Acta Protozool. (2002) 41: 309 - 342 A Century of Research on the Amoeboflagellate Genus Naegleria Johan F. De JONCKHEERE Protozoology Laboratory, Scientific Institute of Public Health, Brussels, Belgium Summary. The amoeboflagellate genus Naegleria contains pathogenic and nonpathogenic species. As most species are morphologically indistinguishable, species are defined and identified by molecular methods. For routine identification, isoenzyme analyses are performed. For the description of a new species, sequences of ribosomal DNA are increasingly used and the analyses of these sequences also allow us to define the phylogenetic relationships between species and strains. In the present monograph 27 Naegleria lineages are discussed and identified as separate species. Using molecular methods, Naegleria spp. have been identified which either form dividing flagellates or which do not form flagellates at all, thus contradicting the accepted definition of the genus. Willaertia, which forms dividing flagellates, is the genus that is the closest relative of the genus Naegleria. The genus Naegleria has some particularities in its molecular biology, such as circular ribosomal DNA plasmids, group I introns in the small and large subunit ribosomal DNA, and an unusual pyrophosphate-dependant phosphofructokinase. The phylogeny of the Naegleria spp. is compared to the situation concerning the other genera of the familyVahlkampfiidae. Also discussed is the state of affairs concerning species designation based on phylogeny in the genus Acanthamoeba, another free-living amoeba with species pathogenic to man. Key words: Acanthamoeba, Balamuthia, dividing flagellates, group I introns, Hartmannella, Naegleria pagei sp. n., N. pringsheimi sp. n., N. tihangensis sp. n., non-flagellating, phylogeny, ribosomal DNA, Vahlkampfia. Abbreviations used: AP - acid phosphatase, ATCC - American Type Culture Collection, bp - basepairs, CCAP - Culture Collection of Algae and Protozoa, CSF - cerebrospinal fluid, EMBL - European Molecular Biology Laboratory, IC - intracerebral, IN - intranasal, ITS - internal transcribed spacer, LSU - large subunit, mt - mitochondrial, NACM - Naegleria amoebae cytopathogenic material, NRS - non-ribosomal sequence, ORF - open reading frame, PAM - primary amoebic meningoencephalitis, PE - propionyl esterase, PCR - polymerase chain reaction, PPi-PFK - pyrophosphate-dependant phosphofructokinase, RAPD - random amplified polymorphic DNA, rDNA - ribosomal DNA, RFLP - restriction fragment length polymorphism, SSU - small subunit. INTRODUCTION A century ago Schardinger (1899) discovered an Amoeba. lobosa that could transform into a flagellate stage, and called it Amoeba gruberi. The genus name Naegleria was suggested much later by Alexeieff (1912). Before 1970 Naegleria was studied mainly as a model for transformation because the amoebae easily transform into flagellates (Willmer 1956, Chang 1958, Fulton and Dingle 1967). However, the genus attracted much more attention, especially from the biomedical world, when it was found that some Naegleria isolates cause a fatal brain infection, primary amoebic menin- goencephalitis (PAM) in humans. The infection almost invariably results in death. The Naegleria isolates that Address for correspondence: Johan F. De Jonckheere, Research Unit for Tropical Diseases (TROP), Christian de Duve Institute of Cellular Pathology (ICP), Avenue Hippocrate 74-75, B-1200 Brus- sels, Belgium Fax: 32-2-762.68.53 E-mail: jdjonckh@ben.vub.ac.be

310 J. F. De Jonckheere cause PAM were given species status and named N. fowleri, after Malcolm Fowler who first recognized the disease in Australia (Carter 1970). Cases of PAM were soon afterwards detected all over the world. The most recent review on the diseases produced by N. fowleri and other opportunistic free-living amoebae, belonging to the genera Acanthamoeba and Balamuthia, can be found in Martinez and Visvesvara (1997). Between October 1970 and October 1972 four PAM cases were diagnosed in Belgium in children aged between 11 and 14 years old (Hermanne et al. 1973). All cases were diagnosed around the city of Antwerp. Although N. fowleri was never isolated from the swim- ming pools where the four children had been swimming before becoming ill, it was presumed that these swim- ming pools were implicated. Because the swimming pools are filled with water from the drinking water suppliers, the latter were accused of introducing the pathogenic amoeba into the swimming pools. Therefore, the Belgian water distributors decided to have the water of their network investigated for the presence of N. fowleri. This is where my involvement with Naegleria started. As a young biologist, with no experience in protozoa at all, I was hired to work on the grant that was assigned to a university laboratory, that also had no experience in protozoology