Sizing the fungal and algal genomes of the lichen Cladonia ...people.duke.edu/~darmaleo/Sizing lichen symbiont genomes.pdf · This is the first genome size determination for lichens,

SYMBIOSIS (2009) 49, 43–51

©Springer Science+Business Media B.V. 2009 ISSN 0334-5114

Sizing the fungal and algal genomes of the lichen Cladonia grayi through quantitative PCR Daniele Armaleo* and Susan May Department of Biology, Duke University, Durham, NC 27708, USA, Email. [email protected] (Received November 19, 2008; Accepted February 10, 2009)

Abstract Using a method based on quantitative PCR, we determined that the nuclear genome sizes for the mycobiont and photobiont of the lichen Cladonia grayi are 28.6 Mb and 106.7 Mb, respectively. This is the first genome size determination for lichens, and suggests that between 20,000 and 25,000 genes function in C. grayi. The mycobiont genome size is near the middle of the range observed within the Pezizomycota, the subphylum containing all known ascomycete lichen fungi. The genome size of the photobiont (the green alga Asterochloris sp.) is near the lower end of its class, the Trebouxiophyceae. Genomes in this size range can be sequenced at relatively low cost with current pyrosequencing-based methods. The genome sizing method requires very small amounts of precisely quantified DNA and should be applicable to any lichen whose symbionts can be reliably isolated from one another. Since the symbionts used in this project were isolated from soredia, the lichen’s vegetative propagules, we also describe a method for the establishment of axenic symbiont cultures from large numbers of soredia. Keywords: Genome size, qPCR, Cladonia, lichens, lichenoids, soredia culture, symbiosis, mycobiont, photobiont,

Asterochloris

1. Introduction

The importance of sizing lichen genomes The symbiotic interaction of specialized fungi with

specialized photosynthetic partners produced the evolutionary success story of lichens, that comprise about 20% (13,500 species) of all known fungi (Kirk et al., 2001). Yet a molecular understanding of lichens is lacking. The sequencing of lichen genomes will be of great help in filling that gap, and a prerequisite is the knowledge of the sizes of the component genomes. Little is known about lichen genomes, while those of many parasitic, pathogenic and model fungi have been sequenced (http://fungal.genome.duke.edu/), as well as those of four species of unicellular algae (Misumi et al., 2007), all of them free-living organisms. We report here the first genome size determination for a pair of lichen symbionts: the Cladonia grayi mycobiont and its associated alga Asterochloris sp.

*The author to whom correspondence should be sent.

Cladonia grayi is a well studied member of the

Cladoniaceae, a world-wide and species-rich family. It is classified within the Lecanoromycetes, a class that includes more than 70% of the lichen-forming fungal diversity (Eriksson, 2006). Its unicellular photobiont, Asterochloris sp., belongs to the most common family of lichen algae, the Trebouxiaceae (Rambold et al., 1998). A detailed molecular knowledge of this lichen could be a model for the understanding of this common mutualistic symbiosis.

Among the Ascomycota, lichen fungi occur only within the subphylum Pezizomycota (James et al., 2006; Hibbett et al., 2007). If one surveys current databases (http://fungal.genome.duke.edu/; Kullman et al., 2005), the genome sizes of most Pezizomycota cluster between the 15-megabase (Mb) range of the Pezizomycetes and the 40-Mb range of the Sordariomycetes. Lichen genomes are therefore expected to fall within 15 and 40 Mb, a band slightly narrower than that for all fungi, where most genomes cluster between 10 and 60 Mb (Gregory et al., 2007). Within the class Trebouxiophyceae, known haploid genome sizes range from 50 to 600 Mb (Kapraun, 2007). However, we do not expect such wide class-level variation within Asterochloris and Trebouxia, two closely related photobiont genera shared among many lichen lineages

DOI 10.1007/s13199-009-0012-3

44 D. ARMALEO AND S. MAY

(Rambold et al., 1998; Piercey-Normore and DePriest, 2001; Miadlikowska et al., 2006).

We measured the haploid genome sizes of the mycobiont and photobiont isolated and cultured from the lichen C. grayi, using a method based on quantitative PCR (qPCR). We devised the method independently, but it is similar to one developed earlier by Wilhelm et al. (2003). Most commonly, the sizing of eukaryotic genomes requires cells with good cytological staining properties for microscopy-based methods (Hardie et al., 2002), or cells whose nuclei or chromosomal DNA can be isolated intact for flow cytometry or pulse-field electrophoresis (Gregory et al., 2007). We attempted cytological analysis using fluorescent DNA probes with the cultured C. grayi symbionts, but did not pursue it due to variability of nuclear sizes and shapes (not shown). The qPCR based procedure described here is in principle applicable to any organism with a single type of nuclear genome, as it requires only nanogram quantities of genomic DNA isolated with standard methods, and the knowledge of a few hundred basepairs of sequence; for its application to lichens, it requires DNA extracted from the separated symbionts.

