INTRODUCTION
The liver fluke, Fasciola hepatica is a major parasite of livestock in temperate regions of the world. There has been a dramatic rise in the incidence of disease in recent years, which has been attributed to climate change (Mitchell, Reference Mitchell2002). Furthermore, fascioliasis is now recognized as a major zoonotic infection of increasing importance (Mas-Coma et al. Reference Mas-Coma, Bargues and Valero2005; WHO, 2007). Current control methods effectively centre round the use of anthelmintic drugs, of which the most commonly employed is the benzimidazole derivative, triclabendazole (TCBZ, ‘Fasinex’) (Fairweather, Reference Fairweather2009). It is favoured due to its remarkably high efficacy against both adult and juvenile stages of F. hepatica (Boray et al. Reference Boray, Crowfoot, Strong, Allison, Schellenbaum, Von Orelli and Sarasin1983). Unfortunately, due to the over-dependence on this drug and its continued use as a monotherapy, resistance is steadily on the increase. Resistance to triclabendazole was first noted on Australian farms in 1995 (Overend and Bowen, Reference Overend and Bowen1995). Since then, resistance has been detected throughout much of Europe (Fairweather, Reference Fairweather2005, Reference Fairweather2009). Resistance in the UK and Ireland is believed to be spreading and is probably significantly underestimated (Wolstenholme et al. Reference Wolstenholme, Fairweather, Prichard, von Samson-Himmelstjerna and Sangster2004). This means that there is an urgent need to find novel fluckicidal compounds.
One group of drugs that has attracted considerable attention in the last few years is the artemisinins. The artemisinins are a family of natural or semi-synthetic compounds derived from the plant, Artemisia annua which is grown in eastern Asia (O'Neill, Reference O'Neill2004) and have established uses as anti-malarial and anti-schistosomal drugs (Borstnik et al. Reference Borstnik, Paik, Shapiro and Posner2002; Woodrow et al. Reference Woodrow, Haynes and Krishna2005; Xiao, Reference Xiao2005; Utzinger et al. Reference Utzinger, Xiao, Tanner and Keiser2007). The main semi-synthetic derivatives of artemisinin are artemether, artesunate and the active metabolite, dihydroartemisinin (Woodrow et al. Reference Woodrow, Haynes and Krishna2005). Artemether and artesunate have been shown to be effective fasciolicides in the treatment of rodent infections (Keiser et al. Reference Keiser, Xiao, Tanner and Utzinger2006a). Due to the unfavourable pharmacokinetic properties of the artemisinins, many synthetic anti-malarial peroxides have been prepared, including the synthetic 1,2,4-trioxolanes, such as OZ277 and OZ78 (Vennerstrom et al. Reference Vennerstrom, Arbe-Barnes, Brun, Charman, Chiu, Chollet, Dong, Dorn, Hunziker, Matile, McIntosh, Padmanilayam, Tomas, Scheurer, Scorneaux, Tang, Urwyler, Wittlin and Charman2004). Synthetic peroxides retain the peroxide pharmacophore essential for anti-parasitic activity, yet are more simple molecules, easier to synthesize and have improved pharmacokinetic properties (Jefford, Reference Jefford2007). Worm burden reductions of 100% have been achieved with both juvenile and adult F. hepatica following a single oral dose of 100 mg/kg OZ78 (Keiser et al. Reference Keiser, Utzinger, Tanner, Dong and Vennerstrom2006b). Significantly, high efficacy has also been recorded for OZ78 and artemether in treatment of triclabendazole-resistant fluke infections in rats (Keiser et al. Reference Keiser, Utzinger, Vennerstrom, Dong, Brennan and Fairweather2007a). These results suggest that OZ78 has a high potential as a flukicidal drug and warrants further evaluation. Relatively little work has been performed so far in relation to the morphological effects of artemisinin-type compounds on F. hepatica and it has been confined to surface changes (Keiser et al. Reference Keiser, Xiao, Tanner and Utzinger2006a, Reference Keiser, Utzinger, Tanner, Dong and Vennerstromb; Keiser and Morson, Reference Keiser and Morson2008a, Reference Keiser and Morsonb).
The present study was carried out to determine the changes induced by treatment in vitro with the synthetic peroxide OZ78 on the surface morphology of adult F. hepatica, as visualized by scanning electron microscopy (SEM), and on the internal ultrastructure by means of transmission electron microscopy (TEM). The aim was to identify possible tissue targets of drug action.
MATERIALS AND METHODS
The experiment was carried out in 2 parts: the rat infections, in vitro incubation in OZ78 and initial fixation of fluke material was performed at the Swiss Tropical Institute, Basel and then the fluke material was sent to The Queen's University of Belfast for further processing and examination by scanning and transmission electron microscopy.
