Conventionally there are two methods of expressing amount of phytoplankton. Number of individuals by species is often used for community composition and amount of components such as chlorophyll and carbon for function of community. The former is essential to study species succession including blooming patterns of specific species. For example, observation of change of cell numbers in relation to environmental parameters is necessary to clarify blooming mechanisms of harmful microalgal species. This is traditionally done by light microscopy (LM). But a problem arises in some new species of phytoplankton as they are classified based on ultra structures visible only by electron microscopy. This is evident in diatoms such as Pseudo-nitzschia, a genus which contains some species known to produce toxin causing Amnesic Shellfish Poisoning (ASP). Species in this genus are currently delineated based on morphological features visible only by electron microscopy. As of this year 2009, there are 32 species of the genus Pseudo-nitzschia, 11 of which are known to produce ASP causing toxin. The objective of this study is to establish quantification protocols for Pseudo-nitzschia for ecological studies and also for monitoring purposes. The protocols developed must be effective, practical and can answer to different situations such as the wide variety of research objectives and availability of facilities among others. In this study, identification groups to facilitate inference of Pseudo-nitzschia species under LM and species count under TEM were developed based on morphological characters of species visible under LM. Fluorescence in situ hybridization (FISH) method was developed especially using newly designed probes and new fluorescent labels. A novel protocol is proposed based on combination of improved molecular (FISH) and morphological attributes (groups).

Pseudo-nitzschia samples to test protocols were collected by 20 μm mesh size plankton net and/or Van Dorn sampler and/or bucket from Tokyo Bay at Odaiba and/or Chiba sites from April 2008 - December 2009, San Pedro Bay, Philippines from December 2006 - May 2008. Samples were also collected from four other bays in Japan, and Cat Ba Bay, Vietnam. 12 species among 105 strains in cultures of Pseudo-nitzschia were established for development of FISH.

To develop protocol 1, lengths (L) widths (W), ratio of lengths versus widths (L/W), tip and cell shapes among others were collected from original descriptions of all species. L, W and L/W were graphed to establish parameters for groups of species. Morphologically similar species based on these characteristics which are discernable under LM were grouped. Groupings were tested on field samples. Cell counts of Protocol 1 were compared with that of protocol 2. Readjustments of some groups were done after a series of comparisons were made. Protocol 2 was developed to assess results of protocols 1, 3 and 4. Number of striae and fibulae in 10μm, number of poroids in 1 μm, rows of poroids, inner poroid features were examined by TEM for critical species identification. A way to count Pseudo-nitzschia species under TEM that will include species occurring in low densities is devised. In protocol 3, FISH method was tested. Hybridization time and temperature, time of fixation, washing conditions, growth conditions of population and proper cell density for counting were experimented to find the optimum conditions using cultures prior to field application. Newly designed probes for FISH including one genus targeted probe and nine species targeted probes, which were based on D1-D3 region of LSU rDNA sequences were designed and used. All probes were tested by cross reaction checks using cultures. Experiments to develop protocol 4 involved examinations of efficacy of 10 fluorescent dyes and selection of combination of dyes that did not exhibit cross talk.

Four protocols were developed. Protocols 1 and 2 are morphologically based. Protocol 3 is molecularly based. Protocol 4, the most advanced of the three protocols is a combination of both morphological and molecular techniques.

Protocol 1 involves counting of morphologically similar species by groups under LM by the following steps: 1. Sample collection; 2. concentration of fixed samples; 3. observation by cell and tip shapes, measuring L, W and comparison to groupings flowchart; 4. identification of groups which the cells belong; 5. counting of cells by groups; 6. acquisition of results of cell number of groups of Pseudo-nitzschia.

The seven groups with some species belonging to two groups as follows:

AME (P. americana group; contains 3 species): no bulge in the center; L not 10 or more times longer than W.

AUS (P. australis group; contains 14 species): bulging central W; with L 10 or more times longer than W; about more than 50μm in L.

CAC (P. caciantha group; contains 11 species): straight species with no bulge at the center; elongated with L 10 or more times longer than W.

GAL (P. galaxiae group; contains 6 species): L 10 or more times longer than W; less than 50μm in L.

MICRO (P. micropora group; contains 2 species): bulging centers; L are not 10 or more times longer than W.

MULTA (P. multistriata ; contains 1 species): cells with sigmoid shape and curving tips in girdle view.

SUBC (P. subcurvata ; contains 1 species): cells are curved, bulging central W abruptly constricted towards the tip.

Protocol 2 involved observing, classifying and counting species by groups established for LM quantification as in Protocol 1, steps 1 to 5. Then additional steps are as follows: 6. isolation of cells by groups and placing these in tubes designated for each group; 7. verification of species in groups by TEM; 8. counting of all Pseudo-nitzschia species found in TEM grid; 9. extrapolation of species cell density by multiplying relative abundance with total Pseudo-nitzschia cell number from LM; 10. acquisition of results of cell number of Pseudo-nitzschia species.

