Transcript levels of the Aspergillus fumigatus Cdc42 module, polarisome, and septin genes show little change from dormancy to polarity establishment

Abstract

Aspergillus fumigatus is the most common airborne pathogen causing fatal mycoses in immunocompromised patients. During the first 8 hours of development A. fumigatus conidia break dormancy, expand isotopically, establish an axis of polarity, and begin to extend germ tubes in a polar manner. The transition from isotropic to polar growth is critical for tissue invasion and pathogenesis. In the current work, we used two-color microarrays to examine the A. fumigatus transcriptome during early development, focusing on the isotropic to polar switch. The most highly regulated transcripts in the isotropic to polar switch did not include known polarity genes. Transcripts encoding the Cdc42 module, polarisome components, and septins, known to be critical players in polarity, showed relatively steady levels during the isotropic to polar switch. Indeed, these transcripts were present in dormant conidia, and their levels changed little from dormancy through germ tube emergence. Not only did the isotropic to polar switch show little change in the expression of key polarity genes of the Cdc42 module, polarisome, and septins, it also showed the lowest overall levels of both up- and downregulation in early development.

Introduction

The filamentous fungus Aspergillus fumigatus is one of the leading agents of invasive fungal disease.¹ Like many other filamentous fungi, A. fumigatus reproduces by means of abundant conidia. When inhaled by the immunocompromised, these conidia break dormancy and eventually develop into highly polar hyphae that invade tissue and disseminate causing disease.² Previous in vitro work has shown that when multiple A. fumigatus dormant conidia are simultaneously inoculated to carbon containing medium they break dormancy and maintain roughly synchronous development through the first several nuclear divisions.³ Conidia expand isotropically for several hours before germ tubes emerge and extend to form hyphae. The paradigm for polar growth in filamentous fungi is that isotropically expanding cells first pick a spot at which to establish an axis of polarity and deposit cortical markers. Next products of key polarity genes (e.g., the Cdc42 module and polarisome; Table 1) are recruited to the cortical markers, and finally the cell growth machinery (e.g. actin, septins, and cell wall biosynthetic enzymes) is reoriented to add new cellular components in the correct place resulting in asymmetric growth.^4–8 This paradigm suggests that if polarity establishment is transcriptionally controlled, the transcription of polarity genes should be greatly upregulated shortly before early polarized growth. To investigate the switch between isotropic and polar growth, we profiled the transcriptome of A. fumigatus during early vegetative development.

Table 1.

Selected genes directing polarity in filamentous fungi.

Polarity component	Gene name	A. fumigatus gene
Cdc42 module	modA, CDC42	Afu2g05740
	CDC24	Afu4g11450
	RGA1	Afu1g12680
	BEM2	Afu3g06280
	BEM3	Afu6g06400
	claA, CLA4	Afu5g05900
Polarisome	sepA/cdc11 formin	Afu6g04940
	sepG, spaA	Afu2g03710
	budA	Afu1g09640
	bemA	Afu4g04120
	STE20	Afu7g04330
Actin	actA	Afu6g04740
Septins	aspA septin	Afu5g08540
	aspB septin	Afu7g05370
	aspC septin	Afu5g03080
	aspD septin	Afu1g08850