Female Wistar rats (purchased from RCC, Itingen, Switzerland) were experimentally infected with metacercarial cysts of the Cullompton isolate, shown in previous experiments to be susceptible to TCBZ in vivo and to its sulphoxide metabolite in vitro (Robinson et al. Reference Robinson, Trudgett, Hoey and Fairweather2002; McCoy et al. Reference McCoy, Fairweather, Brennan, Kenny, Ellison and Forbes2005; Halferty et al. Reference Halferty, Brennan, Hanna, Edgar, Meaney, McConville, Trudgett, Hoey and Fairweather2008, Reference Halferty, Brennan, Trudgett, Hoey and Fairweather2009; McConville et al. Reference McConville, Brennan, McCoy, Castillo, Hernández-Campos, Ibarra and Fairweather2006, Reference McConville, Brennan, Flanagan, Edgar, Hanna, McCoy, Gordon, Castillo, Hernández-Campos and Fairweather2009a; Meaney et al. Reference Meaney, Allister, McKinstry, McLaughlin, Brennan, Forbes and Fairweather2006, Reference Meaney, Allister, McKinstry, McLaughlin, Brennan, Forbes and Fairweather2007; Devine et al. Reference Devine, Brennan, Lanusse, Alvarez, Trudgett, Hoey and Fairweather2009; Toner et al. Reference Toner, McConvery, Brennan, Meaney and Fairweather2009). All animals were housed in environmentally-controlled conditions in groups of 4 and afforded 1 week to acclimatize. Aforementioned conditions were: temperature, 25°C; humidity, 70%; 12 h light/dark cycle. Free access to water and rodent diet was also maintained. Twelve rats were infected intragastrically with 25 cysts each and, after a period of 9–10 weeks, adult flukes were recovered from the bile ducts at necropsy. Recovered flukes were then incubated in 6-well plates (Costar, Corning, USA), placing 2 fluke specimens per well. Each well contained 5 ml of NCTC 135 culture medium (Invitrogen, Carlsbad, USA) containing streptomycin and penicillin at concentrations of 50 μg/ml and 50 IU/ml, respectively. A stock solution of OZ78 (synthesized at the College of Pharmacy, University of Nebraska Medical Centre, USA) was prepared in 60% (v/v) DMSO and flukes were incubated with the drug for a period of 48 h at a concentration of 100 μg/ml. The concentration used was based on the results of previous studies with OZ78 involving F. hepatica, Echinostoma caproni and Clonorchis sinensis (Keiser et al. Reference Keiser, Utzinger, Tanner, Dong and Vennerstrom2006b, Reference Keiser, Utzinger, Vennerstrom, Dong, Brennan and Fairweather2007a, Reference Keiser, Xiao, Dong, Utzinger and Vennerstromb). Fluke motility was observed at regular intervals by way of a dissecting microscope. Specimens were judged dead if no movement was observed for 2 min. The experiment was terminated at 48 h, when the flukes were considered dead. Controls consisted of incubating flukes for 48 h at 37°C in NCTC 135 culture medium in the presence of streptomycin and penicillin. The concentration of DMSO in the control medium (0·05%) was matched to that of the drug medium. The control flukes were still active and appeared normal after 48 h incubation. Controls at 0 h were also prepared. Following treatment, the flukes were fixed with 2·5% (v/v) glutaraldehyde in a PBS buffer (pH 7·3) for 24 h at room temperature. After rinsing and storing in PBS buffer, the specimens were sent to The Queen's University of Belfast for further processing and examination.
For scanning electron microscopy (SEM), 8 flukes were washed in 0·1 m sodium cacodylate buffer (pH 7·4) containing 3% (w/v) sucrose. Following a post-fixation in 1% osmium tetroxide for 1 h, the flukes were given 3 brief washes in 70% (v/v) ethanol and dehydrated in an ascending series of ethanol. After this, they were dried in hexamethyldisilazane, mounted on aluminium stubs and sputter-coated with gold-palladium. The specimens were viewed in an FEI Quanta 200 scanning electron microscope, operating at an accelerating voltage of 10 keV.
For transmission electron microscopy (TEM), 5 flukes were washed in 0·1 m sodium cacodylate buffer (pH 7·4) containing 3% (w/v) sucrose. The apical cones and tail regions of the flukes were discarded and the remaining midbody region was divided transversely into 2 parts; each part was further subdivided transversely into 6 pieces. The pieces were washed in 0·1 m sodium cacodylate buffer (pH 7·4) containing 3% (w/v) sucrose. Following post-fixation in 1% osmium tetroxide for 1 h, the tissue was washed in fresh buffer, dehydrated in an ascending series of ethanol and infiltrated and embedded in Agar 100 resin. Ultrathin sections, 60–70 nm in thickness, were cut on a Reichert Ultracut E ultramicrotome, mounted on bare 200-mesh copper grids, double-stained with alcoholic uranyl acetate (5 min) and aqueous lead citrate (3 min) and viewed in a FEI CM 100 transmission electron microscope operating at an accelerating voltage of 100 keV.
