Mycologia, 97(4), 2005, pp. 804±811. q 2005 by The Mycological Society of America, Lawrence, KS 66044-8897 A phylogenetic study of the genus Haligena (Halosphaeriales, Ascomycota) Jariya Sakayaroj1 INTRODUCTION Department of Microbiology, Faculty of Science, Prince Haligena Kohlm. was described by Kohlmeyer (1961), of Songkla University, Hat Yai, Songkhla, 90112, Thailand with the type species H. elaterophora Kohlm. The National Center for Genetic Engineering and unique characteristic of the species was the long bi- Biotechnology, 113 Thailand Science Park, polar strap-like appendages and multiseptate asco- Paholyothin Road, Khlong 1, Khlong Luang, Pathum spores that characterize and clearly distinguish the Thani, 12120, Thailand genus from other members of the Halosphaeriaceae Ka-Lai Pang (Kohlmeyer 1961). A number of species later were Department of Biology and Chemistry, City University assigned to the genus: H. amicta (Kohlm.) Kohlm. & of Hong Kong, 83 Tat Chee Avenue, Kowloon Tong, E. Kohlm., H. spartinae E.B.G. Jones, H. unicaudata Hong Kong SAR School of Biological Sciences, University of Portsmouth, E.B.G. Jones & Le Camp.-Als. and H. viscidula King Henry Building, King Henry I Street, Kohlm. & E. Kohlm. ( Jones 1962, Kohlmeyer and Portsmouth, PO1 2DY, UK Kohlmeyer 1965, Jones and Le Campion-Alsumard Souwalak Phongpaichit 1970). Shearer and Crane (1980) transferred H. spar- tinae, H. unicaudata and H. viscidula to Halosarpheia Department of Microbiology, Faculty of Science, Prince of Songkla University, Hat Yai, Songkhla, 90112, because of their hamate polar appendages that un- Thailand coil to form long thread-like structures. Recent phy- logenetic studies showed that they are not related to E.B. Gareth Jones Halosarpheia and were transferred to Magnisphaera J. National Center for Genetic Engineering and Biotechnology, 113 Thailand Science Park, Campb. et al and Ascosalsum J. Campb. et al (Ander- Paholyothin Road, Khlong 1, Khlong Luang, Pathum son et al 2001, Campbell et al 2003). Haligena amicta Thani, 12120, Thailand is distinct from Haligena in having appendages that arise from the episporium at various points in the spore wall ( Johnson et al 1987). In Haligena append- Abstract: The genus Haligena (Halosphaeriales, As- ages are polar, arising as outgrowths of the ascospore comycota), with two accepted species, is encountered wall. Thereforea new genus Appendichordella R.G. frequently in marine habitats, especially on wood in Johnson et al was introduced to accommodate H. temperate regions. Phylogenetic analyses of Haligena amicta ( Johnson et al 1987). elaterophora (type species) and H. salina were under- Another species that has been accepted in Hali- taken, with partial large subunit ribosomal DNA se- gena is H. salina C.A. Farrant & E.B.G. Jones, which quences, to determine their relationships with other originally was identi®ed as a Remispora-like species closely related genera in the order. The genus was ( Jones 1985, Farrant and Jones 1986). Haligena sal- shown to be polyphyletic within the Halosphaeriales ina differs from the type species in ascospore size, with the type species forming a basal clade to the septation and especially appendage morphology; ap- order. Haligena salina constituted a sister clade with pendages spoon-shaped at the base, initially coiled weak support of Neptunella longirostris in all analyses. and attached closely to the spore wall and separating Haligena elaterophora and H. salina differ signi®cant- to form a long thread-like ®lament (Farrant and ly in the nature of their ascospore appendages: wider, Jones 1986). Thus only two species, H. elaterophora more sticky and strap-like in H. elaterophora and and H. salina, are retained in Haligena. Therefore spoon-shaped at the point of attachment; in H. salina in this study relationships of Haligena elaterophora they are longer and narrower, ®nely drawn out ®la- and H. salina and their af®nity to other genera in ments. A new genus, Morakotiella, is introduced to the Halosphaeriales are investigated with sequences accommodate H. salina. of the large subunit ribosomal DNA (LSU rDNA). Key words: Halosphaeriales, LSU rDNA, molecu- lar systematics, Morakotiella MATERIALS AND METHODS Accepted for publication 23 March 2005. Fungal isolates and culture characteristics.