Rationale of method

Since our method differs in several ways from that of

Wilhelm et al. (2003), although its basic idea is the same, we present it here in some detail and point out the differences. The technique is based on the ability to determine by qPCR (Wong and Medrano, 2005) how many copies of a particular sequence are present at the start of a PCR reaction. The core concept is the comparison of two parallel qPCR reactions involving the same primer pair, buffer and conditions, and differing only in the nature of the starting template DNA (Fig. 1). In one reaction, the starter (defined here as “P”) is an isolated PCR fragment of known amount and copy number representing a short unique sequence from the test genome. In the other, the starter sequence (defined here as “G”) is the same, but is part of the test genome. The amount of genomic DNA in the G tube is known but not its copy number, that depends on genome size. Thus, the determination of copy number in the G tube by comparing the G and P reactions allows to determine genome size. For illustration purposes, each tube in Fig. 1 is shown to contain the same number of starter copies (either G or P), thus ideally producing overlapping amplification curves.

Under these conditions, the length of the test genome is given by equation (1): Length of genome (in basepairs) =

!

quantity G( )quantity P( )

× length of P (in basepairs)

Figure 1. Simplified scheme of the genome-sizing principle. In this example, each reaction tube contains two starter DNA molecules. The black arrowheads represent the same primer pair in both reactions, the bold segments represent the PCR target region. P tube: the starters are two isolated copies of the target region. G tube: the starters are two copies of the test genome; each copy comprises the target region and the unknown part (thin line). In this idealized situation, the qPCR curve from P will overlap with the curve from G (bottom graph; ∆Ct = 0; see also text). The known quantities are the amount of DNA introduced in P and G (for example, 4.32 × 10–19 g in P and 2.16 × 10–18 g in G) and the length of the amplified region (for example, 200 bp). From equation (1), the genome length = (2.16 × 10–18 / 4.32 × 10–19) × 200 = 1000 bp.

In practice, the number of starter copies in the G tube is

unknown and different from the known number introduced in the P tube. Thus, (1) will have to be multiplied by the ratio of P starter to G starter copies, commonly

!

" 1. Applying the parameters of quantitative PCR to this case, the ratio [P starter copies]/[G starter copies] equals E^(CtG–CtP), where E is the amplification efficiency during the exponential phase of the reaction, ideally 2, and CtG and CtP are the threshold cycles of the G and P reactions, respectively (see Fig. 2 for a definition of Ct). Thus equation (1) becomes equation (2):

Length of genome =

!

quantity G( )quantity P( )

× length of P ×

!

E^(CtG"CtP)

SIZING THE GENOMES OF LICHEN SYMBIONTS 45

Aside from the formal layout of the equations, the main

difference between our method and that of Wilhelm et al. (2003) involves the way to bypass a potential problem not mentioned in the schematic outline above, where it is assumed that the amplification pattern per template copy is identical for the G and P templates. This assumption may not hold for the first few cycles of the G reaction, when some genomic sequences near the starter may slow down the amplification process before the vast majority of the template becomes the amplified sequence. This initial transient inhibition is ascribed to interferences with priming caused by occasional secondary structures in the template outside of the amplified region (Wilhelm et al., 2003). The inhibition will not occur in the P reaction (due to the absence of sequences external to the amplicon), and will not influence E, the amplification efficiency measured later during the exponential phase when practically all the template in both G and P reactions is the same PCR amplicon. However, an initial slower amplification in the G reaction will increase CtG and therefore E^(CtG–CtP), thereby inflating the genome size estimate from equation (2).

Wilhelm et al. (2003) sought to reduce potential discrepancies between G and P amplifications by using as template in the P reaction a fragment about 200 bp longer than the target amplicon, with the amplicon primers nested each about 100 bp from either end of the template. If the genomic sequences around the amplicon inhibited the early cycles in the G reaction, a similar inhibition was expected in the P reaction due to the presence of the same sequences in the template fragment, thus preserving E^(CtG–CtP).

Figure 2. qPCR tracings of a primer inter-ference test. The traces indicated by the arrows represent the exponential DNA increase (fluorescence, on the Y axis) in the reactions used to test ten algal primer pairs. All reactions contained the same amount of starter genomic DNA. Four replicate reactions were run per primer pair. The threshold line is dashed. The Ct for each trace is the cycle number at which that trace crosses the threshold line. While most primers (eight pairs, block arrow) produce traces that cluster closely together on the left, two (small arrows) are obvious outliers due to strong interference during the initial cycles. The primer pairs differing less than 0.3 Cts at the leftmost edge of the majority cluster were chosen for genome sizing. A similar test was performed with primer pairs designed for mycobiont genes.