RESULTS
Scanning electron microscopy
The surface architecture of the control flukes appeared normal (Fig. 1A). It matched the images presented by Bennett (Reference Bennett1975), Fairweather et al. (Reference Fairweather, Threadgold, Hanna and Dalton1999) and McConville et al. (Reference McConville, Brennan, Flanagan, Hanna, Edgar, Castillo, Hernández-Campos and Fairweather2009b).
Scanning electron micrographs (SEMs) of adult Fasciola hepatica incubated in vitro for 48 h. (A) Control, incubated in NCTC culture medium; (B–D) flukes treated with OZ78 at a concentration of 100 μg/ml. (A) Control fluke. Low-power image, showing the oral sucker (OS), ventral sucker (VS), midbody region (MB) and tail region (TR). Inset shows spines (S) from the oral cone region, which bear distinct spinelets (arrow). (B) Low-power image, showing the oral sucker (OS), ventral sucker (VS) midbody region (MB) and tail region (TR). Inset shows spines (S) from the oral cone region, which bear distinct spinelets (arrow). (C) Ventral midbody region. The inter-spinal tegument appears swollen (broken arrows). The arrangement of the spines (S) is more irregular than normal. Inset shows that the spines (S) are slightly sunken due to the swelling of the inter-spinal tegument (arrows). The tegumental surface has a roughened appearance (*). (D) Dorsal midbody region. Swelling of the inter-spinal surface is evident (broken arrows). The arrangement of the spines (S) appears irregular. Inset is a higher-power image showing that the spinelets (broken arrow) are less well-defined. The inter-spinal surface, although swollen (arrows), appears smooth. S, spine.
Following treatment with OZ78, little change was observed anteriorly (Fig. 1B). Spines were grouped in a regular, tight-knit arrangement and clearly protruded from the inter-spinal surface. The spinelets at the tips of the spines (Fig. 1B) were distinct and there was no spine disruption or swelling of the tegument.
The ventral midbody region showed slight swelling of the interspinal tegument (Fig. 1C). This swelling lent an uneven appearance to the surface of the fluke. At higher magnification (Fig. 1C inset), the spines were seen to be slightly sunken, indicating some swelling of the tegument surrounding them. The interspinal surface appeared rough and was covered with small blebs. The dorsal surface of the midbody region exhibited general swelling of the interspinal tegument (Fig. 1D). The swelling again lent an uneven appearance to the surface of the fluke. Spines in this region appeared sunken, due to the swelling of the tegument around them (Fig. 1D inset) and the spinelets were less clearly defined (Fig. 1D inset).
The ventral surface of the tail region (Fig. 2A) showed a relatively normal morphology and the tegument was not swollen. At higher magnification (Fig. 2A inset), the interspinal surface was seen to be roughened. Spines were prominent and identical in shape and structure to those of control specimens. The dorsal tail (Fig. 2B) showed a regular pattern of spines protruding from a roughened tegumental surface. There was no evidence of swelling of the tegument.
Scanning electron micrographs (SEMs) of the posterior region of adult Fasciola hepatica following 48 h treatment in vitro with OZ78 (100 μg/ml). (A) Ventral tail region. The spines (S) project from the surface of the tegument, which is not swollen. Inset shows the roughened appearance of the inter-spinal surface, due to the covering of small blebs (*). S, spine. (B) Dorsal tail region. The spines (S) are regularly-arranged and protrude from the surface. Inset shows that the tegumental surface appears roughened (*) at a higher magnification. S, spine. (C) The main image shows the presence of a bulbous swelling (*) on the ventral surface in the posterior midbody region. EP, excretory pore. Inset shows the swelling (*) at higher magnification. (D) Severe disruption to the tail end of the fluke, which appears to be covered with large blebs, many of which have burst (oval outline). Inset shows the large and burst blebs (*) at high magnification. One of the burst blebs contains a spine (S).
Figures 2C and D demonstrate unusual features seen in 2 of the specimens. A bulbous swelling was observed in the dorsal anterior midbody region of 1 specimen, the identity of which is not clear (Fig. 2C and inset). Apart from this swelling, the surface of the fluke appeared normal. More dramatic disruption was evident in another specimen, in which the posterior end of the fluke was covered with large blebs, many of which had burst (Fig. 2D). One of the burst blebs was seen to contain a spine (Fig. 2D inset).
Transmission electron microscopy
The ultrastructure of the tegumental syncytium, tegumental cells and reproductive structures in the control specimens remained normal. For images, the reader is referred to the papers by Threadgold (Reference Threadgold1963, Reference Threadgold1967) for tegument and parenchyma; Irwin and Threadgold (Reference Irwin and Threadgold1970) and Colhoun et al. (Reference Colhoun, Fairweather and Brennan1998) for vitellaria; and Stitt and Fairweather (Reference Stitt and Fairweather1990) for testis.
