Introduction

Cystic echinococcosis (CE) caused by the larval stages of Echinococcus granulosus sensu lato s.l is one of the most important, globally distributed zoonotic diseases. E. granulosus s.l is a cryptic species complex that includes Echinococcus granulosus sensu stricto (s.s) (genotypes G1–G3), Echinococcus felidis, Echinococcus equinus (G4), Echinococcus ortleppi (G5) and Echinococcus canadensis (G6/G7, G8, G10) (Nakao et al. 2007; Thompson 2008; Hüttner et al. 2008; Romig et al. 2015). E. granulosus s.l circulates among domestic and/or wild carnivores and herbivores, especially dogs and livestock animals (sheep, goats, cattle, camels, pigs) (Thompson 2008). Humans are accidental hosts infected through the ingestion of eggs excreted in dog faeces but do not contribute to the perpetuation of the life cycle.

High CE incidence has been reported from several countries neighbouring Turkey. For example, incidence rates of 5.71 (Jordanova et al. 2015), 4.5 (Abdulhameed et al. 2018) and 1.18–3 (Rokni 2008; Tavakoli et al. 2008) cases per 100,000 inhabitants were documented in recent years from Bulgaria, Basrah (Iraq) and Iran respectively. CE is endemic in Turkey with a high public health impact in rural areas where the local population is actively engaged in animal breeding and remains in close contact with ruminants and dogs (Altintas 2003). Although CE is a notifiable disease in Turkey, data on human infection is principally obtained from hospital records. The incidence rate of CE from 1990 to 2005 was estimated to be 0.8–2 per 100,000 (Yolasıgmaz et al. 2006; Altintas 2008).

Hydatidosis in ruminants is widespread and has been reported from many regions in Turkey (Altintas 2003; Umur 2003; Yildiz and Gurcan 2003; Yildiz and Tunçer 2005; Esatgil and Tuzer 2007; Utuk et al. 2008; Kose and Sevimli 2008; Simsek and Eroksuz 2009; Beyhan and Umur 2011). Meat production losses due to CE for 2008 were US$89.2 million and nationwide annual losses attributed to CE in sheep, cattle and goats were estimated to be US$ 54.1, 32.4 and 2.7 million respectively (Sariozkan and Yalcin 2009). Losses relating to offal condemnation estimated for a 6-month period during 2012 in Bursa Province were US$4015 and US$12,321 for sheep and cattle respectively (Yibar et al. 2015). The domestic dog is presumed to be the most important definitive host and is responsible for contamination of the environment with Echinococcus eggs. Variable prevalence rates of canine echinococcosis from different regions following necropsy (reviewed by Altintas 2003), coproantigen ELISA (Guzel et al. 2008) and parasitological faecal examination (Öter et al. 2011) have been recorded. Recent reports have used molecular methods to identify E. granulosus in dog faeces from Turkey (Acioz 2008; Utuk et al. 2008; Kuru et al. 2013; Öge et al. 2017).

During the last decade, several reports on the molecular confirmation of the causative agents of CE in ruminants have been published (Utuk et al. 2008; Šnàbel et al. 2009; Simsek and Eroksuz 2009; Beyhan and Umur 2011; Simsek et al. 2011a; Eroğlu et al. 2016). Using various DNA-based molecular approaches, these studies have shown that E. granulosus s.s (G1 and G3 genotypes) is the most prevalent species affecting sheep, cattle, goats, camels and pigs in Turkey. Similarly, several reports on the molecular identification of human CE causative agents in Turkey confirmed the predominance of E. granulosus s.s (G1 and G3 genotypes) (Utuk et al. 2008; Ergin et al. 2010; Simsek et al. 2011b; Eryildiz and Şakru 2012; Bakal et al. 2015; Eroğlu et al. 2016). In addition, E. canadensis G6 and G7 were reported from humans and sheep (Šnàbel et al. 2009; Simsek et al. 2011b).

The aim of this study was to investigate the genetic variation and population structure of E. granulosus s.s using metacestode isolates removed from surgically confirmed patients originating from several regions in Turkey. This work also involves the investigation of the occurrence of autochthonous transmission based on the comparison of our human CE nucleotide sequences with those reported from Turkish ruminants and deposited in GenBank.

Materials and methods

Ethics statement

The current study was approved by the Ethical Committee of the Faculty of Medicine, Hacettepe University, Ankara, Turkey (GO 14/293-37).