Rather than equalizing potential inhibitory effects

between the P and G reactions, we sought to minimize or eliminate the inhibition. This involved the selection of primer pairs showing no interference from surrounding genomic sequences. We first designed several primer pairs for different gene sequences, all with similar Tm and producing amplicons of similar length. Then, to select the primer pairs with minimal or no interference, we performed a screening qPCR with each primer pair, using the same amount of genomic DNA as starter. The different primer pairs produced a range of Cts, most clustering within one Ct from each other and a few resulting in dramatically higher Cts (Fig. 2). We assumed that the clustering of Cts with the majority of primers was due to the fact that in most cases interference was low or absent, but that it increased as Cts increased, occasionally dramatically. Finally, we chose for genome sizing the primer pairs with the lowest Ct values within the majority cluster (Fig. 2).

The method requires that the amplified sequence be a single copy per haploid genome (Fig. 1), because the presence of multiple identical copies within a genome would decrease CtG and thereby “deflate” genome size by a factor equal to copy number, all else being equal. To reduce the possibility of amplifying repeated regions, we first chose several gene sequences identified through BLAST (Altschul et al., 1997), rather than random genomic fragments. Then, the initial screening qPCR mentioned above provided a way not only to select for no-interference primers, but also to identify rare multiple-copy genes. Even the smallest repeat number, a simple gene duplication,


would decrease the Ct of the duplicated sequence by 1 relative to the majority cluster. No such sequences were found among the tested amplicons. This process ensures that genome sizing is based on single copy amplicons with no interference from surrounding genomic DNA.

2. Materials and Methods

Soredia culture and isolation of the lichen symbionts Podetia of C. grayi were collected in 1997 on Mount

Sinai Road near Durham, NC. The presence in the podetia of the depsidone grayanic acid, a secondary compound characteristic of the species (Culberson et al., 1985), was verified by thin layer chromatography (TLC) (Culberson et al., 1981). A TLC-tested podetial sample collected from the same site is deposited in the Duke University cryptogamic herbarium (Cladonia grayi Sandst., 7/29/2003, Daniele Armaleo s.n.). The mycobiont and the photobiont were cultured from soredia removed from podetia.

After a number of preliminary tests, the following procedure was found to reduce the overwhelming number of bacterial and fungal contaminants in field-collected soredia. Sterile water, solutions, pipets and tubes were used throughout. After surface dirt removal, about 20 podetia were vortexed 2 × 30 min in 6 ml water in a 15-ml tube. Leaving the podetia behind, the two water suspensions of soredia (~11 ml total) were pooled into a new tube. The suspension was briefly vortexed, and residual dirt and heavier soredia were allowed to settle for 1 min. The upper 10 ml were carefully removed with a pipet and transferred to a new tube. Stocks of EGTA and Tween in water were added to 1 mM and 0.01% final concentration, respectively. The suspension was vortexed for 1 min, left shaking lying flat for 2 min, returned to the vertical position, and the soredia were allowed to settle for 5 min. The detergent and EGTA allow rapid settling of soredia and help in removing some surface contaminants. The supernatant was completely withdrawn by careful pipetting without disturbing the lose soredial pellet, and 10 ml water were added to the tube. The soredia were vortexed for 1 min and allowed to settle again for 5 min. The water wash was repeated three more times. The last soredial pellet was resuspended in 10 ml of modified minimal Lilly and Barnett (LB) medium (Ahmadjian, 1993) containing tetracycline (15 µg/ml), fluconazole (300 µg/ml) and fluorocytosine (500 µg/ml). LB lacks cytosine that otherwise would compete with fluorocytosine for uptake. Tetracycline (Sigma) and fluconazole (donated by Pfizer) were used to kill bacteria and non-lichen fungi, respectively. Fluorocytosine (Sigma) was used to kill very rapid growers. The suspension was transferred to a 50-ml tube and incubated gently shaking at room temperature for 9 hrs. Soredia were then pelleted by centrifugation and

washed 2 times with 40 ml of water, to remove any unincorporated fluorocytosine. The last soredial pellet was resuspended in 5 ml of water. 500 µl aliquots were each spread onto LB plates containing tetracycline (15 mg/ml) and fluconazole (300 mg/ml).

Every day for two weeks plates were screened under a dissecting microscope, and soredia from which fungal contaminants would emerge were cauterized in situ with a hot needle mounted onto the tip of a soldering iron. Towards the end of the two weeks, the mycobiont was easily distinguished from contaminants due to its uniform phenotype of a compact hyphal “corona” growing slowly and evenly around most soredia. This pattern heralds the dissolution of the differentiated soredial structure, to yield an unstructured mixed colony of fungus and alga, which we label “lichenoid”. At this time, individual uncontaminated lichenoids were transferred to fresh MY plates (Ahmadjian, 1993) without antibiotics, and the occasional rare contaminants were cauterized every few days until they stopped appearing. Individual lichenoids were permanently cultured this way. Where the mycobiont and photobiont segregated from the lichenoids, several samples of hyphae and algae were isolated and grown separately on MY agar. After several weeks, individual mycobiont initials were combined into a single culture, and the same was done with the photobiont.