Following drug treatment, severe alterations were evident throughout the tegumental syncytium, subtegumental region, gut and reproductive tissues, in sections taken from the midbody region and as illustrated in Figs 3 and 4. In the syncytium, the basal infolds were severely swollen and occupied the entire basal region of the syncytium (Fig. 3A and C). Blebbing of the apical plasma membrane was observed (Fig. 3B) and autophagic vacuoles were present in the syncytium (Fig. 3B). The numbers of T1 and T2 secretory bodies in the syncytium were considerably less than normal. The mitochondria were swollen and electron-lucent (Fig. 3C). The crystalline structure of the spines was seen to be disrupted: it had a rippled and striated appearance (Fig. 3B).
Transmission electron micrographs (TEMs) of the tegumental syncytium and subtegumental region of adult F. hepatica following 48 h treatment in vitro with OZ78 (100 μg/ml). (A) The basal infolds (BI) are severely swollen. Beneath the basal lamina (BL), the circular (CM) and longitudinal (LM) muscle layers can be seen. APM, apical plasma membrane. (B) High-power image of the apical region of the tegumental syncytium. Small blebs (B) can be seen on the surface. An autophagic vacuole (AV) is present. APM, apical plasma membrane. Inset shows disruption (broken arrows) to the crystalline structure of a spine (S). (C) High-power image of the swollen basal infolds (BI) above the basal lamina (BL). The mitochondria (m) are swollen and electron-lucent. (D) Circular muscle (CM) layer showing the muscle bundles surrounded by large spaces (*). Inset shows the large spaces (*) between the longitudinal muscle (LM) bundles. (E) Low-power image of the subtegumental region. The tegumental cell bodies (TCB) contain large numbers of lipid droplets (L) and are separated from each other by large spaces (*). (F) High-power image of a tegumental cell body (TCB). The irregularly-shaped nucleus (N) can be seen located centrally within the cell. Inset i shows the swollen and electron-lucent appearance of the mitochondria (m). The cisternae of the granular endoplasmic reticulum (ger) are also distended. Inset ii shows the severe swelling of the cisternae of the granular endoplasmic reticulum (ger) at higher magnification. An autophagic (AV) vacuole is present in the cell.
Transmission electron micrographs (TEMs) of adult Fasciola hepatica following 48 h treatment in vitro with OZ78 (100 μg/ml). (A) Gastrodermal cells, showing a large intercellular space (ICS) between two adjacent cells. Within the cells, are vacuoles (Vac), autophagic vacuoles (AV), lipid droplets (L) and a small number of secretory bodies (SB). The cisternae of granular endoplasmic reticulum (ger) are swollen. The mitochondria (m) are swollen and electron-lucent. La, gut lamellae. (B) Gastrodermal cell, containing empty vacuoles (Vac), autophagic vacuoles (AV) and lipid droplets (L). Few secretory bodies (SB) are present. The mitochondria (m) are swollen and some are electron-lucent and appear to be breaking down. The cisternae of granular endoplasmic reticulum (ger) are swollen and vesiculated. ICS, intercellular space. (C) Vitelline follicle. The follicle is heavily-disrupted, with no definable early developmental stages present. The shell protein globules in the shell globule clusters (SGC) show signs of coalescence. Lipid droplets (L) are present in the cells. Inset is a high-power image showing the abnormal shell globule clusters (SGC). (D) Vitelline follicle (oval outline) is severely disrupted and is surrounded by widespread degeneration of the parenchymal tissue. Inset is a high-power image showing a follicle in the final stages of breakdown. Almost no discernible structures remain, and autophagic vacuoles (AV) are present throughout. (E) Testis follicle, showing disruption of the spermatogenic cells. The spermatogonial cells (SG) appear rounded and abnormal and the primary spermatocytes (SC) show division of the nuclei (*) without cytoplasmic division of the cells themselves. General degeneration of the follicular tissue is evident throughout the follicle and autophagic vacuoles (AV) are present. (F) Testis follicle, which is almost completely devoid of spermatogenic cells. There is widespread disruption throughout. SG, spermatagonia; AV, autophagic vacuole.
Beneath the basal lamina, large spaces separated the muscle blocks belonging to the circular and longitudinal muscle layers (Fig. 3D). Large spaces surrounded the tegumental cell bodies, too (Fig. 3E and F). Branches of the excretory ducts contained many lipid droplets (Fig. 4A). Within some of the tegumental cell bodies themselves, there were large numbers of lipid droplets as well (Fig. 3E). In all of the cells, the cytoplasm was filled with swollen and electron-lucent mitochondria, with distinct cristae and some of the mitochondria appeared to be in the process of breaking down (Fig. 3F). The cisternae of granular endoplasmic reticulum (ger) were swollen as well (Fig. 3F). Autophagic vacuoles were present in the cells (Fig. 3F). Type 1 and type 2 tegumental secretory bodies were generally absent from the cells and there were no discernable Golgi complexes present.