Sample collection, DNA extraction, PCR amplification and sequencing

Hydatid material (cyst fluid and/or germinal layer) retrieved between November 2014 and December 2016 from 46 percutaneously treated CE patients was included in this study. Demographic features such as sex, age and origin of patients were retrieved from patients’ clinical records.

Genomic DNA extracted from cyst material using the Qiagen DNeasy Blood and Tissue DNA Extraction Kit (Qiagen, Hilden, Germany) was used to amplify an 828-bp fragment within the cytochrome c oxidase subunit 1 (cox 1) mitochondrial gene using primers F/COI (5′ TTGAATTT-GCCACGTTTGAATGC 3′) and R/COI (5′ GAACCTAACGACATAACATAATGA 3′) (Nakao et al. 2000). Amplified PCR products were visualised under ultraviolet light (Syngene G: Box gel documentation system, Cambridge Biosciences) following electrophoresis on 1.5% (w/v) ethidium bromide-stained agarose gels (Cleaver Scientific Limited) and purified using QIAquick PCR Purification Kit (Qiagen) according to the manufacturer’s instructions. Bi-directional commercial sequencing was achieved using the PCR primers (Macrogen EZ-Sequence, Amsterdam, The Netherlands). FinchTV viewer (Geospiza, Seattle, WA, USA) was used to examine chromatograms and ascertain the quality of generated nucleotide sequences. A BLAST algorithm search (http://www.ncbi.nlm.nih.gov/BLAST/) was performed to compare our sequenced data with that present on the NCBI database. In addition, an alignment of a 366-bp fragment of our generated sequences with reference sequences published for E. granulosus G1-G3 genotypes (Bowles et al. 1992) was also conducted.

Nucleotide sequence data analysis

This was carried out using published methodology (Boufana et al. 2014). In brief, nucleotide DNA sequences were aligned using ClustalX2 software (Larkin et al. 2007) and transported into DnaSP 5 (Librado and Rozas 2009) to determine DNA polymorphisms and haplotype distributions. The best-fit nucleotide substitution model using the Akaike Information Criterion (AIC) (TrN + 1) was obtained through the application of Modeltest 3.7 (Posada and Crandall 1998) executed in Paup 4.0 (Swofford 1998). Population diversity indices such as haplotype numbers (hn), haplotype (hd) and nucleotide diversities (πd) were calculated in Arlequin version 3.5. (Excoffier and Lischer 2010) which was also used to test for population expansion or bottleneck using Tajima’s D (Tajima 1989) and Fu’s Fs (Fu 1997). Hapview (Salzburger et al. 2011) was used to generate parsimony haplotype networks and maximum likelihood trees were constructed by DNAML program in PHYLIP (Felsenstein 1989).

A total of 59 cox 1 mitochondrial nucleotide sequences of E. granulosus s.s (G1 genotype) from Turkish intermediate hosts (KU925352–KU925384, KU925386–KU925387, KU925389–KU925394, KU925396–KU925398, KU925400–KU925412) (sheep n = 22; cattle = 35) in addition to two shared sequences (KU925385, KU925395; sheep and cattle) (Kinkar et al. 2016) deposited in GenBank (http://www.ncbi.nlm.nih.gov) were also included in this study. These were aligned with the human CE sequences generated in this study, trimmed to equal lengths (827 bp) and used to create a new dataset (n = 105) for a human/animal parsimony haplotype analysis to investigate autochthonous transmission.

Results

Demographics of patients

Of the 46 patients included in this study, 56.5% (26/46) were from central Anatolia (Ankara, n = 16; Aksaray, n = 2; Çankırı, n = 2; Eskișehir, n = 1; Karaman, n = 1; Kırıkkale, n = 1; Sivas, n = 1; Yozgat, n = 2) (Fig. 1). The remaining patients originated from eastern Anatolia (Ağrı, n = 1; Bitlis, n = 2; Hakkari, n = 1; Kars, n = 1; Malatya, n = 1; Van, n = 1), the Black Sea Region (Corum, n = 3; Tokat, n = 1), Marmara (Istanbul, n = 1; Balikesir, n = 2), south-eastern Anatolia (Şanliurfa, n = 4), the Mediterranean region (Hatay, n = 1) and Aegean (Kütahya, n = 1). Our sample consisted of 29 female (63%) and 17 (36.9%) male patients with a mean and median age of 40.4 and 40.5 respectively (range 4–78 years).