The genetic identity of the symbionts was confirmed by sequencing ITS rDNA amplified with universal primers: ITS1-F (Gardes and Bruns, 1993) and ITS4 (White et al, 1990) for the mycobiont, and nr-SSU-1780-5' (Piercey-Normore and DePriest, 2001) and ITS4 for the photobiont (not shown). The two strains are designated as Cgr/DA1myc (the C. grayi mycobiont), and as Cgr/DA1pho (the Asterochloris sp. photobiont). Originating from somatic tissues potentially derived from multiple individuals, they cannot be considered genetically clonal. The mycobiont was propagated on solid or liquid MY, and the photobiont on solid MY plates under light (photon flux = 33 µmol/m2/sec). Cultures were maintained at room temperature and transferred to new media every 3–4 months. Mycobiont DNA was extracted from liquid cultures, photobiont DNA from plate cultures, each harvested about two weeks after the last transfer.

DNA extraction from the mycobiont

The mycobiont was lyophilized, ground in liquid

nitrogen and lysed in a DTAB-based buffer with RNAse as described in Armaleo and Miao (1999). The final volume (µl) of DTAB buffer was equal to 20× the dry weight (mg) of the sample. Some of the subsequent steps were modified and are described here. One volume of 5M NaCl was mixed with three volumes of DTAB lysate, and the extract was centrifuged to remove cell debris and precipitated polysaccharides. The supernatant was emulsified for 2


minutes with an equal volume of phenol/chloroform, and phases were separated by centrifugation. The upper (aqueous) phase was transferred to a new tube and 1 volume of a CTAB solution (10% CTAB; 0.7M NaCl) was mixed with 9 volumes of aqueous phase, to precipitate residual polysaccharides. To this solution, an equal volume of chloroform was added and the mix was emulsified by vigorous agitation for 2 minutes. Phases were separated by centrifugation. The upper phase was transferred to a new tube, mixed with an equal volume of isopropanol and, after a 5 minute pause, the DNA was pelleted by centrifugation. After thorough removal of the supernatant, a volume of DTAB-RNAse buffer was added to the pellet, equivalent to 20% of the volume used at the beginning of the procedure. The somewhat tough pellet was not immediately resuspended in the buffer by pipetting, but was first allowed to hydrate at 65oC for 3 minutes. To avoid losses due to inadequate resuspension of the pellet, its complete breakup was carefully observed during pipetting. Treatment with 1/3 volume of 5M NaCl was repeated and, after centrifugation, the supernatant was transferred to a new tube and emulsified for 2 minutes with an equal volume of chloroform. After centrifugation, the upper phase was transferred to a new tube and 2 volumes of EtOH were added to precipitate DNA. After centrifugation and rinsing with 70% EtOH, the pellet was dried under vacuum for 5 minutes. The pellet was left to hydrate in TE buffer (1–3 µl/ mg of original sample dry weight) either overnight on ice or at 65oC for 5 minutes, and thoroughly resuspended by careful pipetting. Concentration was determined with a Nanodrop spectrophotometer (Thermo Scientific) but was approximate, due to the presence of residual UV-absorbing impurities. See “Final cleanup” for the actual quantitation.

Table 1. Primers used for genome sizing. The six primer pairs listed were chosen from a total of 17 primer pairs screened. The gene designations are based on information obtained through BLAST (Altschul et al., 1997). The primers designated as Clgr_Cut, Clgr_btub, Assp_efla were designed by Suzanne Joneson (Joneson, Armaleo and Lutzoni, submitted), the others by the authors. The first three are for fungal, the last three for algal genes.

Name 5’-3’ Sequence Gene

Clgr_Cut_0191F CAGCTATCACTTGGCTTATGCAGTCG Cutinase Clgr_Cut_0291R GCGAGCCGAAGTAGAAATGGAGTAA Pks42F AGCAAGGGAAGCTTACCCCCCTGCG Polyketide synthase Pks43R CTGCCGCCTTTGTGCTCGATTGCGAACTC Clgr_btub_0105F AGTGTGTGATCTGGAAACCTTGGAG Beta tubulin Clgr_btub_0210R AAAGGTCATTACACAGAGGGTGCAG Assp_ef1a_0233F CACACTGGGCAGTCACTTCTTCTTG Elongation factor 1-alpha Assp_ef1a_0340R GACATACGTCAGACTGTGGCTGTTG PelrF1 CTGGAGGAGCCAGTGGGCGCAG ER Lumen protein-retaining receptor PelrR1 GGGTCTTGTGGGCCATGTGCGTC PflsF1 GCCGGAGGTGGAGCATAGGCTTG Flavonol synthase PflsR1 CGGAGGTTCTGCCTGCCTTCAGG