Materials and methods

Strains and growth conditions

The clinical isolate A. fumigatus Af293 (Nierman et al. 2005) was used in all experiments. The strain was cultured on CM agar (0.6% NaNO₃, 0.052% KCl, 0.052% MgSO₄•7H₂O, 0.152% KH₂PO₄, 0.1% Glucose, 0.2% Peptone, 0.1% Yeast Extract, 0.1% Casamino acid, 0.0001% Biotin, 0.0001% Pyridoxine, 0.0001% Thiamine, 0.0001% Riboflavin, 0.0001% ρ-aminobenzoic acid, 0.0001% Nicotinic acid, and trace elements). To measure synchrony of early development, 3.0 × 10⁶ conidia/ml were inoculated to 50 ml of liquid CM and incubated at 37°C with shaking at 220 rpm. Cells were harvested at 0, 2 h, 4 h, 6 h, and 8 h of incubation and fixed in 3.7% formaldehyde, 50 mM phosphate buffer (pH 7.0) and 0.2% Triton X-100 for 30–60 min. Cells were washed with water, incubated for 5 min with 10 μg/ml calcofluor white (Bayer) to label septa and 100 ng/ml Hoechst 33258 (Sigma) to label nuclei, washed again and mounted on a microscope slide for viewing.⁹ In each of three independent assays, 200 conidia were microscopically examined, nuclei were counted, and morphology was scored as dormant (small round conidia the same size as in inoculum), isotropically expanding (round cells larger than dormant conidia), polar (visible small germ tube), and branched (any visible branch emerging). Cells were observed and photographed using a Zeiss Axioplan microscope and Zeiss MC100 microscope camera system.

Microarray analysis

Flasks containing 150 ml complete medium were inoculated with conidia to a final concentration of 3.0 × 10⁶ / ml and incubated for 0, 2, 4, 6, and 8 h at 37°C with constant shaking (220 rpm). Following incubation, cells were harvested using a Steritop filter (Millipore), frozen, ground with liquid nitrogen, and RNA purified using TRIzol (Invitrogen) and the RNeasy MiniElute Clean up kit (Qiagen) according to the manufacturers’ instructions. The cDNA synthesis and indirect Cy dye incorporation were performed using the standard J. Craig Venter Institute (JCVI) protocol (standard operating procedure [SOP] M007) with the following modifications: Six micrograms of total RNA were used in the first-strand cDNA reaction with a 3:1 aminoallyl-dUTP to dTTP ratio. Dried aminoallyl labeled cDNA was resuspended in 9 ml of 50 mM sodium carbonate (pH 9.0) and added directly to the appropriate dye vial from the Amersham CyDye postlabeling reactive dye pack (RPN5661). The resulting labeled cDNA was hybridized to an A. fumigatus glass slide microarray spotted with 70-mers representing all predicted open frames (PFGRC/TIGR A. fumigatus microarray, version 3) according to the standard JCVI protocol (SOP M0008). Three biological replicates were performed using cultures grown in parallel, and technical dye swap replicates were carried out for each biological replicate for a total of four hybridizations. The slides were scanned using a Perkin-Elmer ScanArray microarray scanner, and the resulting tagged-image format file images were imported into TIGR Spotfinder (version 3.1.1). Raw intensities were calculated for each detectable spot using the Otsu method, and quality control filtering was used to eliminate values from spots with poor morphology or low signal-to-noise ratios. The resulting intensities were then imported into TIGR MIDAS (version 2.20), and spots with values of <10,000 were removed from the data set using the low intensity filter.

For analysis of relative transcript abundance, normalization was done using the LOWESS (Locifit) algorithm (global mode; 0.33 smoothing parameter), and print tip bias was addressed using standard deviation regularization. The dye swap consistency filter was used to remove data for genes with inconsistent expression levels between dye swap replicates. Finally, the in-slide replicate function was used to average the normalized intensity values representing replicate spots on the array. The resulting data were analyzed using Microsoft Excel, and genes that had a greater than twofold difference in expression were identified. Microarray data are available at Gene Expression Omnibus (accession no. GSE7541).

For analysis of temporal transcript profiles, data were quantile normalized using Genespring GX 11.02 (Agilent). Spotfinder data sets from two-color microarray hybridization analysis described above were separated into Cy3 and Cy5 channels and ranked using the method of Bolstad et al.¹⁰ Expression levels in adjacent timepoints were compared to determine fold change. Heat maps were constructed using hierarchical clustering.

