Using a multiple sequence alignment, shared primer binding sites were identified in the conserved mtSSU regions among mycobiont genera, that are absent from other ascomycetous fungal genera. Special focus was given to 3’ end unique amplification. The multiple sequence alignment was compiled (Suppl. material 1) from 48 mycobiont mtSSU sequences (10 in-house curated and 38 downloaded from GenBank) and six non-mycobiont/non-lichen-forming fungal sequences of different fungal classes, including Mycocalicium subtile (Pers.) Szatala (Class: Eurotiomycetes), Taphrina flavorubra W.W. Ray (Class: Taphrinomycetes) and Botryotinia fuckeliana (de Bary) Whetzel (Class: Leotiomycetes). Non-mycobiont fungal sequences also include one reference sequence of the ascomycete Triangularia anserina (Rabenh.) X. Wei Wang & Houbraken (Basionym: Podospora anserina (Rabenh.) Niessl; Class: Sordariomycetes; GenBank accession No. X14734), as well as two environmental fungal sequences from Cladophialophora carrionii (Trejos) de Hoog, Kwon-Chung & McGinnis (Class: Eurotiomycetes) and Lichenothelia convexa Henssen (Class: Dothideomycetes), both of which were found to co-inhibit with Melanelia mycobionts. Melting temperature and primer dimer formation were estimated using Multiple Primer Analyzer (ThermoFisher, MA, USA).

EcoPCR (Ficetola et al. 2010) was used for simulated in silico amplification of the mtSSU locus and also to verify amplicon possibilities against an in-house reference ecoPCR database containing overall 2,233,856 fungal sequences. We followed the published procedure (Bellemain et al. 2010) to construct the in-house reference database: all fungal sequences were downloaded from the EMBL fungal database of standard targeted annotated assembled sequences (STD), and sequences were annotated using NCBI taxonomy. Data containing the vast number of annotated fungal sequences were transformed into ecoPCR format before in silico simulation. Two pairs of primers were tested: the commonly used pair mrSSU1 and mrSSU3R, and our newly designed mycobiont-specific primer pair mt-SSU-581-5’ and mt-SSU-1345-3’. In the setting of simulations, amplicon sizes were accepted between 200 bp to 2500 bp, and only up to three nucleotide mismatches between primers and templates were allowed (except for the last two positions at the 3’ end), according to described parameters (Bellemain et al. 2010; Riaz et al. 2011; Liu and Erséus 2017).

The current study included 24 Melanelia herbarium specimens collected from 1997 to 2014, consisting of M. agnata (n=8), M. hepatizon (n=12) and M. stygia (n=4), all of which were morphologically identified and verified with fungal nrITS DNA barcoding and chemotaxonomic analyses in a previous study (Xu et al. 2017). In addition to the Melanelia specimens, 79 specimens of other genera in the same family were also included to test primer universality in the family. The specimen list is provided in Appendix 1. Visible substrates attached to thalli were removed with sterile tweezers or brushes before DNA extraction. Whole genomic DNA was extracted from lichen thalli (ca. 15–20 mg per specimen) using the CTAB method (Cubero et al. 1999).

The PCR master mix and thermal cycler conditions were followed from our published protocol (Xu et al. 2020). Two touchdown programmes were used, where the annealing temperature ramp 61–57 °C (decreasing 1 °C per cycle) was used for mrSSU1 and mrSSU3R, and 54–50 °C for the newly designed primer pair, mt-SSU-581-5’ and mt-SSU-1345-3’, according to predicted melting temperatures (Table 1). Presence and sizes of amplicons were determined by performing 2% agarose gel electrophoresis, using SYBR safe stain (Invitrogen, CA, USA). Amplicons showing single bands were purified with ExoSAP (Fermentas Inc., Hanover, MD, USA) and sequenced in both directions using Sanger sequencing (Macrogen Europe BV, the Netherlands). The same primers were used for both PCRs and Sanger sequencing.

Ambiguous sequences at both ends of the raw sequencing data were trimmed with the software PhyDE v0.9971. Sequence contigs were assembled from both directions and ambiguous base calling was checked. Sequences were identified by BLAST searches. Successful PCR amplification was defined as on-target/mycobiont-specific amplification and clean mycobiont mtSSU sequences without ambiguous base calling. Success rates in percentages were calculated as the number of specimens with successful PCR amplification divided by the total number of specimens. Multiple sequence alignment was performed using MAFFT (Katoh and Standley 2013) and then manually adjusted.