There was severe disruption to the gastrodermal cells. Large spaces were present between adjacent cells, the cells only being held together by the tight junction at their apices (Fig. 4A). Within the cells themselves, there were large empty vacuoles, together with autophagic vacuoles and lipid droplets, but few secretory bodies (Fig. 4A and B). The cisternae of granular endoplasmic reticulum were swollen and vesiculated (Fig. 4B).
The condition of the vitelline follicles is shown in Fig. 4C and D. Within the follicles it was difficult to distinguish any stem and early immature cells and the follicle was loosely-packed. The predominant cell type was the mature cell, although it displayed a number of abnormalities. That is, the shell globule clusters were either loosely-packed or the shell protein globules appeared to have coalesced (Fig. 3C). The cisternae of the ger were swollen and a number of lipid droplets were present (Fig. 3C). In some cells, the cytoplasm appeared to be breaking down (Fig. 3C). In more extreme cases, the vitelline follicles were greatly reduced in size and contained no recognizable cell types, as the cells had broken down and were full of autophagic vacuoles (Fig. 3D). In these situations, the surrounding tissue was very disorganized and empty looking (Fig. 3D).
The testis follicles are pictured in Fig. 4E and F. Only spermatogonia and primary spermatocytes were present. The spermatogonia were rounded, the space between the inner and outer nuclear membranes was swollen and vacuoles were present in the cytoplasm (Fig. 4E). The spermatocytes appeared to have undergone nuclear division without undergoing cytoplasmic division, giving rise to multinucleate cells (Fig. 4E). In the most extreme cases, the follicle was almost totally devoid of spermatogenic cells and the tissue that remained was very disorganized and autophagic (Fig. 4F).
DISCUSSION
Following treatment in vitro of F. hepatica with the synthetic peroxide OZ78 for 48 h at a concentration of 100 μg/ml, there was a marked difference between what disruption was evident externally and what disruption occurred internally. That is, changes to the surface morphology were quite limited. In stark contrast to the SEM observations, however, there was severe internal disruption of the tegumental, subtegumental and reproductive tissues. The basal infolds of the tegumental syncytium were swollen, flooding the basal region of the syncytium. Flooding was also evident internally, with large spaces separating the subtegumental cells. The gastrodermal cells were severely affected, with evidence of metabolic disruption and breakdown. In both the vitelline and testis follicles, the normal development of the cells was disrupted and, in more extreme cases, the vitelline follicles appeared to be in the process of breaking down. These findings will now be discussed in detail and compared with the findings of previous, related studies.
Previous morphological studies in vitro and in vivo with OZ78 and other artemisinins have concentrated solely on the tegument of F. hepatica using SEM (Keiser et al. Reference Keiser, Xiao, Tanner and Utzinger2006a, Reference Keiser, Utzinger, Tanner, Dong and Vennerstromb, Reference Keiser, Xiao, Dong, Utzinger and Vennerstrom2007b; Keiser and Morson, Reference Keiser and Morson2008a, Reference Keiser and Morsonb; O'Neill et al. Reference O'Neill, Johnston, Halferty, Brennan, Keiser and Fairweather2009). While the changes seen can vary between flukes, they are generally relatively mild, with swelling, localized areas of blebbing and occasional sloughing. The results of the present study were consistent with this pattern, with swelling and limited blebbing over part of the body surface. There was not the severe and extensive disruption seen with other fasciolicides (e.g. Fairweather et al. Reference Fairweather, Anderson and Baldwin1987; Skuce and Fairweather, Reference Skuce and Fairweather1990; Stitt and Fairweather, Reference Stitt and Fairweather1993; McKinstry et al. Reference McKinstry, Fairweather, Brennan and Forbes2003; Meaney et al. Reference Meaney, Fairweather, Brennan, McDowell and Forbes2003; McConville et al. Reference McConville, Brennan, McCoy, Castillo, Hernández-Campos, Ibarra and Fairweather2006; Halferty et al. Reference Halferty, Brennan, Trudgett, Hoey and Fairweather2009). More severe changes were seen in 2 of the flukes. The large swelling observed in one of them is similar to that described previously (Keiser et al. Reference Keiser, Utzinger, Tanner, Dong and Vennerstrom2006b), but it is not known what internal structure it relates to. Such individual variation in response to artemisinin treatment seems to be typical of F. hepatica, as it has been described in other studies (Keiser et al. Reference Keiser, Utzinger, Tanner, Dong and Vennerstrom2006b; Keiser and Morson, Reference Keiser and Morson2008a, Reference Keiser and Morsonb). Surface changes with artemisinins are greater when haemin is included in the in vitro culture medium (Keiser and Morson, Reference Keiser and Morson2008a, Reference Keiser and Morsonb) and following in vivo treatment, which may be related to proposals concerning the activation of the drug following its ingestion by the fluke, as has been suggested for schistosomes (Xiao et al. Reference Xiao, Wu, Tanner, Wu, Utzinger, Mei, Scorneaux, Chollet and Zhai2003). That is, interaction with haemin or another iron-containing compounds derived from the haemoglobin in the blood that the worms feed upon may lead to activation of the artemisinin compound. Activation leads to cleavage of the peroxide bridge in the artemisinin molecule (which is the essential moiety