DNA extraction from the photobiont We know from experience that Asterochloris releases

less impurities than the mycobiont, and DNA extraction is therefore simpler. The alga was transferred from a plate to a 1.5 ml microfuge tube, lyophilized, weighed, and powderized in liquid nitrogen. For weights lower than 20 mg, the grinding was as described for the mycobiont in Armaleo and Miao (1999). Larger amounts were ground in a mortar. The ground cells were thoroughly resuspended in lysis buffer (15/1, v/w) (100 mM Tris HCl, pH 8; 2% SDS; 400 mM NaCl; 40 mM EDTA, pH 8) containing RNAse (100 µg/ml) and incubated at 65oC for 5 minutes, with occasional mixing. The solution was emulsified with an equal volume of phenol/chloroform by vigorous shaking for 2 minutes, and phases were separated by centrifugation. The upper (aqueous) phase was transferred to a new tube, emulsified with an equal volume of chloroform for 2 minutes, and phases were separated by centrifugation. The upper phase was transferred to a new tube, the DNA was precipitated by mixing with two volumes of EtOH and pelleted by centrifugation. The pellet was rinsed with 70% EtOH and lyophilized. The pellet was left to hydrate in TE buffer (5 µl/mg of original sample dry weight) either overnight on ice or at 65oC for 5 minutes, and thoroughly resuspended by careful pipetting. Also for algal DNA, the approximate concentration was determined with a Nanodrop spectrophotometer. See “Final cleanup” for the actual quantitation.

Primer pair screening

Primers were designed from seven gene sequences


cloned from the mycobiont and ten from the photobiont. The polyketide synthase (pks) gene sequence was identified by the authors, the others by Suzanne Joneson (Joneson, Armaleo and Lutzoni, submitted). Primers were screened by qPCR as follows (only the primers selected for genome sizing are listed in Table 1). A qPCR master mix was prepared, containing genomic DNA and all the other components, except primer pairs. The master mix was then subdivided into submixes to which the different primer pairs were added. Four 15-µl replicates for each submix were aliquoted into a white 96-well PCR plate (Biorad) and the plate was sealed with microseal ‘B’ film (Biorad). Each replicate reaction contained: 5 µl of genomic DNA (3 ng/µl); 1.5 µl of Invitrogen 10× PCR buffer (200 mM Tris HCl, pH 8.4, 500 mM KCl); 5.62 µl of water; 0.3 µl of Invitrogen 50× Rox Dye; 0.03 µl of a 100× dilution of Invitrogen Sybr Green I (50,000× stock); 0.9 µl of MgCl2 (50 mM stock); 1.2 µl of a dNTP mix (each nucleotide at 1.25 mM); 0.3 µl of Taq DNA polymerase (Apex, 1 unit/µl); 0.075 µl of each primer (10 µM stock). PCR was performed in a Chromo 4 real time PCR detector (Biorad) on a PTC-200 thermal cycler (MJ Research, Inc). Cycling parameters were: 10 min at 95ºC followed by 40 cycles, each of 30 s at 95ºC, 30 s at 60ºC, 30 s at 72ºC. The fluorescence signal was measured after each extension at 72ºC. Ct values were obtained with the MJ OpticonMonitor Analysis software version 3.1. Fig. 2 shows the results of such an experiment. Table 1 lists the primer pairs selected using this screen.

Preparation of starter PCR fragments (“P”)

The primer pairs selected through the screen described

in the previous paragraph (Table 1) were used to prepare the PCR fragment starters needed for genomic sizing. Each 50-µl PCR reaction contained: 39.5 µl of mycobiont or photobiont genomic DNA (5–10 ng) in water; 5 µl of Fisher 10× PCR buffer (100 mM Tris-HCl, pH 9.0; 500 mM KCl; 15 mM MgCl2); 4 µl of a dNTP mix (each nucleotide at 1.25 mM); 0.5 µl of Taq DNA polymerase (Fisher, 5 units/µl); 0.5 µl of each primer (10 µM stock). Cycling parameters were: 10 min at 95ºC followed by 40 cycles, each of 30 s at 95ºC, 30 s at 60ºC, 30 s at 72ºC.

Final cleanup, quantitation, and storage of genomic DNA and of PCR fragments

DNA quantitation errors due to possible residual UV

absorbing impurities will produce corresponding errors in genome sizing. All the starter DNAs for qPCR (PCR fragments as well as genomic DNA) were therefore filtered through Qiaquick columns (Qiagen), and the final DNA quantitation was done using a Qubit fluorometer (Invitrogen) whose principle does not involve UV absorption. Qiaquick columns are primarily designed for

the cleanup of PCR fragments, but we found that they work also with genomic DNA. The cleanup was carried out following the manufacturer’s instructions, except for two modifications designed to ensure complete removal of ethanol from the column matrix: a) the “dry” centrifugation was performed for 90 seconds using a new collection tube; b) the cartridge was kept open at 50oC for 10 minutes before eluting the sample. The sample was finally eluted twice with 30 µl of water (for a total volume of 60 µl). Concentrations were determined by Qubit fluorometry and confirmed by gel electrophoresis, also used to verify the absence of RNA. To avoid repeated freezing and thawing, samples were subdivided into single-use aliquots (2–3 µl) and stored at -80o C. Each aliquot contained 200–250 ng of genomic DNA or 10–30 ng of PCR fragment.

