Broad infectivity of Leidynema appendiculatum (Nematoda: Oxyurida: Thelastomatidae) parasite of the smokybrown cockroach Periplaneta fuliginosa (Blattodea: Blattidae)

Abstract Host specificity of parasites is important for the understanding of evolutionary strategies of parasitism that would be a basis of predictions of the disease expansion when parasitized hosts invade new environments. The nematode order Oxyurida is an interesting parasite group for studying the evolution of parasitism as it includes parasites of both invertebrates and vertebrates. In our survey, we found that the smokybrown cockroach Periplaneta fuliginosa was primarily infected with only one nematode species Leidynema appendiculatum. In two cases, L. appendiculatum was isolated from two additional cockroach species Pycnoscelus surinamensis, sold in Japan as a reptile food, and Blatta lateralis, captured in the field and cultured in the laboratory. Inoculation of L. appendiculatum into three additional cockroach species P. japonica, Blattella nipponica, and P. surinamensis also resulted in parasitism. Infection prevalence was high, and timing of postembryonic development from hatched nematode larva to mature adult in these hosts was identical with that in P. fuliginosa. While ecological interactions strongly determine the host range, such broad infectivity is still possible in this parasitic nematode.


| INTRODUC TI ON
The host specificity of parasites is important as it reflects the evolutionary strategy of parasitism. The nematode order Oxyurida consists of a number of families that are divided into two superfamilies.
The first superfamily Oxyuroidea, which consists of vertebrate parasitic nematodes (also called pinworms that include the human parasite Enterobius vermicularis), and second Thelastomatoidea, which consists of invertebrate parasitic nematodes (Carreno, 2014;Nadler et al., 2007). Nematodes belonging to the order Oxyurida share a simple infection cycle: Nematode eggs laid by adult females are deposited within their host feces and released by defecation, next ingestion of eggs by new host individuals leads to infection (Adamson, 1994;Ozawa et al., 2016). While Oxyurida parasites sometimes cause diseases in humans, such as colitis, perianal abscess, and ectopic infections in females (Ariyarathenam et al., 2010;St Georgiev, 2005), these conditions are not regarded as serious compared to infectious diseases caused by another gastrointestinal worms including Onchocerciasis, Elephantiasis, and Ancylostomiasis (Bungiro & Cappello, 2004;Pilotte, Unnasch, & Williams, 2017). Moreover, Oxyurida parasites in invertebrates are seemingly harmless to their hosts (Adamson, 1994;Ozawa et al., 2014), but consequences of carriage are largely unknown. Cospeciation is well conserved in vertebrates and pinworms (Falk & Perkins, 2013;Hugo, 1999;Murata,
Here, we show the broad infectivity of L. appendiculatum through natural and artificial infection of five cockroach species within three families and two suborders. L. appendiculatum was basically isolated from P. fuliginosa collected in three area of Japan with high prevalence. Infection prevalence, intensity, and infrapopulation of L. appendiculatum in five host cockroach species were similar. Moreover, the developmental timing of this nematode from hatched larvae to matured adult was identical. While ecological interactions with P. fuliginosa seemingly strongly determine the host range of L. appendiculatum, such broad infectivity is still possible in this parasitic nematode.
We used two independent strains of P. fuliginosa: P. fuliginosa EE strain was supplied by the Earth Environmental Service (Ako, Hyogo Prefecture, Japan) and reared since 2012. The P. fuliginosa UF strain, also laboratory-culturing strain, has been established and reared for more than 60 years in the Philip G. Koehler laboratory, University of Florida. P. japonica, the Japanese cockroach or Yamato cockroach, is a native of Japan and belonging to the same Periplaneta genus with P. fuliginosa. The P. japonica Miyoshi strain was established from 10 adult males and 10 females collected at the Miyoshi city (Aichi Prefecture, Japan) in June 2012. The P. surinamensis Pet strain, sold for pet reptile food, was obtained by Japanese online shop (no detailed information obtained). The P. surinamensis Yaedake strain was established from three adult females hand-picked at the Yaedake (Okinawa prefecture, Japan) in March 2013. B. nipponica is also a native of Japan. The B. nipponica CU was collected at the Chubu University campus (Kasugai, Japan) during April to September 2014 and 2017; because this species was impossible to be reared in laboratory for over a year, all individuals were collected and reared only temporary before experiments. The B. lateralis KX strain was established from three adult females collected at Knoxville (Tennessee, USA) in July 2013.
Nematode-free cockroach strains were established as follows: oothecae oviposited were collected, surface wiped with 70% ETOH, and kept in a plastic dish until hatching. Hatched nymphs were transferred into the plastic cage and reared. P. japonica Miyoshi, B. nipponica CU, and P. surinamensis Yaedake were originally nematode-free.
Individuals of the established nematode-free strains were often dissected to confirm the absence of parasitic nematodes.

