Endophytic fungi are microbes that reside in plants, but without any visible symptoms on the host plant [1] -[3] . Ecological features of these fungi are still largely unknown, but some confer functions that positively affect host plants. These functions include stimulation of plant growth [4] -[6] , resistance to environmental stress (e.g., drought or edaphic stresses, such as heavy metal ions; [7] -[9] ) and defense mechanisms against herbivores or diseases [10] [11] . Thus, at least some of the endophytic fungi play important roles in the evolution of plants and plant-fungus interactions, and the roles could be more important than previously known. We need more insight into the interactions of endophytic fungi and their host plants to fully understand their evolutionary roles.

Holoparasitism is a special nutritional feature of flowering plants. Holoparasitic plants are echlorophyllous and all substrates (including photosynthesites as carbon sources) necessary for growth and reproduction are drawn from the host plants, with a direct plant-to-plant connection [12] [13] . As for photosynthetic flowering plants, holoparasitic plants have endophytic fungi, although reports are limited (Rafflesia [14] ; Cuscuta [15] ). With the addition of endophytic fungi, cost-benefit relationships in those tripartite interactions are quite complicated. If the holoparasitic plants have their own symbiotic endophytes, the hosts also provide carbon resources to the fungi and are parasitized by both plants and fungi. Symbiotic endophytes of holoparasitic plants also might support parasitization of host plants. Thus, to uncover the ecological and evolutionary roles, studies focusing on species composition of endophytic fungi in various holoparasitic plants are needed.

Balanophoraceae (Santalales) is a large family of holoparasitic plants, comprising about 17 genera and 40 species distributed mainly in tropical regions worldwide [16] . An updated description of the taxon was presented previously [17] . All members of this family fully depend on carbon resources from woody plants growing around them. Although Balanophoraceae is a major family of holoparasitic plants, there is little information about their endophytic fungi. There is only one report of B. harlandii Hook.f. in China [18] . The main objective of that study was to isolate from a medicinally useful plant the fungi effective against bacteria. Not all fungi isolated were examined and no comparisons were made with host plants.

In this study, we isolated and examined endophytic fungi from Balanophora japonica Makino (Figure 1). Balanophora is the largest genus in the Balanophoraceae family (15 spp.), distributed mainly in the Old World tropics [19] . Balanophora japonica is the most common of the five species of Balanophora in Japan. It is widely distributed in the western part of the Japanese archipelago (including the Ryukyus) and Taiwan [19] [20] . Thus, this species is appropriate for the first step in investigating the composition of endophytic fungi and their ecological and evolutionary roles in Balanophoraceae.

We isolated a total of 23 fungal strains from three individuals (one from Kami and two from Sakawa; Table 2). The most frequently isolated genus from B. japonica was Trichoderma-Hypocrea (8 of 23, 34.8%), but this genus was not isolated at Sakawa. The next frequently isolated genus was Penicillium (6 of 23, 26.1%), and all samples used in this study had the fungal genus as endophytes. Phialemonium (five strains) was isolated only from Sakawa individuals. Additional genera were Podospora (two strains), Mortierella and Bionectia (one strain each). A total of 13 strains were isolated from inflorescences and the largest number (5 of 6) of Penicillium strains and all Phialemonium were isolated from inflorescences. On the other hand, 10 residual strains were isolated from tubers, and six of the eight Trichoderma-Hypocrea strains and all Podospora strains were isolated from tubers.

We also isolated eight strains from a host plant of a Kami sample (identified as Symplocos lancifolia Siebold et Zucc.; Table 1). Similar to the parasite, Trichoderma-Hypocrea was the most common fungus isolated (3 of 8;

Note: T001, T003, and T004 are individual IDs of B. japonica. T001 was originated from Kami and the others from Sakawa. HT002 is an ID of a host plant of B. japonica (Symplocos lancifolia from Kami). The subsequent letters “F” and “R” indicate the parts of plants used for isolation (“F” refers to inflorescences and “R” to tubers for B. japonica and roots for a host plant).