Genome sizing

The same procedure was used for either fungal or algal

DNA. Siliconized barrier tips and tubes were used for the dilutions. One concentration (3 ng/µl) of genomic DNA was used in all “G” reactions. The “P” reactions were performed using four 10× dilutions of each starter PCR fragment, in order to calculate its amplification efficiency coefficient E (Fig. 3). To minimize variation, each PCR fragment was diluted using the same protocol, regardless of the fragment’s initial concentration, determined as indicated previously and generally between 5 and 20 ng/µl. First, 2 µl fragment were serially diluted 100× three times. The third dilution of this series was then used as the highest concentration (designated 1) used for qPCR. Concentrations 2, 3 and 4 were prepared as 10× serial dilutions of concentration 1, diluting 10 µl with 90 of water. Each 15-µl qPCR reaction contained 5 µl of DNA (either genomic or PCR fragment) and 10 µl of mix components identical to those used for primer screening. Also the 96-well plates and cycling parameters were the same. For each primer pair, four replicates of the “G” reaction were run, and three replicates each of concentrations 1, 2, 3, 4 of the “P” reaction (Fig. 3). Ct values, obtained with the MJ OpticonMonitor Analysis software, were exported to Microsoft Excel files and, in the case of the P reactions, used to plot Ct as a function of the log of fragment dilution for each primer pair (Fig. 3). The slope of the linear function provided the E value for each primer pair according to E = 10 ^-1/slope (Wong and Medrano, 2005). For genome sizing, E and Ct values (Fig. 3) were used according to equation (2).

3. Results and Discussion For axenic culture, lichen mycobionts are commonly

obtained from spores, and algae from single somatic cells (Ahmadjian, 1993, and references therein). The C. grayi


symbionts used for this study were isolated while developing a method to obtain long-term axenic cultures from large number of soredia on agar media. Lichen soredia have been cultured under non-sterile conditions on natural substrates (Schuster et al., 1985; Stocker-Wörgötter and Türk, 1988) or on low-moisture, low-nutrient filter substrates that reduce, but don’t eliminate, the growth of contaminants (Armaleo, 1991). Soredia have been also grown on minimal agar media for short term study (Koopmann et al., 2007; Hauck et al., 2002). The method described here removes the massive bacterial and fungal contamination of natural soredia, and allows the long term propagation of many mixed mycobiont-photobiont cultures (“lichenoids”) and the axenic isolation of symbionts from them.

The validity of the qPCR method for genome sizing was verified with genomes of known size (Wilhelm et al., 2003). We applied a modification of that method to size the hitherto unexplored genomes of a pair of lichen symbionts, the mycobiont Cladonia grayi and its associated alga Asterochloris sp. We used three genes/primer pairs each for the fungus and the alga. The calculated genome sizes are shown in Fig. 4. The average for the mycobiont is 28.6 Mb and for the photobiont 106.7 Mb. Within each genome, the largest and smallest values differ by about 20%, a fact that emphasizes the need for multiple determinations with several different primer pairs. This variation probably reflects the combined effects of small differences in primer efficiency, DNA quantitation and sample dilution. Among the replicates with one primer pair, most of the variance is due to slight differences in Ct, which affect the exponent in equation (2).

Figure 3. Use of qPCR data for genome sizing. One of the experiments with the Pelr primer pair is shown as example (each experiment was repeated four times to generate each bar in Fig. 4). The values for the actual Cts and the log of the concentrations (data columns on the top left) were used to generate the graph for the P reactions (top right). The slope was used to calculate the qPCR efficiency (see text) and the line equation was used to calculate the trendline Cts at each of the four P concentrations. The resulting input parameters for equation (2) are listed on the bottom. The G reaction Ct used in equation (2) was the average of the four replicates (16.492 in this case). The P reaction Cts used in equation (2) were trendline Cts rather than average Cts, as the former reflect all the data points rather than just those at each concentration. Any of the four trendline Cts, paired with its corresponding P concentration, yields the same genome size, 1.02 × 108 bp in this case.

Figure 4. The estimated genome sizes for the Cladonia grayi symbionts. The height of the bars represents the genome size (in Mb on the Y axis) calculated using the primer pairs for the genes listed along the X axis (Pks, Tub and Cut are fungal genes, the others are algal; see Table 1). Each determination comprised four replicates (error bars). Arrows indicate the average genome size for the fungus (28.6 Mb) and for the alga (106.7 Mb).