| Cockroach dissection and parasitic nematode observation
The hindgut of cockroach adult male and female was extracted, placed in a Syracuse watch glass containing cockroach Ringer's solution (NaCl 9.32 g, KCl 0.77 g, NaHCO 3 0.18 g, NaH 2 PO 4 0.01 g, CaCl 2 0.5 g in 1 L of distilled water), and split longitudinally with tweezers to release nematodes. Nematode species, sex, and stage were observed under a stereomicroscope (SMZ600, Nikon, Japan).
When detailed morphological information was needed, nematodes were picked up with a mouth pipette, transferred onto an agar pad (Shaham, 2006), covered, and sealed with a silicon grease-rimmed coverslip for viewing by Nomarski DIC optics (Eclipse E600, Nikon).
Pairwise comparisons of % differences (D) between each sequence combination were performed using the formula D = (M/L) × 100 (Chilton, Gasser, & Beveridge, 1995), where M is the number of alignment positions at which the two sequences have a base in common, and L is the total number of alignment positions.

| Artificial infection test of parasitic nematodes
Artificial infection experiments were performed in four nematodefree cockroach strains, P. fuliginosa EE, P. japonica Miyoshi, P. surinamensis Yaedake, B. nipponica CU. L. appendiculatum adult females were collected from nematode-infected P. fuliginosa EE as described above. The transferred L. appendiculatum gravid females were placed in a Syracuse watch glass containing cockroach Ringer's solution and cut in the center by sterilized scalpel blade to release eggs (~150 eggs per individual). Released eggs were kept in the cockroach Ringer's solution for 14 days at 25°C until all reached to the L2 resting stage. The last instar of nematode-free cockroaches was fasted for 3 days before artificial infection. Ten fasted cockroaches were reared in a cylindrical plastic case (130 mm diameter × 225 mm height) with 0.4 g of the bait mixed with about 500 of the L2 resting stage nematode eggs. After artificial infection, cockroaches were fed with bait without nematode eggs ad libitum. To observe the nematode prevalence (% of the infected cockroaches among all cockroaches examined), intensity (mean number of nematodes in the infected cockroaches, excluding the number "zero" of uninfected host), and developmental stages, one or two cockroaches of each group from P. fuliginosa EE were dissected 1, 3, 7, 14, 21, 28, and 35 days after infection, or one or two cockroaches of each group from P. japonica, B. nipponica, and P. surinamensis were dissected every week (from 7 days until 35 days after infection). All nematodes isolated in these experiments were mounted on the agar pad (described above) and observed by Nomarski DIC optics (Eclipse E600, Nikon).
After experimental infection, the rest of P. japonica was pooled and reared for several generations, and then checked the infrapopulation of L. appendiculatum male, female, and juvenile by the Quantitative Parasitology 3.0 (Rózsa et al., 2000).

| Two exceptional cases of L. appendiculatum hosts
We dissected 14 species of cockroaches collected in the field, and as well laboratory-cultured strains. Basically L. appendiculatum were isolated only from P. fuliginosa in our experiments. We always isolated only L. appendiculatum from laboratory-cultured (Ozawa et al., 2014) and field-captured (Table S1) P. fuliginosa with high prevalence. In this study, we showed that the two different laboratory strains of P. fuliginosa EE (Japan) and UF (USA) were infected with L. appendiculatum with 100% prevalence (Table 1). As these two P. fuliginosa strains were established independently and maintained for a long period in different countries and laboratories, we conclude that this parasitic association of L. appendiculatum with P. fuliginosa is quite stable.
Interestingly, we also found two rare cases in the cockroach species P. surinamensis and B. lateralis. We established three strains of P. surinamensis; the first strain P. surinamensis Pailand was infected with unknown nematode species Suifunema sp. with almost 100% of prevalence (data not shown). The second strain, P. surinamensis Yaedake, was free from nematode, but the third P. surinamensis Pet was infected with L. appendiculatum with 100% prevalence. We established one strain of B. lateralis KX and found that was also infected with L. appendiculatum with 100% prevalence.
We confirmed that these nematodes were L. appendiculatum from morphological characteristics of the adult male and female (data not shown).