Fig. 1
figure 1

Map of Turkey showing the origin and number of Echinococcus granulosus sensu stricto metacestodes used in this study (n = 46)

Full size image

Echinococcus species

Using DNA extracted from a total of 46 human-derived CE isolates, we successfully analysed an 827-bp fragment within the cox1 mitochondrial gene. The use of BLAST algorithm confirmed the causative agent of human cystic echinococcosis in patients included in this study to be E. granulosus s.s. Most isolates (39/46, 84.8%) belonged to E. granulosus G1 genotype whereas the remaining 7 isolates (15.2%) were identified as E. granulosus G3 genotype (Bowles et al. 1992).

DNA polymorphisms, diversity indices and parsimony network

Within the 827 bp analysed nucleotide sequences, we detected a total of 28 polymorphic sites of which 7 were parsimony-informative. No indels or gaps were observed. The overall haplotype and nucleotide diversities are shown in Table 1. The neutrality indices were both negative and significant (Tajima’s D − 2.1263; p < 0.01; Fu’s Fs − 26.80; p < 0.001) indicating negative selection or population expansion.

Table 1 Diversity and neutrality indices for Echinococcus granulosus sensu stricto isolates from Turkish patients using nucleotide sequence data of the cytochrome c oxidase subunit 1 mitochondrial gene
Full size table

The haplotype parsimony network consisted of 28 haplotypes arranged within three main clusters separated by one to three mutational steps (Fig. 2). The E. granulosus s.s dominant cluster with TUK04 as its main haplotype had 27/46 human CE isolates which were identified in patients from central Anatolia (n = 15), the Black Sea Region (n = 3), Marmara (n = 2), eastern Anatolia (n = 5), south-eastern Anatolia (n = 1) and the Mediterranean region (n = 1). The nucleotide sequences of haplotype TUK04 were 100% identical to those of the globally distributed ‘ancestral’ E. granulosus haplotype (Accession no. AB491414) first described by Nakao et al. (2010). TUK20 was the main haplotype in the second most dominant cluster (12/46) which gave 100% identity to E. granulosus G1 (Accession no. KU925431) and encompassed isolates from patients from central Anatolia (n = 7), the Black Sea Region (n = 1), eastern Anatolia (n = 1), south-eastern Anatolia (n = 2) and the Aegean Region (n = 1). It was also 100% identical to the second most common haplotype recently reported from Kurdistan, Iraq (Accession no. MF004305) (Hassan et al. 2017). The third cluster within this network of E. granulosus s.s (haplotypes TUK24–TUK28) (7/46) occurred in CE patients from central Anatolia (n = 4), Marmara (n = 1), south-eastern Anatolia (n = 1) and eastern Anatolia (n = 1) and gave a 99% homology to E. granulosus G3 genotype (Accession no. KY766903) (Fig. 2).

Fig. 2
figure 2

Haplotype network generated using mitochondrial nucleotide sequences of the cytochrome c oxidase subunit 1 (827 bp) of Echinococcus granulosus sensu stricto from human CE patients from Turkey

Full size image

We used nucleotide sequences of E. granulosus s.s (G1 genotype) derived from sheep (n = 24) and cattle (n = 37) hosts originating from Erzurum and Elazig provinces in eastern Turkey (Kinkar et al. 2017) to investigate autochthonous transmission through the generation of a network to include sequences from these animals and those generated from human CE patients in this study. This network consisted of 39 haplotypes with 18/24 (75%) (sheep) and 29/37 (cattle) (78.4%) isolates respectively sharing 7/39 haplotypes with the human CE isolates from this study (data not shown).

Discussion

In the current study, we used CE material derived from patients originating from seven Turkish regions from which human CE was previously reported. A retrospective study based on the examination of hospital, health directorship and Ministry of Health records from 2001 to 2005 identified 14,789 CE cases from central Anatolia (38.57%), Aegean Region (16.94%), Mediterranean region (16.09%), Marmara (13.13%), eastern Anatolia (6.8%), Black Sea Region (5.70%) and south-eastern Anatolia (2.75%) (Yazar et al. 2008). In addition, a study conducted to determine the etiological agent responsible for CE in patients originating from most of these regions showed that infection was due to E. granulosus G1 genotype (Ergin et al. 2010). Most CE cases described in this study were from female patients, which is consistent with results previously reported from Turkey (Hakverdi et al. 2008; Ozekinci et al. 2009).