whatsoever. It was quite a coincidence that a new case of PAM was diagnosed in Belgium (Van Den Driessche et al.1973) only six months after I started to investigate methods for the identification and isolation of N. fowleri. Not only was the timing useful, the place where the 14 year-old boy probably became infected gave a clue to the ecology of the pathogenic N. fowleri. The deceased boy had been swimming in a brook that received cooling water from a metallurgical factory, and we were able to demonstrate the presence of N. fowleri in that water (De Jonckheere et al. 1975). Therefore, an investigation was started on the presence of N. fowleri in warm water discharges of different industries in Belgium. This investigation demonstrated that cooling waters were indeed the places where this pathogen could proliferate (De Jonckheere and van de Voorde 1977a). In addition, the absence of N. fowleri in drinking water and swim- ming pools in Belgium was also subsequently proven (De Jonckheere 1979a). This is in contrast to the situation in other parts of the world where higher annual water temperatures are prevalent and the presence of N. fowleri in drinking water is not uncommon (South Australia is the most notorious example Dorsch et al. 1983). In the attempts to isolate N. fowleri, many different Naegleria strains were isolated that have properties that did not fit the descriptions of either N. gruberi (non- pathogenic) or N. fowleri (pathogenic), the only two species described at that time. At the time of writing this monograph, 20 Naegleria spp. have been fully de- scribed. Three strains are given species status here, and a few more descriptions are in preparation (Table 1). It is mostly due to the use of molecular biology techniques that species descriptions are possible in a genus where morphology is not discriminative. However, this latter statement may now have to be reconsidered. Until recently all vahlkampfiids with dividing flagellates had been classified in genera other than Naegleria. It has been found that some amoeboflagellates whose flagellates can divide (Dobson et al. 1993, B. Robinson personal communication) are in fact Naegleria spp. (De Jonckheere and Brown 1995, 1999b). In addition there are two Naegleria strains that do not form flagel- lates under laboratory conditions (De Jonckheere et al. 2001), and a few N. fowleri strains from one location in France have never formed flagellates. A Naegleria strain also exists that fails to form cysts, but this seems to be due to the presence of a bacterial parasite (Michel et al. 2000). Infection of other Naegleria strains with the parasite impaired their capacity to form cysts. This bacteria did not interfere with the transformation to the flagellate stage. As cyst morphology is informative for identifying amoeboflagellate genera, and because the bacteria also infects other genera of amoebae, the investigation of whether the originally-infected amoeboflagellate does indeed belong to the genus Naegleria is recommended. In 1988 the definition of the Naegleria genus was: these are vahlkampfiids whose flagellate stage normally has two flagella, lacks a cytostome, and does not divide. The cysts have plugged pores through which the amoeba excysts (Page 1988). Although the statement about the plugs in the cysts remains valid the rest of the definition of Naegleria should be emended as follows: these are vahlkampfiid amoebae with a temporary flagellate stage in most species, but lacking or difficult to induce in some species and in individual strains of others. Where present, the flagellate stage lacks a cytostome, is usually biflagel- late and incapable of division. In at least two lineages, flagellates initially have four flagella and divide once to

The amoeboflagellate Naegleria 311 form typical biflagellate cells. Naegleria can only be identified to species level by biochemical and molecular techniques. MATERIALS AND METHODS Culture Established cultures of Naegleria strains are grown either monoxenically on non-nutrient agar plates with Escherichia coli (Page 1988) or axenically in a liquid medium (De Jonckheere 1977). Isoenzyme analysis Protein extracts are prepared by adding 0.25% Triton X-100 to amoebae concentrated by centrifugation. The suspensions are frozen and thawned several times to make the amoebae burst. For isoenzyme analyses the proteins are separated by agarose gel isoelectric focusing (De Jonckheere 1982a) or cellulose acetate electrophoresis (Robinson et al. 1992), and the bands of enzyme activity were visualized according to procedures published by these authors. DNA sequence analysis DNA is extracted from cell pellets using either a phenol-chloro- form-isoamyl alcohol method or a guanidium thiocyanate-sarkosyl method (Pitcher et al. 1989). The small subunit ribosomal DNA (SSU rDNA), large subunit ribosomal DNA (LSU rDNA) and