RT-PCR and qPCR

RNA isolation was as described for microarray analysis except that instead of grinding in liquid nitrogen, frozen samples were mixed with 1 ml of 0.5 mm glass beads (Biospec no. 11079105z) in 2 ml tubes and homogenized using the Retsch Mixer Mill 400 (30 Hz, 1 min). Three independent experiments were performed for each time point (biological replicates). The iScript Reverse Transcription Supermix for RT-qPCR (BioRad) was used to prepare cDNA from 1 ug of total RNA according to the manufacturer's instructions. The resulting cDNA was diluted 1:5. Primer sequences for qPCR were determined using the software QuantPrime available online.¹¹The qPCR was carried out using 1.5 ul of diluted cDNA in 15 ul reactions, 1.5 ul of 200 nM primer pairs, and 7.5 ul of 2× Maxima SYBR Green qPCR Master Mix (Thermo Scientific). The following PCR protocol was used on Agilent Stratagene Mx3005P (96-well plates) real-time PCR systems: 2 min at 50°C, 10 min at 95°C, 30 s at 95°C, and 1 min at 60°C repeated in 40 cycles, followed by melting curve analysis. Each qPCR analysis was repeated three times (technical replicates). Cycle threshold (Ct) values for each reaction were calculated using Agilent Mx3005P software, with baseline set to cycle 10–15 and threshold to 0.1 Rn. Real-time PCR amplification efficiencies were calculated using the LinRegPCR tool.¹² The expression levels of all genes of interest were the normalized to the geometric mean of gpdA and hho1 (histone H1) expression.¹³ Primers are shown in Table 2.

Table 2.

qRT-PCR primers.

Gene name	Sequence number	Fw	Rv
aspD	Afu1g08850	GAGATCCAGACTGTCTCCCACATC	ACCCGGTGTGTCGACAATGTTC
cdc24	Afu4g11450	ATGGAACTGCGCAAACAAACCG	AGTCGAGAACAGACTGGATGGTG
cdc42	Afu2g05740	TCTTTGATGAGGCGATTGTTGCG	GACGCATCTGGACTTCTTCTTCGG
claA	Afu5g05900	TTGCGACAATCTGTGCCGAGAC	CGTAAGTTTGGCGCAGAAACCG
gdpA	Afu5g01970	ACGAGATCAAGCAGGCCATC	TCAGTGTAGCCGAGGATGTTC
hho1	Afu3g06070	AGAGCGTAACGGTAGCAGTC	CGGCCTTGATAGCCTTGTTG

Results

To investigate the transcriptional program underlying the isotropic to polar switch, we inoculated conidia of A. fumigatus clinical isolate Af293 to complete medium (CM) with shaking at 37°C. We monitored morphology and isolated RNA at set time points. As expected, early development was relatively synchronous (Fig. 1). At 0 h dormant conidia had a single nucleus and were small and round. At 2 h and 4 h conidia had a single nucleus and were isotropically expanding. At 6 h most cells had four nuclei (92%), and all had polarized, showing a teardrop shape. At 8 h most cells had eight nuclei (85%), and all had short germ tubes.

Figure 1.

Aspergillus fumigatus early growth is synchronized. (A) A. fumigatus conidia were inoculated to complete medium and microscopically monitored for small dormant conidia (open circles with solid line), isotropically- expanding conidia (open squares with broken line), germ tube emergence (triangles with bold solid line), and branch emergence (filled squares with broken line); n = 200. Error bars represent the standard error of three biological replicates. (B) Representations of the morphology found in most of the population at the corresponding time point.

To investigate the isotropic to polar switch, relative transcript abundance among mRNAs during isotropic expansion, between isotropic expansion and polarity establishment, and during early polar growth was analyzed by two color hybridization against A. fumigatus oligonucleotide microarrays (Fig. 2). We expected that the expression levels of many genes would show large changes coinciding with the large morphological shift of polarity establishment at 6 h. Because we were most interested in the genes with the largest expression changes during polarity establishment, we used an arbitrary greater than twofold change to identify the most differentially expressed genes. To our surprise, we saw the fewest changes in gene expression during the isotropic to polar switch. We saw a greater than twofold change in only 80 genes in 4 h samples relative to 6 h samples, with 46 being upregulated at 6 h and 34 being downregulated at 6 h (Fig. 2, Table 3). To our surprise, no known polarity genes were among the most highly regulated during the isotropic to polar switch (Tables 1 and 3). However, among the 40 most highly upregulated genes were several associated with the cytoskeleton and cell wall that likely play roles in remodeling of cell morphology at this time. The upregulated genes include the actin-binding protein profilin, chitinase, two putative GPI-anchored proteins, two predicted adhesin-like proteins, and the allergen AspF4 (Table 3). A number of iron-associated genes were also upregulated in early polar growth including a putative siderophore biosynthesis lipase, a putative ortholog of Sit1 (a siderophore-iron transporter), and a high affinity iron permease. Also of interest is the up-regulation of an ortholog of Hex1, the protein that makes up the Woronin bodies that plug septa. Upregulation is consistent with previous work that showed that these bodies are visible well before septation.¹⁴