The two morphology based quantification protocols described above, although limitations abound, may still be used in unique circumstances. Protocol 1 maybe used in cases when only rough identification of Pseudo-nitzschia is required such as in preliminary identification during field surveys. Species verification by TEM may be done at a later time. Protocol 2 is the most accurate method in species identification. Counting is possible. But picking up about 104 cells and acid washing is time consuming. Results will be acquired only after several days. Protocols 1 and 2 will be useful for bays with less diverse composition of Pseudo-nitzschia e.g. San Pedro Bay more so when bays have occurrence of high densities of Pseudo-nitzschia but low species diversity scattered in different groups. A bloom of similar composition just described will also find this method helpful.

The seven groups in Protocol 1used other characteristic aside from L and W such as general shape of the cell and tip and ratio of L/W. This is more useful than the currently practiced counting of Pseudo-nitzschia in two groups delineated only by 3μm of W since W varies in many strains within species. Species are spread into seven groups thus there is higher possibility to infer species correctly under LM. Results however will be expressed as "groups" of Pseudo-nitzschia only. Results of Protocols 1 and 2 correlated for P. multistriata very highly significant r=0.982928, AUS significant r=0.730685 and CAC very highly significant r=0.785045 with samples from Odaiba Site in Tokyo Bay.

General steps for protocol 3 are as follows: 1. counting of number of Pseudo-nitzschia by LM; 2. Filtration of amount of seawater to achieve at least 300-600 cells/ml of Pseudo-nitzschia; 3. processing of samples by FISH method using single colored probe; 4. observation and counting in epifluorescent microscope noting cellular size and shape by groups established from protocol 1 to check if cross reactions occur; 5. acquisition of results of cell number of Pseudo-nitzschia species. Only one probe out of five probes by Miller and Scholin, frD1 for P. fraudulenta was found to be a good probe for the same species in Tokyo Bay. On the contrary designed probes in this study worked well. Three out of nine designed probes had cross reactions, however by calculations the cell number in each species were obtainable. The nine species-specific probes were for the following species: P. americana, P. brasiliana, P. caciantha, P. calliantha, P. delicatissima, P.galaxiae, P. multistriata, P. multiseries and P. pungens respectively. A probe for the genus Pseudo-nitzschia, P-n2 could catch the species or strains in the bay that were not designed probes.

Protocol 4, the final protocol named group guided multicolour FISH (g-MFISH) is the result of development of three previous protocols. This provides a novel technique of simultaneous observation of both morphological and molecular attributes in a species. This follows the steps described in protocol 3. Except that application of probes involves uniformly labelling members of one group, described in Protocol 1 with one dye color. Members of another group were separately labelled with another dye color. One sample aliquot was applied with probes of member species from different groups. This protocol combines differences of sizes and shapes that reflect morphological variations and the different colors reflect molecular variations in each species. As the case in Tokyo Bay, nine species with developed probes fitted to a three-colored system i.e. three species in three separate groups. The following settings of Olympus IX71 epifluorescence microscope were used to observe the following fluorescence: NIBA3 (green bandpass) for viewing FITC or AF488 (green dye), Filter Cy3 (red band pass filter) for viewing Cy3 or AF568 (red dye) and WU2 (blue UV long pass) for viewing AF350 (blue dye). Cross reaction problems particularly with P. pungens and P. multiseries were solved with this strategy. Another advantage of this method from ordinary FISH is the decrease in time and materials spent in processing samples. One limitation was that the third dye either exhibits cross talk or had lower percentage recovery (Alexa Fluor 350).

Correlations of data by protocols 2 and 3 for total Pseudo-nitzschia and cells reacting with genus probes were always at very highly significant r= 0.986799, for example in Odaiba Site of Tokyo Bay. There is always possibility to find species which could not be detected in the efforts of group component species analysis. Sometimes species will be found in a research area that was not encountered before. Therefore observation of difference between total cell number detected by specific probes and detected by genus probe is important. Correlations of data for species P. calliantha significant r=0.589945, P. fraudulenta significant r=0.612067, P. multistriata very highly significant r=0.98769, and P. multiseries significant r= 0.681602 with samples from the same site were also achieved. Except for some probe cross reaction and non reaction problems, results of Kamaishi, Okirai, Ofunatu, Kobe and Cat Ba Bay samples by the three protocols described above matched accordingly. Thus probes can trace species in a bay when these are designed based on the same species or strains in the same bay. Protocols 3 and 4 have an advantage of getting results in a shorter time than Protocol 2. The process takes only about five hours to finish.

Application of g-MFISH or FISH using single colored probes will depend on the variability, number and composition of expected Pseudo-nitzschia species in the bay. If there is less or equal to five species in separate groups with no cross reaction of probes among species, a single color of the probe will suffice. Observation by groups should be made to ensure that no cross reaction of unexpected species occurred. Protocol for g-MFISH is useful if simultaneous enumeration of all Pseudo-nitzschia species is necessary in a bay with diverse Pseudo-nitzschia species concentrated in three groups or less with probes cross reacting to species within groups such as Pseudo-nitzschia species composition in Tokyo Bay.