required for activity) and generation of free radicals which are damaging to the parasite (Xiao et al. Reference Xiao, Wu, Tanner, Wu, Utzinger, Mei, Scorneaux, Chollet and Zhai2003). In relation to the role of haemin in drug action, an interesting observation has been made that, although disruption is greater in the presence of haemin, it does not accelerate the ‘death’ of the fluke (Keiser and Morson, Reference Keiser and Morson2008b). A similar observation has been made for artemether and artesunate (Keiser and Morson, Reference Keiser and Morson2008a). This suggests that cleavage of the peroxide bridge is not an absolute pre-requisite for drug action and that an iron-independent mechanism may be involved, views expressed by Xiao et al. (Reference Xiao, Keiser, Chollet, Utzinger, Dong, Endriss, Vennerstrom and Tanner2007) and Keiser and Morson (Reference Keiser and Morson2008a). Indeed, it has been suggested that the trioxolanes have different actions from the semi-synthetic artemisinins (Uhlemann et al. Reference Uhlemann, Wittlin, Matile, Bustamante and Krishna2007).
In a separate in vivo study on F. hepatica involving artemether, the gut was consistently more severely affected than the tegument, lending support to the notion of oral uptake of the artemisinin compounds (O'Neill et al. Reference O'Neill, Johnston, Halferty, Brennan, Keiser and Fairweather2009). In the present study, the gut was severely disrupted and this would make the condition of the fluke worse. There was little secretory activity and the ger was swollen and vesiculated. Vesiculation of the ger has been linked to a state of starvation or stress in the fluke (Robinson and Threadgold, Reference Robinson and Threadgold1975; Skuce and Fairweather, Reference Skuce and Fairweather1990; Meaney et al. Reference Meaney, Fairweather, Brennan and Forbes2004, Reference Meaney, Haughey, Brennan and Fairweather2005; McKinstry et al. Reference McKinstry, Brennan, Halferty, Forbes and Fairweather2007). Moreover, there were signs that the cells were beginning to break down, as evidenced by the presence of autophagic vacuoles and degenerating mitochondria.
In the tegumental system, the main changes seen were a severe swelling of the basal infolds, surface blebbing, swelling of the mitochondria and cisternae of ger, the presence of autophagic vacuoles and lipid droplets and a lack of secretory bodies in the tegumental cells. Similar changes to the tegument, together with the widespread internal oedema, have been observed in schistosomes following treatment with artemether (Xiao et al. Reference Xiao, Binggui, Utzinger, Chollet and Tanner2002a, Reference Xiao, Binggui, Utzinger, Chollet and Tannerb). In the present study, the blebs present on the apical plasma membrane were also seen with SEM and the swelling of the basal infolds (together with the internal flooding) will explain the tegumental swelling visible externally. Greater swelling of the basal infolds would lead to the sloughing of the tegument, as has been observed following treatment with other fasciolicides (e.g. Fairweather et al. Reference Fairweather, Anderson and Threadgold1986; Skuce and Fairweather, Reference Skuce and Fairweather1990; Stitt and Fairweather, Reference Stitt and Fairweather1993; Anderson and Fairweather, Reference Anderson and Fairweather1995; Meaney et al. Reference Meaney, Fairweather, Brennan and Forbes2004). Clearly, there is disruption of the tegumental osmoregulatory system, which spreads internally to affect the subtegumental tissues. This may be due to inhibition of the energy-dependent ion pumps on the tegumental membranes and is evidenced by the swollen mitochondria and cisternae of ger and the presence of lipid droplets, together with the swelling of the basal infolds. Similar changes have been seen following treatment with fasciolicides that are believed to target energy metabolism, including clorsulon, closantel and nitroxynil (Skuce and Fairweather, Reference Skuce and Fairweather1990; Meaney et al. Reference Meaney, Fairweather, Brennan and Forbes2004, Reference Meaney, Haughey, Brennan and Fairweather2005, Reference Meaney, Allister, McKinstry, McLaughlin, Brennan, Forbes and Fairweather2007; McKinstry et al. Reference McKinstry, Brennan, Halferty, Forbes and Fairweather2007, Reference McKinstry, Halferty, Brennan and Fairweather2009). Alternatively, the flooding may stem from damage to the surface membrane as a result of impairment of production of secretory bodies in the tegumental cells. Maintenance of the apical membrane is dependent on the continual production and release of secretory bodies (Fairweather et al. Reference Fairweather, Threadgold, Hanna and Dalton1999). The lack of secretory bodies in the tegumental cells and reduced numbers in the syncytium suggest that they are no longer being produced, with the consequence that the surface membrane will be affected, leading to the changes observed. Similar morphological changes have been seen with fasciolicides that are not believed to affect energy metabolism, such as TCBZ and diamphenethide (Fairweather et al. Reference Fairweather, Anderson and Threadgold1986; Skuce et al. Reference Skuce, Anderson and Fairweather1987; Stitt and Fairweather, Reference Stitt and Fairweather1994; Halferty et al. Reference Halferty, Brennan, Trudgett, Hoey and Fairweather2009). These two drugs have been shown to inhibit protein synthesis in the fluke (Anderson et al. Reference Anderson, Fairweather, Bamford and Montgomery1993; Stitt et al. Reference Stitt, Fairweather and Mackender1995), an action also attributed to artemisinin in the malarial parasite (Gu et al. Reference Gu, Warhurst and Peters1983).