Figure 2.

Transcriptional profiling by microarray. Dotted lines indicate sample pairs analyzed by two-color hybridization on microarrays. The number of genes whose transcripts showed at least twofold increase is indicated on the top. The number of the genes whose transcripts showed at least twofold decrease is indicated on the bottom. Arrows indicate strategy for comparative gene expression analysis whereby arrow head points from reference sample to test sample. All changes in gene expression are expressed relative to test sample.

Table 3.

Transcripts with largest fold change in isotropic to polar switch (4 to 6 h).

	Locus	Aspergillus fumigatus annotation	log2a
Upregulated
1	Afu5g03010	conserved hypothetical protein	4.60
2	Afu5g08830	HEX1	4.24
3	Afu5g03760	class III chitinase ChiA1	3.35
4	Afu4g07300	hypothetical protein	3.23
5	Afu1g05790	GPI anchored protein, putative	2.92
6	Afu3g03390	Putative siderophore biosynthesis lipase/esterase	2.71
7	Afu2g07680	L-ornithine N5-oxygenase	2.41
8	Afu2g07800	GPI anchored protein, putative	2.18
9	Afu5g11380	rho-gdp dissociation inhibitor	2.04
10	Afu4g03760	glycine dehydrogenase	1.93
11	Afu7g05950	EF-hand protein	1.88
12	Afu2g07810	cytosolic hydroxymethyltransferase, putative	1.81
13	Afu1g10780	glycine cleavage system T protein	1.76
14	Afu7g02570	heterokaryon incompatibility protein (Het-C), putat.	1.74
15	Afu4g14380	conserved hypothetical protein	1.73
16	Afu4g08580	antioxidant protein LsfA	1.71
17	Afu2g06100	DUF907 domain protein	1.69
18	Afu4g03050	profilin	1.67
19	Afu4g11730	glycerol dehydrogenase (GldB), putative	1.67
20	Afu1g12070	glycine cleavage system H protein	1.66
21	Afu2g10030	vip1 protein	1.64
22	Afu6g05210	malate dehydrogenase, NAD-dependent	1.64
23	Afu4g09080	C2H2 transcription factor (Seb1), putative, putative	1.62
24	Afu8g07090	Predicted adhesin-like protein	1.59
25	Afu6g08530	sister chromatid separation protein (Src1), putative	1.54
26	Afu2g06150	disulfide isomerase, putative	1.54
27	Afu1g17200	Putative nonribosomal peptide synthetase (NRPS) involved in	1.53
		ferricrocin siderophore biosynthesis
28	Afu3g11830	phosphoglucomutase	1.51
29	Afu1g03720	UPF0136 domain protein	1.49
30	Afu7g06060	siderochrome-iron transporter (Sit1), putative	1.46
31	Afu5g03800	high-affinity iron permease CaFTR2	1.46
32	Afu1g15300	choline transport protein, putative	1.45
33	Afu7g02340	L-PSP endoribonuclease family protein (Hmf1), putative	1.42
34	Afu3g10300	galactokinase	1.41
35	Afu3g05530	condensin complex component cnd1	1.39
36	Afu5g10780	UDP-glucose 4-epimerase	1.36
37	Afu3g08650	Putative C1-tetrahydrofolate synthase	1.35
38	Afu1g13940	Predicted adhesin-like protein	1.34
39	Afu2g01040	formaldehyde dehydrogenase	1.31
40	Afu2g03830	allergen Asp F4	1.30
Downregulated
1	Afu3g09690	Laminin-binding protein with extracellular thaumatin domain	−3.21
2	Afu4g08960	GPI anchored protein, putative	−3.16
3	Afu6g03910	integral membrane protein, Mpv17/PMP22 family, putative	−2.39
4	Afu4g11250	carbonic anhydrase family protein	−2.27
5	Afu5g10890	DNA replication licensing factor Mcm6, putative	−1.96