37.5%), and Penicillium and Podospora were the least common (2 of 8; 25%). Umbelopsis (1 strain) was the only fungus that was isolated from a host plant, but not from B. japonica. All results for molecular identifications will appear in DDBJ/EMBL/GenBank International DNA Data Bank under the accession nos. LC109274- LC109304.

All isolated fungal genera are known as common soil fungi and frequently found in various woody plant species as endophytes [23] -[25] . The similarity in fungal composition on B. japonica and its host suggested either a similar process of acquiring endophytic fungi from soil, horizontal transmission from hosts to parasites or both. In the case of Penicillium, however, most isolates were obtained from inflorescences. Detailed identification indicated that the species from host roots differed from those isolated from parasites. These facts suggest that most endophytic Penicillium strains were acquired after the aboveground parts of B. japonica.

This interpretation did not coincide with the B. harlandii results from China, as Tu et al. [18] did not isolate fungi from flowers or leaves. We could not identify any reason for these differences, but factors, such as age and longevity of aboveground parts or differences among species, might be related. Further examinations, particularly time-series analyses of fungal infections, should be conducted.

Although the major infected parts differed from each other, Penicillium was the only endophytic fungus in common with results from the previous study of B. harlandii [18] . However, the dominance status differed between the two species. Penicillium was one of the major fungi isolated, while only 3 of the 28 identified strains were Penicillium in B. harlandii. The major fungus found in China was the basidiomycotan anamorphic fungus Rhizoctonia [18] . This difference might be caused by differences in how the hosts are used. The host of B. harlandii is Ficus spp. (Moraceae), while B. japonica mainly uses Symplocos spp. as the host [20] . The endophytic assemblage of Ficus spp. [26] [27] indicated the presence of some fractions of “sterile forms” that might correspond to Rhizoctonia. Tu et al. [18] characterized the anamorph genus as different from conidiospores. Further studies, especially exhaustive molecular identification of fungi, are needed to elucidate the relationships between fungi and hosts in different parasitic plants.

The only example of a root holoparasite, other than Balanophoraceae, was Rafflesia cantleyi [14] . However, the results of Rafflesia were completely different from those of Balanophoraceae. Major isolates were Colletotrichum spp. that were never isolated from Balanophora (Tu et al. [18] and this study). Colletotricum is also a major endophytic fungus isolated from various plants [28] [29] . This fact might be caused by differences in host taxa (Tetrastigma [Vitaceae] is a host of Rufflesia spp.), distribution range (tropical vs. temperate evergreen forests) and/or phylogenetic position (Malpighiales vs. Santalales). Additional studies are required to identify the factors that contribute to these differences through an investigation of fungal composition of other holoparasitic species.

In this study, we could not identify roles for endophytic fungi of B. japonica in relation to its parasitic lifestyle. High similarity at the generic level indicated that B. japonica might obtain beneficial effects similar to those of host plants. In particular, the predominance of Trichoderma-Hypocrea might affect root growth stimulation [5] [6] and, thus, is advantageous to both hosts and parasites. Trichoderma-Hypocrea was not isolated in this study. This variation among B. japonica individuals, in terms of fungal composition, might provide a good opportunity to test the effects of endophytes. We only found a hint of tripartite relationships among host plants, parasitic plants and endophytic fungi. Further study is required to elucidate these relationships and the roles of each organism.

We thank Y. Kumekawa, K. Matsuyama, Y. Muramatsu, M. Muroi, K. Ohga, and N. Yokoyama for their help in the course of this study. This study was partly supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, JAPAN (KAKENHI: No. 26291076 for J.Y. and T.F.)

The English in this document has been checked by at least two professional editors, both native speakers of English. For a certificate, please see: http://www.textcheck.com/certificate/A8VLeD.

HidekoIkeda,TatsuyaFukuda,JunYokoyama,11, (2016) Endophytic Fungi Associated with a Holoparasitic Plant, <i>Balanophora japonica</i> (Balanophoraceae). American Journal of Plant Sciences,07,152-158. doi: 10.4236/ajps.2016.71016