CE infection in all 46 patients included in this study was confirmed to have been caused by E. granulosus s.s (G1 and G3 genotypes). Our results are consistent with the predominance of CE caused by G1 and G3 genotypes in Turkey (Schneider et al. 2008; Utuk et al. 2008; Šnàbel et al. 2009; Ergin et al. 2010; Simsek et al. 2011b; Eryildiz and Şakru 2012; Bakal et al. 2015; Eroğlu et al. 2016). Although E. granulosus G2 genotype was not detected in this study, it had been reported from two water buffaloes from the Black Sea Region of Turkey (Beyhan and Umur 2011). Interestingly, a re-appraisal of the G1–G3 genotype cluster using 112 metacestodes from sheep and cattle intermediate hosts from Turkey did not find evidence to support the continued presence of G2 genotype within E. granulosus s.s (Vural et al. 2008) and other authors have considered it as a microvariant of G3 genotype (Šnàbel et al. 2009). Further, a recent study questioned the validity of G2 genotype within the E. granulosus species complex whereas G1 and G3 were reported to be valid mitochondrial genotypes, but no difference was found between them at the nuclear level (Kinkar et al. 2017).

It is not surprising that E. granulosus s.s is the most the dominant causative species responsible for CE in humans and animals in Turkey. The Middle East has been described as the ‘ancestral seat’ of E. granulosus s.s (Eckert et al. 2001; Nakao et al. 2010). In addition, E. granulosus s.s is a cosmopolitan species that owes its ubiquitous existence to its ability to parasitize a wide range of animal hosts. Further, 88.44% of 1661 molecularly confirmed CE infections were caused by E. granulosus s.s (Alvarez Rojas et al. 2014). CE human infections in Turkey due to E. granulosus s.s. have been reported in adults as well as paediatric patients (Ergin et al. 2010; Simsek et al. 2011b; Bakal et al. 2015). Other human CE cases from Turkey include those caused by E. canadensis G6, which was reported in two surgically confirmed patients using histopathological tissues, retrieved from an Elazig Province university hospital (Simsek et al. 2011b). In addition, E. canadensis G7 was molecularly confirmed from Selçuk-Izmir Province (Šnàbel et al. 2009) and Edirne in the Thrace region (Eryildiz and Şakru 2012). Additional studies are required to further investigate the epidemiology of E. canadensis in Turkey and identify potential risk factors particularly in rural areas. In a recent study, living in endemic rural areas in association with free roaming dogs with access to infected offal were identified as significant risk factors for acquiring CE infection (Possenti et al. 2016).

Hydatidosis in animals from Turkey has been reported in sheep, cattle, goats, mouflon, water buffaloes and pigs (Umur 2003; Yildiz and Gurcan 2003; Yildiz and Tunçer 2005; Esatgil and Tuzer 2007; Kose and Sevimli 2008; Simsek and Eroksuz 2009; Beyhan and Umur 2011). The assessment carried out in this study using GenBank nucleotide sequence data from Turkey for sheep and cattle demonstrated the importance of autochthonous transmission concerning these two species where sheep, cattle and humans harboured the same haplotypes. Sheep are known as the most important hosts for the perpetuation and transmission of E. granulosus s.s worldwide (Deplazes et al. 2017) with high cyst fertility and viability reported from many countries. On the other hand, cattle generally harbour sterile cysts of E. granulosus s.s. Cyst fertility for sheep from Kirikkale for example was reported to be 81.53 and 76.47% for liver and lung cysts respectively (Yildiz and Gurcan 2003). The same study found that the mean number of viable liver protoscoleces was 12,400 and that for the lung cysts to be 5800 where the authors emphasised the important role sheep play in CE transmission in Turkey. Cattle cyst fertility, however, was much lower for example 6.6% in Kırıkkale (Yildiz and Tunçer 2005) and 5.42% in Afyonkarahisar district, western Turkey (Kose and Sevimli 2008) whereas as far as is known, no information on the viability of hydatid cysts derived from Turkish cattle has been published.

In summary, this is the first report on E. granulosus s.s haplotypes and population structure of metacestodes derived from human Turkish hosts. The only other known study on genetic diversity using human CE isolates from Turkey determined haplotypes within the E. granulosus s.s species through the identification of E. granulosus variants using RFLP and SSCP profiles (Eryildiz and Şakru 2012).

All three haplotype clusters described in this study occurred in patients from central and eastern Anatolia and Marmara regions. The E. granulosus G3 haplotype cluster (TUK24–TUK28) was observed in CE patients from central, eastern and south-eastern Anatolia and Marmara only, although this may be related to the small sample size for some of these regions. Further studies are required to investigate the biological significance if any, of E. granulosus s.s haplotypes and the genetic variability of CE from human patients using longer nucleotide sequences and a larger sample set.