6	Afu4g03360	GPI anchored protein, putative	−1.75
7	Afu6g00510	NADP-dependent alcohol dehydrogenase	−1.70
8	Afu4g05900	conserved hypothetical protein	−1.65
9	Afu8g02430	NADP-dependent alcohol dehydrogenase	−1.65
10	Afu3g11430	arginase	−1.64
11	Afu2g09060	DNA replication licensing factor Mcm4, putative	−1.60
12	Afu2g15440	integral membrane protein, putative	−1.56
13	Afu8g01710	antigenic thaumatin domain protein, putative	−1.52
14	Afu1g00530	thermoresistant gluconokinase family protein	−1.47
15	Afu5g13020	DNA polymerase alpha/primase associated subunit	−1.34
16	Afu7g00920	hypothetical protein	−1.32
17	Afu5g14210	glucose-repressible gene protein-related protein	−1.32
18	Afu3g08990	cspA, Repeat-rich glycophosphatidylinositol (GPI)-anchored cell wall protein	−1.28
19	Afu4g04270	thymidylate synthase	−1.25
20	Afu6g11130	possible replication factor-a protein	−1.21
21	Afu5g02520	DNA replication licensing factor Mcm5, putative	−1.19
22	Afu3g14010	DNA replication licensing factor Mcm2, putative	−1.18
23	Afu2g10020	hypothetical protein	−1.18
24	Afu7g05630	hypothetical protein	−1.17
25	Afu1g01190	hypothetical protein	−1.14
26	Afu4g07970	DNA polymerase delta subunit 2, putative	−1.11
27	Afu1g06100	glutaredoxin	−1.11
28	Afu6g05040	DNA replication factor C subunit Rfc4, putative	−1.07
29	Afu4g08830	telomere length regulator protein (Rif1), putative	−1.07
30	Afu1g15610	oxidoreductase, zinc-binding dehydrogenase family, putative	−1.07
31	Afu2g11260	3-isopropylmalate dehydratase, putative	−1.03
32	Afu3g08710	protein kinase domain-containing protein	−1.03
33	Afu1g14930	hypothetical protein	−1.02
34	Afu6g03050	oleate delta-12 desaturase	−1.00
35	Afu2g13210	DnaJ domain protein	−0.99
36	Afu5g03230	stress response RCI peptide, putative	−0.99
37	Afu4g08600	aldehyde dehydrogenase, putative	−0.98
38	Afu2g05560	exonuclease, putative	−0.97
39	Afu3g11030	histone chaperone (ASF1), putative	−0.97
40	Afu2g10420	branched-chain amino acid aminotransferase, cytosolic	−0.96

^aratio average

The most highly down-regulated genes in early polar growth (4 to 6 h) included seven predicted to encode proteins that are cell wall or membrane resident: a laminin-binding protein, three putative GPI anchored proteins, two putative integral membrane proteins, and an antigenic protein. Among the highly downregulated genes was also a large group predicted to encode proteins involved in DNA synthesis: two putative DNA polymerase associated subunits, a putative DNA replication factor and four putative DNA replication licensing factors, three of which had been upregulated in isotropic growth. The downregulation of DNA synthesis genes is consistent with our observation that the first mitosis leading from a single nucleus to two nuclei begins after 4 h and was completed by 6 h. Not only did the isotropic to polar switch (4 to 6 h) show little regulation of polarity genes, it also showed the lowest overall levels of both up- and downregulation of gene expression in early growth (Fig. 2). The highest level of differential expression occurred between the dormant and early isotropic stages (0 to 1 h).