| Molecular identification and phylogenetic analysis of L. appendiculatum
We obtained ribosomal DNA sequence data, D2/D3 LSU and SSU from L. appendiculatum isolated from cockroach strains used in these experiments including P. fuliginosa CU (Ozawa et al., 2014), P. fuliginosa EE, P. fuliginosa UF, P. surinamensis Pet, and B. lateralis KX. We also obtained sequence data for L. appendiculatum Tokyo 20140825SB, isolated from P. fuliginosa captured in Minato city, Tokyo in 2014 (Table S1).
These data were deposited in the NCBI GenBank (accession numbers, Table 2). Several published sequences of Leidynema spp. were included in this phylogenetic analysis. All SSU sequences of L. appendiculatum were 100% identical (Figure 1a), whereas D2/D3 LSU sequences showed unique polymorphisms that divided into two clades. One clade contained L. appendiculatum isolated from cockroaches in Japan.  Table 2). Pairwise sequence differences between the two clades were 1.5%, although 0.0% within the clades (Table 3).

| Infrapopulation of L. appendiculatum in different cockroach strains
We analyzed the prevalence, intensity, and infrapopulation structure of L. appendiculatum males, females, and juveniles in four cockroach strains: P. fuliginosa EE, P. fuliginosa UF, P. surinamensis Pet, and B. lateralis KX. All cockroaches in the laboratory-cultured strains were infected with L. appendiculatum with 100% prevalence. We also confirmed the infrapopulation structures were similar in each cockroach strain. First, the distributions of female and juvenile nematodes are positive skewed (Mean > Median); second, the number of nematode males was basically one or zero; third, the mean and median intensity of nematodes in female hosts is higher than that in male hosts; and fourth, there was no difference in nematode prevalence between male and female cockroach hosts (Table 1).

| Artificial infection of L. appendiculatum in P. fuliginosa
Next, we inoculated L. appendiculatum into P. fuliginosa to monitor nematode prevalence and developmental timing from the L2 stage

| Artificial infection of L. appendiculatum in other cockroaches
Next, we inoculated L. appendiculatum into the cockroach species, P. japonica, B. nipponica, and P. surinamensis, which belong respectively to three families (Blattidae, Blattellidae, Blaberidae) and two suborders (Blattoidea, Blaberoidea). We observed the nematode number, developmental stage, and sex. Figure 3b-d shows the infection prevalence of L. appendiculatum (%) and mean intensity (uninfected "zero" data were excluded) of male, female, and juvenile nematodes in a cockroach hindgut at 7 to 35 days after infection.
Although average nematode prevalence over the course of the study was lower in P. japonica (65%, N = 40) (Table 1).

| D ISCUSS I ON
We Onychostylus pallidiolus, Lobopterella dimidiatipes, Opisthoplatia orientalis, Panesthia angustipennis spadica (data not shown). These species were collected in the field as well as cultured in the laboratory.

Although we have not finished identification and description of all
parasitic nematodes yet, host specificity of the nematodes belonging to the family Thelastomatidae was seemingly high. Species combinations of cockroach host and parasitic nematode were almost fixed except for two exceptions described in this work. We always isolated L. appendiculatum from P. fuliginosa with high prevalence.
L. appendiculatum was not isolated from the other cockroaches with the two exceptions; one of our three P. surinamensis strains and one of our one B. lateralis strain (Table 1) In the case of B. lateralis, we only checked one strain for these experiments. However, we did check more than ten individuals of wild P. surinamensis as well as laboratory-cultured strains. All parasitic nematodes isolated from P. surinamensis were unknown species Suifunema sp. (data not shown) except for one laboratory-cultured strain P. surinamensis Pet (Table 1). As this cockroach reproduces parthenogenetically, grows, and reproduces quickly, it is easy to maintain without any special condition.
Therefore in Japan, P. surinamensis is popular for pet reptile breeders as a food of their animals. As P. fuliginosa is the most widely distributed cockroach in Japan, there are likely many chances for contamination with the parasitic nematode from P. fuliginosa if the cockroaches are cultured in environments of general households or pet shops.
We present here that L. appendiculatum was capable of infecting three cockroach species, P. japonica, B. nipponica, and P. surinamensis. The infectivity of L. appendiculatum is quite broad and includes five cockroach species, three families in two suborders ( Figure 4).