the internal transcribed spacer (ITS) regions, including the 5.8S rDNA, are amplified using primers and polymerase chain reaction (PCR) conditions described by De Jonckheere (1994a, 1998). In preparation for sequencing PCR products were treated with exonuclease I and shrimp alkaline phosphatase for 15 min. at 37��C. After inactivating these enzymes by heating at 80��C for 15 min., the PCR products were sequenced using the Sequenase PCR product sequencing kit (Amersham Pharmacia Biotech UK Limited, Buckinghamshire, England) using either [32P] -dATP or [33P] -dATP. SSU rDNA and ITS amplification and conserved internal primers were used (De Jonckheere 1994a, 1998). Approximately 800 basepairs (bp) Table 1. Species of the genus Naegleria Species author, year Max. ��C Flagellates EMBL* N. gruberi Schardinger, 1899, emend. De Jonckheere, this paper 39 + M18732 N. fowleri Carter, 1970 45 + U80059 N. jadini Willaert and Le Ray, 1973 35 + - N. lovaniensis Stevens, De Jonckheere and Willaert, 1980 45 + U80062 N. australiensis De Jonckheere, 1981 42 + U80058 N. italica De Jonckheere, Pernin, Scaglia and Michel, 1984 42 + U80060 N. andersoni De Jonckheere, 1988 40 + U80057 N. jamiesoni De Jonckheere, 1988 42 + U80062 N. clarki De Jonckheere, 1994 37 + - N. galeacystis De Jonckheere, 1994 35 + - N. minor De Jonckheere and Brown, 1995 38 divide X93224 N. pussardi Pernin and De Jonckheere, 1996 41 + - N. carteri Dobson, Robinson and Rowan-Kelly, 1997 45 + Y10189 N. morganensis Dobson, Robinson and Rowan-Kelly, 1997 44 + Y10188 N. niuginensis Dobson, Robinson and Rowan-Kelly, 1997 45 + Y10186 N. sturti Dobson, Robinson and Rowan-Kelly, 1997 44 + Y10185 N. robinsoni De Jonckheere and Brown, 1999 38 divide AJ237786 N. fultoni De Jonckheere, Brown, Dobson, Robinson and Pernin, 2001 35 + AJ243440 N. chilensis De Jonckheere, Brown, Dobson, Robinson and Pernin, 2001 30 - AJ243442 N. indonesiensis De Jonckheere, Brown, Dobson, Robinson and Pernin, 2001 38 - AJ243441 N. tihangensis De Jonckheere, this paper 42 + - N. pringsheimi De Jonckheere, this paper 37 + - N. pagei De Jonckheere, this paper 37 + - N. philippinensis In preparation 40 + NA WA variant N. lovaniensis In preparation 45 + Y10187 NG597 In preparation 42 + Y10184 antarctic Naegleria sp. In preparation 30 + ND * - EMBL accession N�� of SSUrDNA - - not at EMBL, but partial sequences have been published (De Jonckheere 1994a, Pernin and De Jonckheere 1996) NA - not available yet ND - not done (DNA could not be isolated because of poor growth)

312 J. F. De Jonckheere between two conserved Pst I sites within the Naegleria SSU rDNA were sequenced and used for phylogenetic analysis (De Jonckheere 1994a). Sequences of group I introns in the SSU and LSU rDNA are determined using internal rDNA and group I intron primers (De Jonckheere 1993). The nucleotide sequence data reported in this paper are available in the European Molecular Biology Laboratory (EMBL) nucleotide sequence database and the accession numbers are indicated at each species description. Phylogenetic analysis The DNA sequences are aligned by eye using the Eyeball Sequence Editor (ESEE) (Cabot and Beckenbach 1989). Phylogenetic trees are constructed from the aligned sequences using the DNAPARS (parsimony), DNADIST (distance matrix), NEIGHBOR (Neighbor joining and UPGMA), FITCH (Fitch-Margoliash), KITCH (Fitch- Margoliash with evolutionary clock) and SEQBOOT (bootstrapping) programs of the PHYLIP (version 3.572c) package (Felsenstein 1989). For phylogenetic analyses of proteins the PROTPARS and PROTDIST programs of the same package are used. MOLECULAR BIOLOGY OF THE GENUS NAEGLERIA Chromosomes and ploidy Naegleria has an intranuclear mitosis, called promitosis, following the classical pattern of chromo- some separation, but the chromosomes are too small to be counted by conventional histological techniques (Fulton 1970). However, it has been possible to enumerate the chromosomes with the use of pulsed field gel electro- phoresis. The number of chromosomes and their size differ between species and even between strains of the same species. Two strains of N. gruberi sensu lato have 23 chromosomes, but the size of some chromo- somes differ (Clark et al. 1990). These two strains are considered now to belong to two different species, N. gruberi sensu stricto and N. pringsheimi (see below). Within the species N. fowleri differences in number and size of chromosomes are observed with different isolates (De Jonckheere 1989). The ploidy of the Naegleria genome is still not known. The sum of the chromosome sizes (approxi- mately 19 Mb) does not equal the expected genome size (approximately 104 Mb), which indicates that Naegleria might be polyploid (Clark 1990). It has been demon- strated that differences in ploidy exist between strains (Fulton 1993), and isoenzyme studies of Naegleria spp. usually imply diploidy (Cariou and Pernin 1987, Adams et al. 1989). Isoenzyme studies also reveal that genetic exchange occurs in N. lovaniensis but not in other species (Pernin et al. 1992). Of course, genetic recom- bination does not mean sexuality, which involves meiosis to form monoploid cells, and karyogamy. It has been argued that the flagellates of Naegleria are gametes (Fulton 1993) and the fact that in some Naegleria spp. the flagellates divide once (De Jonckheere and Brown 1995) could be in support of monoploidy formation. However, meiosis in Naegleria has not been proven experimentally. rDNA plasmid In N. gruberi, the rRNA genes are carried exclu- sively on a 14-kp circular plasmid, and each plasmid contains only one rDNA repeat unit (Clark and Cross 1987). The number of rDNA circles per cell was estimated to be 4,000. This circular plasmid is a general feature of the rDNA genes in all the vahlkampfiids (Clark and Cross 1988a). The length of the rDNA plasmid varies according to the species and strain inves- tigated. It is not known whether different numbers of rDNA repeats per plasmid, as was found in the anaero- bic Entamoeba histolytica (Bhattacharya et al. 1998), contribute to the plasmid length differences. Length differences in the ribosomal genes themselves are mainly due to the presence of group I introns. The SSU rDNA of several species carry these introns (De Jonckheere 1994b). Length differences in the ITS1 and/or ITS2 also contribute to repeat unit size variability (De Jonckheere 1998). In a few Naegleria strains, group I introns are also present in the LSU rDNA (De Jonckheere and Brown 1998a, 2001). The rDNA plasmid of N. gruberi strain EG B has been completely sequenced. The molecule is 13,996 bp in length with an overall G+C content of 40.7% (Mullican, J. C. Molecular characterization and complete sequence analysis of the extrachromosomal ribosomal DNA ele- ment in Naegleria gruberi. Ph D thesis. The graduate College in the University of Nebraska, Omaha, Ne- braska, USA:1-163,1995). A putative open reading frame (ORF) for a heat shock protein is detected in the non- ribosomal sequence (NRS) of the N. gruberi plasmid (Mullican and Tracy 1993). No similarities were ob- served in the NRS between N. fowleri (strain LEE) and N. gruberi (strain NG B ). Another plasmid has been detected in N. minor (De Jonckheere and Brown 1995). The function of this 6.0 kb plasmid is unknown but it seems not to be involved in flagellate division as it is not found in N. robinsoni, currently the only other Naegleria species with dividing flagellates.

The amoeboflagellate Naegleria 313 Group I introns Group I introns are catalytic RNA molecules that occur within transcribed sequences and are able to self- excise. The group I intron in the SSU rDNA of Naegleria spp. is a twintron (Einvik et al. 1998), consisting of two distinct ribozymes (catalytic RNAs) and an ORF encoding a homing endonuclease with a His-Cys box (Johansen et al. 1993). Endonucleases with His-Cys boxes are uncommon (Johansen et al. 1997). A similar twintron has only been found in the myxo- mycete Didymium (Einvik et al. 1998). In one Naegleria lineage the twintron has lost the ribozyme that carries the endonuclease (De Jonckheere and Brown 1994). The group I introns in the LSU rDNA of Naegleria either carry an endonuclease or do not (De Jonckheere and Brown 1998a, 2001). In the genus Naegleria the group I intron seems to be transferred vertically in the SSU rDNA (De Jonckheere 1994b) and horizontally in the LSU rDNA (De Jonckheere and Brown 1998a, 2001). From this it is inferred that the SSU rDNA group I intron was acquired in an ancestral state and lost in most of the Naegleria spp. In all described Naegleria spp. with a group I intron in the SSU rDNA, the presence of this intron is a property of the species. Only the WA variants of N. lovaniensis could be exception to this rule. An ORF with approxi- mately 30% identity to the ORF in the SSU rDNA group I intron of N. pringsheimi has been found in the NRS of the rDNA plasmid of strain EG B of N. gruberi (Mullican, J. C. Molecular characterization and complete sequence analysis of the extrachromosomal ribosomal DNA ele- ment in Naegleria gruberi. Ph D thesis. The graduate College in the University of Nebraska, Omaha, Ne- braska, USA: 1-163, 1995). It is trancriptionally silent and may be a remnant of the group I intron that was lost from the SSU rDNA of N. gruberi. The His-Cys box is still present in the ORF in strain EG B , but comparison with His-Cys boxes in the SSU rDNA introns of other Naegleria spp. shows it is phylogenetic distinct (Fig. 1). Fig. 1. Phylogenetic tree inferred from the amino acid alignments of the His-Cys box in the ORF of the group I introns Fig. 2. Phylogenetic tree inferred from partial SSU rDNA of species in the genus Naegleria and its closest relative Willaertia magna