The testes and vitelline follicles were severely disrupted by treatment with OZ78. In the testes, only spermatogonia and primary spermatocytes were present and in low numbers. The fluke isolate used for this experiment was the Cullompton isolate: in this isolate, spermatogenesis is known not to proceed beyond the primary spermatocyte stage, with the consequence that this isolate is aspermic and parthenogenetic (Fletcher et al. Reference Fletcher, Hoey, Orr, Trudgett, Fairweather and Robinson2004; Hanna et al. Reference Hanna, Edgar, Moffett, McConnell, Fairweather, Brennan, Trudgett, Hoey, Cromie, Taylor and Daniel2008). Following OZ78 treatment, the spermatogonia were abnormal and the primary spermatocytes were multinucleate, as a result of nuclear, but not cytoplasmic, division. Eosinophilic and apoptotic multinucleate bodies equivalent to aberrant secondary spermatocytes have been described by Hanna et al. (Reference Hanna, Edgar, Moffett, McConnell, Fairweather, Brennan, Trudgett, Hoey, Cromie, Taylor and Daniel2008), but were not seen in the present study. So, even within the limited amount of sperm development that this isolate undergoes, there was disruption of the two early stages and the follicle became very empty.
In those vitelline follicles that could still be recognized, there appeared to be a preponderance of mature cells, suggesting that the normal developmental sequence from stem to mature cell had been disrupted. A similar shift in cell population within the follicle was observed after treatment with TCBZ and diamphenethide (Fairweather et al. Reference Fairweather, Anderson and Threadgold1988; Stitt and Fairweather, Reference Stitt and Fairweather1996). Within the mature cells, the shell globule clusters were loosely-packed; in some clusters the shell protein globules had coalesced; the cisternae of ger were swollen; lipid droplets were present and the cytoplasm appeared to be breaking down. Again, similar changes have been observed in the vitelline cells of Schistosoma mansoni following treatment with artemether (Xiao et al. Reference Xiao, Binggui, Utzinger, Chollet and Tanner2002b). The presence of lipid and the disruption of the ger could be indicative of metabolic disruption by OZ78, whilst the loosely-packed clusters could be the result of a decrease in shell protein synthesis, as occurred with TCBZ and diamphenethide (Fairweather et al. Reference Fairweather, Anderson and Threadgold1988; Stitt and Fairweather, Reference Stitt and Fairweather1996). Premature fusion of the shell protein globules has been described in drug-treated F. hepatica before, but may be related more to pH changes than to alterations in calcium ion levels, as it did not occur after treatment with the calcium ionophore, lasalocid (Colhoun et al. Reference Colhoun, Fairweather and Brennan1998). A potential target of artemisinin action in malarial parasites is the sarco(endo)plasmic reticulum Ca++-ATPase (SERCA), a membrane-bound enzyme that regulates calcium levels within the parasite (Eckstein-Ludwig et al. Reference Eckstein-Ludwig, Webb, van Goethem, East, Lee, Kimura, O'Neill, Bray, Ward and Krishna2003; Jung et al. Reference Jung, Kim, Nam and No2005; Uhlemann et al. Reference Uhlemann, Wittlin, Matile, Bustamante and Krishna2007): swelling of the cisternae of ger in the vitelline (and tegumental) cells may be linked to this action, but comparable changes to the sarcoplasmic reticulum in the muscle cells was not seen. The disruption of the ger would severely affect the synthetic activity of the cells.