M M e g a b a s e s


The DNA extraction protocols did not separate nuclear from organellar DNA. Therefore, the quantity of G in the numerator of equation (2) is an overestimate relative to the actual amount of nuclear DNA added, leading to a corresponding inflation of the calculated genome size. However, this inflation is minimal and within the error intrinsic to the method. In the slow-growing mycobiont, mitochondrial DNA is probably around 1%, closer to the percent in the filamentous ascomycetes Aspergillus nidulans (Lopez Perez and Turner, 1975) and Neurospora crassa (Luck and Reich, 1964), than to the 5–15% (depending on growth conditions) in the fast-growing yeast Saccharomyces cerevisiae (Fukuhara, 1969; Grimes et al., 1974). In the photobiont Asterochloris, chloroplast and mitochondrial DNA together probably do not exceed 5%, based on values determined for other unicellular algae like Chlamydomonas reinhardtii (Wurtz et al., 1977) and Euglena gracilis (Lyman et al., 1975).

The 107 Mb genome of Asterochloris is at the lower end of the Trebouxiophyceae size range. The 29 Mb genome of the mycobiont is near the middle of the Pezizomycota size range, a region occupied also by the 28 to 37 Mb genomes of Aspergilli (http://www.broad.mit.edu/annotation/ genome/aspergillus_terreus/GenomeStats.html). It is of interest that Aspergilli, which are non-lichenized, belong to the Eurotiomycetes that incude the lichenized order Verrucariales and are closely related to two other classes of lichens, the Lecanoromycetes (to which Cladonia belongs) and the Lichinomycetes (James et al., 2006). Aspergilli are therefore thought to be derived from former lichen fungi that lost the symbiotic life style (Lutzoni et al., 2001; James et al., 2006). The sizing of more genomes across the lichen clades populating the Pezizomycota will allow to investigate whether the narrow size range of Cladonia and Aspergillus can be generalized and perhaps correlated with symbiosis.

The genomes of Aspergillus clavatus, A. fumigatus, and A. terreus are all about 29 Mb long and each is predicted to contain 9000–10,000 genes (http://www.broad.mit.edu/ annotation/genome/aspergillus_terreus/GenomeStats.html). The 120 Mb genome of the unicellular alga Chlamydomonas reinhardtii is predicted to contain 15,000 genes (Merchant et al., 2007). This suggest that the fungus-alga total comprises between 20,000 and 25,000 genes in C. grayi. It is fortunate that the genomes of C. grayi and Asterochloris sp., and perhaps those of lichen symbionts in general, are relatively small. Although these genomes still require extensive assembly and annotation efforts, their raw sequencing with good coverage and accuracy is approachable with the newer and less expensive sequencing methodologies based on pyrosequencing (Ronaghi, 2001).

Acknowledgements

We thank Suzanne Joneson for providing most gene

sequences and primers, Vicky Mobley for help in developing the soredia cultures, and Lisa Bukovnik for introducing us to the Qubit system.

REFERENCES Ahmadjian, V. 1993. The Lichen Symbiosis. J. Wiley and Sons,

Inc., New York, 250 pp. Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z.,

Miller, W., and Lipman, D.J. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research 25: 3389–3402.

Armaleo, D. 1991. Experimental microbiology of lichens: mycelia fragmentation, a novel growth chamber, and the origins of thallus differentiation. Symbiosis 11: 163–177.

Armaleo, D. and Miao, V. 1999. Symbiosis and DNA methylation in the Cladonia lichen fungus. Symbiosis 26: 143–163.

Culberson, C.F., Culberson, W.L., and Johnson, A. 1981. A standardized TLC analysis of β-orcinol depsidones. Bryologist 84: 16–29.

Culberson, C.F., Culberson, W.L., and Johnson, A. 1985. Orcinol-type depsides and depsidones in the lichens of the Cladonia chlorophaea group (Ascomycotina, Cladoniaceae). Bryologist 88: 380–387.

Eriksson, O.E. 2006. Outline of Ascomycota. Myconet 12: 1–82. Fukuhara, H. 1969. Relative proportions of mitochondrial and

nuclear DNA in yeast under various conditions of growth. European Journal of Biochemistry 11: 135–139.

Gardes, M. and Bruns, T.D. 1993. ITS primers with enhanced specificity for basidiomycetes – application to the identification of mycorrhizae and rusts. Molecular Ecology 2: 113–118.

Gregory, T.R., Nicol, J.A., Tamm, H., Kullman, B., Kullman, K., Leitch, I.J., Murray, B.G., Kapraun, D.F., Greilhuber, J., and Bennett, M.D. 2007. Eukaryotic genome size databases. Nucleic Acids Research 35: D332–D338.

Grimes, G.W., Mahler, H.R., and Perlman, P.S. 1974. Nuclear gene dosage effects on mitochondrial mass and DNA. Journal of Cell Biology 61: 565–574.