Multiple sequence alignments at the primer binding sites are shown in Fig. 1. The universal primers mrSSU1 and mrSSU3R were designed at the conserved region U2 and U6, respectively, with an expected amplicon size of around 900 base pairs (bp). From the alignments, this primer pair shows little discriminating power between lichen-forming and environmental ascomycetes. Therefore, searching for mycobiont-specific priming sites in the universal regions was not possible, and more variable regions were checked. New primers were designed at the genetic regions where high discriminations were found, particularly at the 3’ end. The new forward primer is located at the variable sites between U2 and U3, showing ca. nine nucleotide differences between Parmeliaceae and environmental fungi. Similarly, the reverse primer was designed at the connecting zone between U5 and U6 with potential discriminating power including roughly nine nucleotide differences.

Figure 1. 

Primer design and sequence alignments at the priming locations. Conservative regions (i.e. U1 to U8) are marked as previously designated (Cummings et al. 1989) with adjustments from the reference. Nucleotide sites are relative to the 1980 bp-long sequence of the fungus Triangularia anserina (GenBank accession No. X14734).

The newly designed primers, mt-SSU-581-5’ with mt-SSU-1345-3’, were named according to the primer nomenclature recommendation (Gargas and DePriest 1996): “mt-SSU” indicates the mitochondrial small subunit ribosomal DNA, and -5’ or -3’ defines the primer annealing to the coding strand (-5’ for forward primers) or the non-coding one (-3’ for reverse primers). The number before -5’ or -3’ is the nucleotide position relative to the reference sequence of the fungus Triangularia anserina at the 5’ end, so these numbers help the estimation of amplicon sizes. For instance, mt-SSU-581-5’ with mt-SSU-1345-3’ will result in amplicons estimated around 700 bp. The numbering of primers, i.e. 581 and 1345, is based on the reference fungus Triangularia anserina, which gives rise to an amplicon size of 764 bp. However, amplicons of Parmeliaceae mtSSU are usually shorter the reference fungus, as seen from Suppl. material 1.

The amplification success was significantly affected by the allowed number of mismatches and positions between archived fungal sequences and primers (Fig. 2). As more mismatches between primers and templates are allowed, the numbers of amplicons increase. The amplicon profiles between the new and universal primer pairs are considerably different. Universal primers give rise to an overwhelming proportion (ca. 98%) of sequences from non-Parmeliaceae fungi, regardless of the number of nucleotide mismatches. Our in silico results show that the amplicons are mainly from three fungal families in the example of three mismatches: Nectriaceae (n=589), Aspergillaceae (n=201) and Trichocomaceae (n=156). A full list of in silico amplicons is provided in Suppl. material 2. In contrast, the new primer pair only produced sequences of Parmeliaceae fungi when less than two mismatches were allowed, and only a single non-Parmeliaceae fungal sequence was amplified with three mismatches.

Figure 2. 

Comparison of primer specificity between the new and the universal primer pairs using in silico PCR. The number of off-target amplifications is classified as others in grey, while desired amplifications of lichen-forming fungi in Parmeliaceae are marked in white. The number of nucleotide mismatches in the priming sites is shown as Mis_0 to Mis_3, indicating 0 to 3 nucleotide mismatches.

Amplicons resulting from the universal primer pair mrSSU1 and mrSSU3R are around 1000 bp in length (lanes 1–3 in Fig. 3A). However, after sequencing, only the amplicons in lane 1 are identified as the lichen-forming fungus M. agnata after BLAST search, while lanes 2 and 3 are off-target amplification of non-lichen-forming fungi, Lichenothelia convexa (GenBank accession No. OQ450499) and Cladophialophora carrionii (GenBank accession No. OQ450500), respectively. The amplicon of C. carrionii is slightly shorter than the other two. Among 24 Melanelia specimens, we only obtained three sequences from the mycobionts (3/24, 12.5% success), while the others showed messy and ambiguous base calling or even no PCR products in gel electrophoresis. Fig. 3B shows Sanger electropherograms resulting from the same DNA extract of one M. agnata specimen but different primers during PCR and sequencing, and the importance of mycobiont-specific primer is highlighted in generating good-quality sequences. Amplicons from the newly designed primers were around 700 bp long. The resulting Sanger electropherograms show unambiguous nucleotides with good quality (Fig. 3B, lower electropherogram), and we obtained 22 mycobiont sequences out of 24 specimens (22/24, 91.7% success). The remaining two specimens yielded no bands after PCR.

Figure 3. 