Our two-color hybridization experiments were designed specifically to compare gene expression between dormant conidia (0 h) and isotropic cells (1 h, 2 h, and 4 h) and between isotropic cells (4 h) and polar cells (6 h and 8 h). To allow comparison between consecutive time points, we quantile normalized our data. In quantile normalization color channels are separated, each spot is ranked by intensity, and intensity rankings for each time point are averaged. Signal rankings can then be compared between any two time points.¹⁰

Consistent with our two-color hybridization results, quantile normalized data showed the highest level of greater than twofold upregulated gene expression between dormant conidia and early isotropic expansion (0 to 1 h) and relatively few genes showed large levels of up- or downregulation in the isotropic to polar switch (4 to 6 h) (Fig. 3). The quantile normalized data revealed an unexpectedly large number of downregulated transcripts in late isotropic expansion (2 to 4 h). However, when we specifically examined the transcription of the Cdc42 module, polarisome, septins, and actin, we found that, with the exception of actin, they showed relatively steady levels of expression throughout early development, and especially in the isotropic to polar switch at 4 h to 6 h (Fig. 4). To validate our microarray results, we performed qRT-PCR on four polarity genes (Fig. 5). The expression levels of two polarity genes, aspD and cla4, showed very little change during early isotropic expansion and early polar growth. The other two genes, cdc42 (modA) and cdc24, rose just after dormancy was broken and then fell in the early timepoints, but were relatively stable during the isotropic to polar switch, consistent with the microarray results.

Figure 3.

Transcriptional profiling by quantile normalization. Two color hybridization data was quantile normalized to allow comparison between time points. The number of genes whose transcripts showed at least twofold increase between adjacent timepoints is indicated on the top. The number of the genes whose transcripts showed at least twofold decrease between adjacent timepoints is indicated on the bottom. Arrows indicate strategy for comparative gene expression analysis whereby arrow head points from reference sample to test sample. All changes in gene expression are expressed relative to test sample.

Figure 4.

Core polarity gene transcripts change little from dormancy to polar growth. Heat map of quantile normalized data shows expression differences between adjacent timepoints. The Cdc42 module, polarisome components and septins do not change much during the isotropic to polar switch (4 to 6 h).

Figure 5.

qRT-PCR of selected polarity genes. Expression levels of selected polarity genes normalized to the geometric mean of two controls (gpdA and hho1).

Discussion

The most striking morphological change in the early development of A. fumigatus takes place during the transition from nonpolar to polar growth with the emergence of the germ tube. The conidium itself is formed by nonpolar growth processes—budding and isotropic expansion on the conidiophore.^15,16 Thus, we predicted that the conidium would carry no or few transcripts encoding the proteins needed to establish polarity and that transcription of the corresponding genes would be highly upregulated before germ tube emergence during the isotropic to polar switch. However, our experiments show that transcripts for the key polarity genes in the Cdc42 module and the polarisome are already present in the dormant conidium at levels that are similar to those in isotropically expanding and early polar cells. Our results also show that the 40 genes showing the most highly differential up- and downregulation do not include any obvious polarity-associated genes. Our results suggest that at least for Cdc42 and the polarisome, transcripts are packed into the dormant conidium, rather than being induced for polar growth.

We saw a relatively flat expression profile during the isotropic to polar switch for many genes and the biggest transcriptional changes were during the first hour after breaking dormancy. A recent study of Aspergillus niger conidia showed a similar pattern with many transcripts packed into the conidium and large changes in transcripts levels both up and down in the first hour after breaking dormancy and much smaller changes at 4 to 6 h.¹⁷

It is certainly possible that other polarity-associated genes that are not among the most highly expressed in the cell act to trigger polarity establishment or that post-transcriptional mechanisms drive polarity establishment. Future work will address these possibilities.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and the writing of the paper.