In extreme instances, the vitelline follicles were greatly reduced in size, contained very few cells (which were in the process of breaking down) and were barely recognizable. There is an old maxim, first advanced by Dawes (Reference Dawes1968) that, in order to survive, flukes will shut down non-essential organs (such as reproductive organs) when under stress – such as that induced by drug treatment. It is possible that the changes to the testes and vitelline follicles simply reflect that maxim. The vitelline follicles occupy a large proportion of the body of the fluke and the vitelline cells produce the shell protein that underpins the production of large numbers of eggs (~25 000 eggs per fluke per day: Happich and Boray, Reference Happich and Boray1969); the cells have a high rate of turnover and are very sensitive to drug action (Fairweather et al. Reference Fairweather, Anderson and Threadgold1988; Skuce and Fairweather, Reference Skuce and Fairweather1990; Stitt and Fairweather, Reference Stitt and Fairweather1996; Colhoun et al. Reference Colhoun, Fairweather and Brennan1998). However, the changes seen were considerable and may be indicative of a more general action of OZ78. A severe degeneration of the vitelline follicles and atrophy (along with almost complete disappearance) of the testes and ovaries have been described following treatment of Schistosoma mansoni with artemether (Xiao and Catto, Reference Xiao and Catto1989; Xiao et al. Reference Xiao, Guo, Chollet, Wu, Tanner and Utzinger2004). Remarkably, the changes were reversible. The changes seen in the current study could be attributed simply to prolonged incubation in vitro, but this is unlikely as they were not observed in the control flukes.
So far, the morphological changes seen following treatment with OZ78 have been discussed largely in relation to those observed with existing fasciolicides. As indicated above, it is believed that artemisinins may act – at least in part – via the production of free radicals and it may be possible to explain some of the changes on this basis. For example, artemisinins and free radicals are known to induce the peroxidation of membrane lipids and oxidation of membrane proteins and membrane-bound enzymes, such as ion pumps (Rohn et al. Reference Rohn, Hinds and Vincenzi1996; Berman and Adams, Reference Berman and Adams1997; Mason et al. Reference Mason, Walter and Mason1997; Sumegi et al. Reference Sumegi, Kovacs, Veres, Radnai, Varbiro, Bognar, Toth, Gallyas, Benedetti, Banhegyi and Burchell2005). Na+/K+-ATPase pumps are associated with the apical and basal tegumental membranes (Skuce et al. Reference Skuce, Anderson and Fairweather1987) and their inhibition by OZ78 could lead to the severe swelling of the basal infolds and the internal tissue flooding observed. The swelling of the ger cisternae could be due to a similar mechanism. Treatment of other organisms and tissues with artemisinins and free radicals also causes collapse of the membrane potential of mitochondria, leading to their swelling and inhibition of electron transfer and oxidative phosphorylation (Dubin and Stoppani, Reference Dubin and Stoppani2000; Wakabayashi and Karbowski, Reference Wakabayashi and Karbowski2001; Li et al. Reference Li, Mo, Shen, Sun, Wang, Lu, Gitschier and Zhou2005). This action could explain the mitochondrial changes seen and would have a major impact on energy production by the fluke. Such changes contribute to the process leading to cell death, or apoptosis (Dubin and Stoppani, Reference Dubin and Stoppani2000; Wakabayashi and Karbowski, Reference Wakabayashi and Karbowski2001; Li et al. Reference Li, Mo, Shen, Sun, Wang, Lu, Gitschier and Zhou2005). The degenerative changes in the tegumental and vitelline cells (and the vitelline follicles) may be indicative of this phenomenon.
Finally, the present study has highlighted an interesting anomaly between the SEM and TEM data: indeed, the lack of significant surface changes belies the widespread disruption evident internally. A similar observation has been made by O'Neill et al. (Reference O'Neill, Johnston, Halferty, Brennan, Keiser and Fairweather2009). The severe disruption of the tegumental syncytium and subtegumental tissues might contribute to the efficacy of this particular synthetic peroxide against F. hepatica (Keiser et al. Reference Keiser, Utzinger, Tanner, Dong and Vennerstrom2006b, Reference Keiser, Utzinger, Vennerstrom, Dong, Brennan and Fairweather2007a; Keiser and Morson, Reference Keiser and Morson2008b). It is difficult to see how the fluke could survive such major impairment to the tegumental system that protects the fluke against immunological, enzymatic, bile and drug attack. Moreover, it is unlikely that the affected flukes could produce any viable eggs, given the marked degeneration of the vitellaria. The present results have revealed some of the morphological changes induced by OZ78 and some of its tissue targets. It will be interesting to see if similar changes take place when haemin is incorporated into the culture medium in vitro and following drug treatment in vivo. Further research on the efficacy and actions of artemisinins and synthetic peroxides against F. hepatica is warranted, to consolidate and extend the promising results obtained to date.
In part, this study was supported by a grant from the European Union (DELIVER grant, no. FOOD-CT-200X-023025). J. Keiser (Project No. PPOOA—114941) is grateful to the Swiss National Science Foundation for financial support.