PCR amplification and sequencing results of the mtSSU locus in the genus Melanelia A 2% agarose gel electrophoresis of PCR products, where lanes 1-3 contain amplicons from the universal primer pair mrSSU1 and mrSSU3R, and lanes 4-6 are from the mycobiont-specific primer pair mt-SSU-598-5’and mt-SSU-1324-3’. Abbreviations: M-molecular ladder 100 bp; NC-negative control B illustration of sequencing results (M. agnata, voucher number LA29683) using the forward primers: the universal primer mrSSU1 (upper) and the mycobiont-specific primer mt-SSU-581-5’ (lower), respectively.

Using the new primers, the mtSSU region was also successfully amplified from DNA extracts of other genera in Parmeliaceae (Table 2), e.g. Alectoria, Evernia, Flavoparmelia, Xanthoparmelia, etc. The exception is the genus Usnea, from which we did not succeed with the newly designed primers (mt-SSU-581-5’ and mt-SSU-1345-3’) or universal primers (mrSSU1 and mrSSU3R). To solve the amplification problem in Usnea, a multiple sequence alignment (Suppl. material 3) was specially made for this genus, which contained 13 reference sequences: six partial mtSSU sequences from PCR, and seven mitochodrial genomes containing the whole mtSSU region. The alignment in (Suppl. material 3) shows that: 1) there are unique and long introns inserted in U4 (e.g. 907 bp for U. trachycarpa (Stirt.) Müll.Arg. and U. antarctica Du Rietz), U5 (e.g. 757-1110 bp for U. halei P.Clerc, U. subgracilis Vain. and U. ceratina Ach.) and U6 (e.g. 535–848 bp for U. subfusca Stirt., U. subscabrosa Nyl. ex Motyka, U. ceratina, U. pennsylvanica Motyka and U. subgracilis) regions, and thus when the primers mrSSU1 and mrSSU3R are used, the estimated size of PCR amplicons may reach ca. 3000 bp; 2) there is a high nucleotide variation in the primer binding sites for the newly designed mycobiont-specific primers mt-SSU-581-5’ and mt-SSU-1345-3’, which prohibit primer binding, especially at the 3’ end (Suppl. material 2: fig. S1).

Therefore, we designed alternative primers (the pair mt-SSU-574-5’ and mt-SSU-897-3’) for the amplification of shorter mtSSU sequences (ca. 400 bp) in Usnea, to avoid amplifying the introns in U4 or U5 region. For Usnea species lacking the intron in the U4 region, we recommend the the primer pair mt-SSU-574-5’ and mt-SSU-1093-3’, which produces amplicons as long as ca. 500 bp. These primers are also useful for old herbarium specimens, for which longer amplicons can not be obtained. The same PCR condition was used for the genus Usnea, except for the adjustment of annealing temperatures: 56–52 °C for touchdown cycles (decrease 1 °C per cycle), and 52 °C for the last 30 to 32 cycles. The primers (mt-SSU-574-5’ and mt-SSU-897-3’/mt-SSU-1093-3’) have been tested with in vitro PCR screening and we got six mycobiont mtSSU sequences out of seven Usnea specimens. The improved success rate is at the cost of variable sites after the U5 zone. The reverse primers, mt-SSU-897-3’ or mt-SSU-1093-3’, can also be used with the mycobiont-specific forward primer mt-SSU-581-5’ for other genera in Parmeliaceae.

We also provide two universal reverse primers (i.e. mt-SSU-1372-3’ and mt-SSU-1492-3’) to replace mt-SSU-1345-3’ when the latter is not working. In this case, we recommend combining a mycobiont-specific forward primer, either mt-SSU-574-5’ or mt-SSU-581-5’, to enhance the specificity for PCR. Two reverse primer mt-SSU-1492-3’ and mt-SSU-1548-3’ were designed to replace the published primer mrSSU3R, for two reasons: 1) the primer mrSSU3R has a much higher melting temperature (T_m 64.2 °C) than the newly designed primers, which have their T_m around 55 °C; 2) mrSSU3R has four consecutive G/C at the 3’ end with a higher risk of non-specific binding.

Notably, the region between U5 and U6 has the highest number of variable sites (Table 3). For the genus Melanelia, nearly all variable sites (18 out of 22 variable nucleotide sites) come from this region, which is the most informative for specimen discrimination. Therefore, we designed an alternative forward primer, mt-SSU-862-5’ in combined use with the reverse primer mt-SSU-1345-3’, which focuses on the amplification of the region between U5 and U6. Although a fairly short amplicon size around 450 bp is obtained from this primer pair (mt-SSU-862-5’ and mt-SSU-1345-3’), it actually contains the majority of the total variable sites (Table 3), which is the most informative region of the mtSSU locus. We have obtained successful PCR products and good sequencing results (i.e. clean mycobiont mtSSU sequence without ambiguous base calling) with this primer pair from specimens which failed in obtaining longer amplicons.