P. japonica is a Japanese domestic species and lives mainly in north-
ern area moving between indoors to outdoors (Tanaka & Tanaka, 1997;Tanaka & Uemura, 1996). P. surinamensis is estimated to have originated in the Indo-Malayan region and is now distributed worldwide in tropical and subtropical regions (Kramer & Brenner, 2009).
We found P. japonica and P. surinamensis were infected with specific nematode Protrellus sp. and Suifunema sp., respectively, with high prevalence (data not shown). B. nipponica is a Japanese domestic and grassland of southwestern Japan (Asahina, 1991;Tsuji, 1985).
Interestingly, B. nipponica is always free from parasitic nematodes.
We also confirmed that L. appendiculatum, when inoculated artificially into B. nipponica could persist in the host hindgut and lay eggs but could not be re-infected. This is likely because B. nipponica does not eat feces. An infection cycle for parasitic nematodes might not be established. This result matches with our data that B. nipponica captured in the field is not infected with parasitic nematodes at all. Therefore, we conclude that L. appendiculatum is not a native parasite of B. nipponica. It was surprising that L. appendiculatum can infect a cockroach species that has a different ecological and systematic position than P. fuliginosa.
Although infection prevalence of L. appendiculatum within these three cockroaches was less than that in P. fuliginosa, values exceeded 50% and the developmental timing of nematode maturation resembled that seen in P. fuliginosa (Figure 3). We believe that there was no difference in infection opportunities as all cockroaches consumed all of the given bait within a few days. These results might suggest that L. appendiculatum has evolved as a parasite of P. fuliginosa, but that its infection mechanism makes it capable of infecting other cockroach species accidentally when exposed.
The remarkable features of L. appendiculatum we could reveal in these experiments are (1) it has a broad infectivity and (2) it has a strong association with a cockroach species that is currently spreading and inhabiting worldwide. P. fuliginosa inhabits all prefectures of Japan, although mainly central and western Japan (Nakano, 2013; The Japanese Society of Pestology, 2015; Tsuji, 1975). As the cockroaches examined in this experiment originally live in different habitats, there would be few opportunities to replace their original parasite in nature.
The field survey revealed host sharing of thelastomatid nematodes and suggested that host specificity is dictated largely not by their taxonomy but by the host ecology (Jex et al., 2006a(Jex et al., ,b, 2007 was isolated from endemic and invasive cockroaches (Sinnott et al., 2015). As there is no record of P. fuliginosa in these islands, these cockroaches can be also considered nematode carriers contributing to their spreading.
Adamson proposed that ancestors of oxyurid nematodes were diversifying in invertebrates long before the parasites of vertebrates, and that they probably arose in diplopods or blattarians (Adamson, 1989(Adamson, , 1994. Cospeciation of host and parasite was well conserved in primates and pinworm nematodes (Hugo, 1999). Host specificity influences the expansion of parasitic diseases when the host invades a new environment. Thus, understanding host range might contribute valuable information (Agosta, Janz, & Brooks, 2010). For example, the human pinworm Enterobius vermicularis only causes itching in the anal area of its native human host, but can be fatal in chimpanzees Pan troglodytes (Murata et al., 2002). In addition, parasitic nematodes greatly affect the composition of microbial communities in cockroaches. It is obvious that there are direct and indirect interactions between hosts and parasites . This nematode is also a good model system for understanding the mechanisms and evolution of mutualistic relationships between parasites and hosts.

ACK N OWLED G M ENTS
We

CO N FLI C T O F I NTE R E S T
None Declared. F I G U R E 4 Phylogenetic relation of five cockroach species examined in these experiments, constructed based on Inward, Beccaloni, & Eggleton (2007). These species belong to three families (Blattidae, Blattellidae, Blaberidae) and two suborders (Blattoidea, Blaberoidea). Blattodea was believed to separate with Mantodea from common ancestor between 250 and 200 mya. Two suborders, Blattoidea and Blaberoidea, were thought to separate between 200 and 150 mya (Misof et al. 2014). American cockroach P. americana another major sanitary pest and reported as a host of L. appendiculatum also belongs to the family Blattidae (Grimaldi & Engel, 2005)