ÐFungal cultures 1 Corresponding author. Email: [email protected] were obtained from City University of Hong Kong Culture 804 SAKAYAROJ ET AL: HALIGENA 805 TABLE I. Fungal isolates sequenced for this study Species Culture No.a Origin GenBank accession No. Haligena elaterophora PP4705 Friday Harbour, USA AY864845 Haligena elaterophora JS147 Portsmouth, England AY864846 Morakotiella salina CY3437 Friday Harbour, USA AY864843 Morakotiella salina BCC12781 Marloes, South Wales AY864844 a CY is from City University of Hong Kong Culture Collection, PP is from Portsmouth University Culture Collection, BCC and JS are from BIOTEC Culture Collection. Collection (CY) and Portsmouth University Culture Collec- and Saagaromyces ratnagiriensis (AF539470). Inclusion and tion (PP) (TABLE I). Additional cultures were obtained by exclusion of all insertion regions had no effect on the tree single-spore isolations from woody material collected at topology in all analyses. Therefore the insertion regions Marloes, South Wales, and Portsmouth, England. Isolates were included in all analyses. are maintained in the BIOTEC Culture Collection and cod- The phylogenetic analyses were performed with PAUP ed as BCC and JS numbers (TABLE I). All cultures were 4.0b10 (Swofford 2002) with maximum parsimony analysis grown in seawater glucose-yeast extract-peptone broth applying heuristic searches with this setting: 100 replicates (GY P) (Abdel-Wahab et al 2001) on a rotary shaker at 200 of random stepwise addition of sequence and tree-bisection- rpm at 25 C. reconnection (TBR) branch-swapping algorithm. Gaps were treated as missing data; all characters were given equal DNA extraction, ampli®cation and sequencing.ÐFungal bio- weight. The consistency indices (CI), retention indices (RI) mass was harvested through cheesecloth, washed several and rescaled consistency indices (RC) were calculated for times with sterile distilled water, frozen in liquid nitrogen each tree generated. Weighted parsimony analysis was per- and ground into ®ne powder with a mortar and pestle. formed with a step matrix to weight nucleotide transfor- Ground fungal mycelia of 50-100 mg were placed into 400 mations based on the reciprocal of the observed transition mL lysis buffer (O'Donnell et al 1997). DNA was extracted to transversion (ti : tv) ratio, which was estimated by maxi- following the instruction of a Nucleospint Plant DNA ex- mum likelihood score setting in PAUP* (Swofford 2002). traction kit (Macherey-Nagel, Germany). Moreover characters were reweighted according to their RC The partial LSU rDNA was ampli®ed with primers LROR- with PAUP* default setting for reweighting character. Boot- LR7 (Bunyard et al 1994) and JS1-JS8 (Landvik 1996). PCR strap analysis (Felsenstein 1985) was performed for un- reactions were carried out in total volume of 50 mL con- weighted and weighted parsimony with full heuristic search taining 50 ng DNA template, 13 PCR buffer, 1.5 mM on 1000 replicates (10 replicates of random stepwise addi- MgCl , 2 mM dNTPs, 0.2 mM each primer and 0.5 units of 2 tion of sequence and TBR branch-swapping algorithm). Taq Polymerase (DyNAzymey II DNA Polymerase Kit, The proportion of invariable sites, gamma distribution FINNZYMES, Finland). Ampli®cation cycles were per- shape parameter and base frequency were estimated from formed following the procedure of Pang et al (2003). The maximum likelihood score setting in PAUP*. Maximum PCR products were puri®ed with NucleoSpint Extract Kit likelihood analysis was employed with a heuristic search (as (Macherey-Nagel, Germany), following the manufacturer's is stepwise addition sequence and TBR branch-swapping al- instructions. PCR products were sequenced directly by the gorithm) according to these estimated values. Bio Service Unit (BSU) laboratory with BigDye on an ABI Bayesian phylogenetic inference was calculated with 377 automated sequencer (Perkin Elmer). Forward and re- MrBayes 3.0b4 with general time reversible (GTR) model verse primers: JS1, NL4, JS5, JS8, LROR, NL3 and NL4R, of DNA substitution and a gamma distribution rate varia- were used for the sequencing reactions (Bunyard et al 1994, tion across sites (Huelsenbeck and Ronquist 2001). Four Landvik 1996). Each sequence was checked for ambiguous Markov chains were run from random starting trees for bases and was assembled with Bioedit 5.0.6 (Hall 2001). 2 000 000 generations and sampled every 100 generations. Sequence alignment and phylogenetic analyses.ÐThe consen- The ®rst 100 000 generations were discarded as burn-in of sus sequences for each species were multiple aligned by the chain. A majority rule consensus tree of all remaining Clustal W 1.6 (Thompson et al 1994) along with other se- trees, as well as the posterior probabilities, was calculated. quences obtained from the GenBank database. The dataset The alignments were deposited in TreeBase: study accession was re®ned visually in Se-Al v1.0a1 (Rambaut 1999) and number 5 S1228, matrix accession numbers 5 M2135, Bioedit 5.0.6, 6.0.7 (Hall 2001, 2004). Daldinia concentrica M2136. (Bolton) Ces. and Xylaria hypoxylon (L) Grev. were chosen as outgroup for all analyses. Two insertion regions were ob- served, one at a position 835-1035 of Halosarpheia trullifera