yokoyama, S. Crystal structure of the alpha subunit of human translation initiation factor 2B. Journal of molecular biology 392, 937–951, https://doi.org/10.1016/j.jmb.2009.07.054 (2009).

  62. 62.

    Ye, J. et al. The GCN2-ATF4 pathway is critical for tumour cell survival and proliferation in response to nutrient deprivation. The EMBO journal 29, 2082–2096, https://doi.org/10.1038/emboj.2010.81 (2010).

  63. 63.

    Fels, D. R. & Koumenis, C. The PERK/eIF2alpha/ATF4 module of the UPR in hypoxia resistance and tumor growth. Cancer biology & therapy 5, 723–728 (2006).

  64. 64.

    Rovini, A., Savry, A., Braguer, D. & Carre, M. Microtubule-targeted agents: when mitochondria become essential to chemotherapy. Biochimica et biophysica acta 1807, 679–688, https://doi.org/10.1016/j.bbabio.2011.01.001 (2011).

  65. 65.

    Dumontet, C. & Jordan, M. A. Microtubule-binding agents: a dynamic field of cancer therapeutics. Nature reviews. Drug discovery 9, 790–803, https://doi.org/10.1038/nrd3253 (2010).

  66. 66.

    Tilburn, J. et al. Transformation by integration in Aspergillus nidulans. Gene 26, 205–221 (1983).

  67. 67.

    Szewczyk, E. et al. Fusion PCR and gene targeting in Aspergillus nidulans. Nature protocols 1, 3111–3120, https://doi.org/10.1038/nprot.2006.405 (2006).

  68. 68.

    Osmani, A. H., Oakley, B. R. & Osmani, S. A. Identification and analysis of essential Aspergillus nidulans genes using the heterokaryon rescue technique. Nature protocols 1, 2517–2526, https://doi.org/10.1038/nprot.2006.406 (2006).

  69. 69.

    McCully, K. S. & Forbes, E. The use of p-fluorophenylalanine with ‘master strains’ of Aspergillus nidulans for assigning genes to linkage groups. Genetical research 6, 352–359 (1965).

  70. 70.

    Thorvaldsdottir, H., Robinson, J. T. & Mesirov, J. P. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Briefings in bioinformatics 14, 178–192, https://doi.org/10.1093/bib/bbs017 (2013).

  71. 71.

    Robinson, J. T. et al. Integrative genomics viewer. Nat Biotechnol 29, 24–26, https://doi.org/10.1038/nbt.1754 (2011).

  72. 72.

    Hervas-Aguilar, A. & Penalva, M. A. Endocytic machinery protein SlaB is dispensable for polarity establishment but necessary for polarity maintenance in hyphal tip cells of Aspergillus nidulans. Eukaryotic cell 9, 1504–1518, https://doi.org/10.1128/ec.00119-10 (2010).

  73. 73.

    Arst, H. N. Jr., Hernandez-Gonzalez, M., Penalva, M. A. & Pantazopoulou, A. GBF/Gea mutant with a single substitution sustains fungal growth in the absence of BIG/Sec. 7. FEBS letters 588, 4799–4806, https://doi.org/10.1016/j.febslet.2014.11.014 (2014).

Download references

Acknowledgements

We thank Juan Fernando Martínez-Leal for critically reading the manuscript and supplying reagents, Herb Arst for help with genetic mapping, Carlos del Cerro for help with visualizing NGS data, Mario Pinar for cDNA, J. Fernando Díaz for eribulin and rhizoxin, Berl and Liz Oakley for GFP::TubA and HhoA::mCherry strains and Elena Reoyo for technical assistance. We also thank 3 anonymous referees for their constructive comments. This work was supported by the Spanish Ministry of Economy and Competitiveness [INNPACTO, project number IPT-2011-0752-900000 and BIO2015-65090R] and by a CSIC-Pharmamar contract.

Author information

Author notes

    • Areti Pantazopoulou

    Present address: Department of Molecular Genetics and Cell Biology, University of Chicago, 920 East 58th Street, Chicago, IL, 60637, USA

Affiliations

  1. Department of Cellular and Molecular Biology, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain

    • Areti Pantazopoulou
    •  & Miguel A. Peñalva
  2. Departamento de Biología Celular y Farmacogenómica, Pharma Mar S.A., Colmenar Viejo, Madrid, Spain

    • Carlos María Galmarini

Authors

  1. Search for Areti Pantazopoulou in:

  2. Search for Carlos María Galmarini in:

  3. Search for Miguel A. Peñalva in:

Contributions

Conceptualization: A.P., M.A.P. and C.M.G. Data Curation: A.P. and M.A.P. Formal Analysis: A.P. and M.A.P. Funding Acquisition: C.M.G. and M.A.P. Investigation: A.P. and M.A.P. Methodology: A.P. and M.A.P. Project Administration: A.P., C.M.G. and M.A.P. Resources: A.P., C.M.G., M.A.P. Supervision: A.P. Validation: A.P. Visualization: A.P. and M.A.P. Writing – Original Draft Preparation: A.P. Writing – Review & Editing: A.P., C.M.G. and M.A.P.

Competing Interests

C.M.G. is an employee of Pharmamar that has commercial interest in the compound used in this work. A.P. had a CSIC-Pharmamar contract. M.A.P. declares no potential conflict of interest.

Corresponding author

Correspondence to Areti Pantazopoulou.

Electronic supplementary material

About this article

Publication history

Received

Accepted

Published

DOI

https://doi.org/10.1038/s41598-018-26736-3

Rights and permissions

Creative Commons BY

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.