Hardie, D.C., Gregory, T.R., and Hebert, P.D.N. 2002. From pixels to picograms: a beginners' guide to genome quantification by Feulgen Image Analysis Densitometry. Journal of Histo-chemistry and Cytochemistry 50: 735–749.

Hauck, M., Paul, A., Mulack, C., Fritz, E., and Runge, M. 2002. Effects of manganese on the viability of vegetative diaspores of the epiphytic lichen Hypogymnia physodes. Environmental and Experimental Botany 47: 127–142.

Hibbett, D.S., and 65 other authors. 2007. A higher-level phylogenetic classification of the fungi. Mycological Research 111: 509–547.

James, T.Y., and 69 other authors. 2006. Reconstructing the early evolution of fungi using a six-gene phylogeny. Nature 443: 818–822.

Kapraun, D.F. 2007. Nuclear DNA content estimates in green algal lineages: Chlorophyta and Streptophyta. Annals of Botany 99: 677–701.

Kirk, P.M., Cannon, P.F., David, J.C., and Stalpers, J. 2001. Ainsworth & Bisby's Dictionary of the Fungi. 9th edition, CAB International, Wallingford.

Koopmann, R., Stevens, H., Franzen-Reuter, I., Frahm, J-P., and Grote, M. 2007. In vitro inhibition of soredial growth in the epiphytic lichen Physcia tenella (Ascomycetes: Lecanorales) by a variety of bark phenols. The Lichenologist 39: 567–572.


Kullman, B., Tamm, H., and Kullman, K. 2005. Fungal Genome Size Database. http://www.zbi.ee/fungal-genomesize.

Lopez Perez, M. and Turner, G. 1975. Mitochondrial DNA from Aspergillus nidulans. FEBS Letters 58: 159–163.

Luck, D.J.L. and Reich, E. 1964. DNA in mitochondria of Neurospora crassa. Proceedings of the National Academy of Science USA 52: 931–938.

Lutzoni, F., Pagel, M., and Reeb, V. 2001. Major fungal lineages are derived from lichen symbiotic ancestors. Nature 411: 937–940.

Lyman, H., Jupp, A.S., and Larrinua, I. 1975. Action of nalidixic acid on chloroplast replication in Euglena gracilis. Plant Physiology 55: 390–392.

Merchant, S.S., and 124 other authors. 2007. The Chlamydomonas genome reveals the evolution of key animal and plant functions. Science 318: 245–250.

Miadlikowska, J., and 29 other authors. 2006. New insights into classification and evolution of the Lecanoromycetes (Pezizomycotina, Ascomycota) from phylogenetic analyses of three ribosomal RNA- and two protein-coding genes. Mycologia 98: 1090–1103.

Misumi, O., Yoshida, Y., Nishida, K., Fujiwara, T., Sakajiri, T., Hirooka, S., Nishimura, Y., and Kuroiwa, T. 2007. Genome analysis and its significance in four unicellular algae, Cyanidioshyzon merolae, Ostreococcus tauri, Chlamydomonas reinhardtii, and Thalassiosira pseudonana. Journal of Plant Research 121: 3–17.

Piercey-Normore, M.D. and DePriest, P. 2001. Algal switching among lichen symbioses. American Journal of Botany 88: 1490–1498.

Rambold, G., Friedl, T., and Beck, A. 1998. Photobionts in lichens: possible indicators of phylogenetic relationships? Bryologist 101: 392–397.

Ronaghi, M. 2001. Pyrosequencing sheds light on DNA sequencing. Genome Research 11: 3–11.

Schuster, G.S., Ott, S., and Jahns, H.M. 1985. Artificial culture of lichens in the natural environment. The Lichenologist 17: 247–253.

Stocker-Wörgötter, E. and Türk, R. 1988. Artificial cultures of the cyanobacterial lichen Peltigera didactyla (Peltigeraceae) in the natural environment. Plant Systematics and Evolution 165: 39–48.

White, T.J., Bruns, T., Lee, S., and Taylor, J.W. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: PCR Protocols: A Guide to Methods and Applications. Innis, M.A., Gelfand, D.H., Sninsky, J.J., and White, T.J., eds. Academic Press, Inc., New York, pp. 315–322.

Wilhelm, J., Pingoud, A., and Hahn, M. 2003. Real-time PCR-based method for the estimation of genome sizes. Nucleic Acids Research 31(10) e56.

Wong, M.L. and Medrano, J.F. 2005. Real-time PCR for mRNA quantitation. BioTechniques 39: 1–11.

Wurtz, E.A., Boynton, J.E., and Gillham, N.W. 1977. Perturbation of chloroplast DNA amounts and chloroplast gene transmission in Chlamydomonas reinhardtii by 5-fluorodeoxyuridine. Proceedings of the National Academy of Science USA 74: 4552–4556.

Sizing the fungal and algal genomes of the lichen Cladonia ...people.duke.edu/~darmaleo/Sizing lichen symbiont genomes.pdf · This is the first genome size determination for lichens,

Documents