G-MFISH can be applied to ecological studies of individual species with similar difficulty of identification as Pseudo-nitzschia such as many other diatoms e.g. Skeletonema or dinoflagellates e.g. Alexandrium. Data verification should still be made using electron microscopy. Future studies may be directed to improving probe specificity by searching for sequences in other domains of the LSU and advancing g-MFISH by testing other dyes or developing new dyes specific for phytoplankton. This strategy may also be applied to designs at automation of species enumeration.

珪藻は海洋の微小プランクトンの主要な一群であり、わが国沿岸はじめ世界の海洋に広く分布して、海洋の基礎生産を担う重要な分類群であるが、一部の属種は赤潮や貝毒の原因となる有害種とされている。特に、浮遊珪藻Pseudo-nitzschia(以下PN)属の11種は、脳神経細胞を不可逆的に破壊するドーモイ酸というアミノ酸の一種を生産し、記憶喪失性貝中毒の原因となるため、その生態や生産毒の貝への蓄積機構は世界各地で研究されてきた。しかし、PN属に属する32種はほとんどが長さ10- 200μmで、幅はその1/10以下という極めて細い針状の外形をもち、電子顕微鏡でのみ観察可能な形態形質による分類がされているため、天然に発生する種の迅速な定量分析ができず、生態研究の大きな妨げとなっていた。このような背景のもと、本研究はPN属を形態および遺伝子塩基配列を基にした特徴を用いて、種の同定計数手法の開発を目指し、生態研究に役立てようとしたものである。

研究材料としては東京湾奥部とフィリピン中部のレイテ島サンペドロ湾で、孔径20μmの目合いのプランクトンネットあるいはヴァンドーン採水器により2006年から2009年の間に定期的に採集した試料を主に用いた。これら試料は、光学顕微鏡あるいは電子顕微鏡下で形態を観察し、原記載など過去の形態観察例と比較して、種を査定した。さらに、これらの種の培養株を作成してLSU rDNAの塩基配列を決定し、プローブ作成に供した。培養株の作れなかった種は過去の文献などの情報を元にプローブを作成した。また、原記載を含む過去の形態観察結果を参考にPN属の32種を、光学顕微鏡で区別ができる6つのグループにわけ、種の同定のための識別検索の基準とした。そして、本研究では、PN属の同定手法は用いることのできる観察機器によって異なるということを基本的な考え方とし、光学あるいは電子顕微鏡による形態観察と、LSU rDNAの遺伝子塩基配列に基づいて作成されたプローブを組み合わせて、4種の観察手法を考案した。

まず第一法として、光学顕微鏡のみを用いて試料の形態を観察し、その結果を識別検索グループにあてはめ、種を同定する方法を考案した。この方法は、発生する種数が少ないサンペドロ湾のような海域の試料には応用可能で、種ごとに計数することができたが、東京湾のように11種が発生する海域の試料の定量には不十分であった。このため第二法として、第一法に基づいて分けた細胞を識別検索グループごとに異なる容器に移し入れて、電子顕微鏡下で種の査定をしながら計数することにしたところ高い再現性が認められ、実用的と考えられた。ただしこの方法は、種ごとに万単位の細胞数を分取して電子顕微鏡観察に供する必要があり、結果を得るまでに時間と多大な労力を要する欠点があった。第三法は、FITC標識した種特異性の高いプローブを用いたFISH法により蛍光顕微鏡下で計数して、同時に第一法の形態情報と合わせて種ごとに確認する方法を考案した。同じ試料を第三法と第二法で計数した結果はよく一致し、遺伝子塩基配列と形態観察結果の組み合わせによって、第二法に比べて短時間で種別に計数することが可能になった。またPN属全体をターゲットにしたプローブを作成して、これまで報告のない種が出現した場合にも対応可能になった。ただし、作成したプローブには交叉反応するものがあること、また1種ごとに試料を作成する必要があり、第一法に比べて試料の準備と観察に時間を要した。そこで、さらにこれらを改良し、第四法として、第三法で用いたプローブに3種類の発色の異なる蛍光標識を結合させ、これらを同時に用いることにより、交叉反応の問題を解決し、観察時間を削減させた。以上のことから、これら4種の観察手法が有用であることがわかった。

以上本研究は、その困難さから研究の立ち遅れていたPN属珪藻類に関し、過去に記載のある32種の詳細な形態観察結果と遺伝子塩基配列を基に、現場で使用できる簡易な同定手順を、必要とされる研究精度に応じて選ぶように4方法考案したものである。この成果はPN属に限らず、光学顕微鏡では同定困難な種を含む、多くの珪藻の属に応用できるものであり、生態研究における実用的同定手法が作られたものともいえる。すなわち、有毒種を含めたPN属の生理生態研究において研究材料の同定の標準化を諮ることが容易になり、水圏生物科学の発展に資するところが極めて大であると考えられた。よって、審査委員一同は本論文をもって、本研究の申請者デジエット レニ ヤップさんが博士(農学)の学位を授与するに値する研究者であると判断した。