Vertical Transmission of Hepatozoon in the Garter Snake Thamnophis elegans

Vertical transmission of blood parasites is a significant mode of infection in humans (e.g., Ouédraogo et al. 2012) and domesticated animals (e.g., Murata et al. 1993). However, only a single anecdotal case has been reported in wild vertebrates: transmission of Hepatozoon sp. to a whole litter of 12 neonates by a heavily infected mother of the banded water snake (Natricia fasciata; Lowichik and Yaeger 1987). Water snakes give birth to live neonates that do not depend on the mother for further nutrition. Thus, infection of neonates is likely to have occurred during gestation, with the hemoparasites passing through the placenta, as has been reported for dogs (Canis lupus familiaris; Murata et al. 1993). We tested for evidence of vertical transmission of Hepatozoon sp. in wild-caught western terrestrial garter snakes (Thamnophis elegans), which naturally harbor the hemoparasite and like the water snake are viviparous.

In reptiles, Hepatozoon parasites are commonly transferred through horizontal transmission (Smith et al. 1996; Telford 2008). Early stages of Hepatozoon development (oocysts and sporocysts) occur in an invertebrate definitive host, generally a mosquito. Parasites progress to a cystic stage when the invertebrate host is ingested by a vertebrate (e.g., an anuran), which becomes the first intermediate host. The second intermediate host (e.g., a snake or lizard) then becomes infected by ingesting the first intermediate host, and within this second vertebrate the parasites (micromerozoites) enter the erythrocytes and mature into gamonts. The cycle is completed when a mosquito feeds from a reptile with infected blood cells. Although horizontal transmission seems to be the general mode of infection, vertical transmission might represent an overlooked alternative route by which Hepatozoon spp. spread among reptilian hosts in natural populations. Unlike horizontal transmission, vertical transmission should not depend on an invertebrate host and a primary intermediate host. We tested whether wild-caught pregnant garter snakes with naturally occurring Hepatozoon infection give birth to infected offspring in a laboratory environment where horizontal transmission is precluded by the absence of invertebrate hosts and potentially infected prey items.

Nine gravid females were caught along Eagle Lake (40°39′15.37′′N, 120°44′47.28′′E) in the California Sierra Nevada Mountains during June 2010. The natural diet of these snakes consists primarily of anurans and fish (Kephart and Arnold 1982). Females were transported to Iowa State University and housed individually in glass aquaria within an indoor facility. Aquaria were positioned with one end on a heating element that provided a thermal gradient for thermoregulation. Snakes were maintained on 12:12 light:dark cycles and fed thawed frozen mice pinkies once per week to satiation. Females were bled twice while in captivity, at preparturition (August) and postparturition (October), and blood smears were prepared, air dried, fixed with methanol, and stained with Wright's stain. For six of the nine females, we also had blood smears prepared at the time of capture in the field as part of another study. Each female gave birth on a single day between August and September. Litters ranged from 8–15 offspring, of which 5–8 randomly selected individuals were available for this study. Offspring were housed individually and fed thawed, frozen mice pinkies. Offspring were bled and blood smears prepared twice, once at around 2 mo and again at around 1 yr of age.

Smears were randomized and assigned ID numbers. Each smear underwent an initial scan at 400× across an area of roughly one fourth to one third of the total slide (approximately 500,000 erythrocytes) to determine the presence or absence of parasites. If no parasites were observed within this area, then parasites were classified as absent (A). If at least one parasite was observed within this area, parasites were determined to be present (P). After finding the first parasite, scanning continued until all of the area was scanned or until a second parasite was found, which always occurred within approximately 20,000 erythrocytes. If only one parasite was observed, intensity was considered to be ≤1/500,000 erythrocytes. If a second parasite was found, parasite load was further quantified by scanning smears at 1,000× and counting the number of parasites present per 2,000 erythrocytes at a random location (Madsen et al. 2005). If parasites were not detected during this scan, then parasitemia for that individual was determined to be <1/2,000 erythrocytes (but still ≥2/20,000), as two parasites were initially detected within this area when scanning at 400×. We tested for differences in parasite intensity between sexes using a Mann-Whitney test and between sampling time points for offspring using a paired Wilcoxon Sign-Rank test. We also tested for sex differences in parasite presence using a chi-square (χ²) analysis.

Parasites were determined to be members of the genus Hepatozoon (Apicomplexa, Hepatozoidae) based on comparisons of the gamonts to those described for other snake species (Telford et al. 2001; O'Dwyer et al. 2013). The species could not be determined, however, due to absence of information on other parasite stages (e.g., oocysts, sporocysts) and/or molecular data. Immature and mature gamonts were the most common forms observed—only two parasites were detected in what appeared to be the extracellular micromerozoite stage (Fig. 1). Four of the nine mothers had parasites present at preparturition and three of them also showed infection postparturition (Table 1). Two of these females (3 and 4) also showed infection at the time of capture in the field (no smears were available for the other two at this time point). The remaining five mothers had no parasites present at any time point sampled. All infected offspring (n=18) were born to the four infected mothers, although not all offspring in a litter were necessarily infected (Table 2). All offspring (n=33) born to the five mothers with no parasites also showed no parasites (Table 2).

The observed correlation between maternal and offspring parasitemia in a laboratory setting, where the necessary components for horizontal transmission were absent, strongly suggested that parasite infection was vertically transmitted from mothers to offspring. Even if an invertebrate host had been present in the laboratory environment, horizontal transmission would still have been precluded by the absence of a primary intermediate host. Smith et al. (1996) demonstrated that eastern garter snakes (Thamnophis sirtalis) did not become infected with Hepatozoon spp. when fed infected mosquitoes but needed to ingest infected frog livers to become infected. Therefore, the laboratory-born snakes in the present study would have had to ingest infected vertebrates in order to contract the infection, the chances of which are negligible given that individuals were fed exclusively thawed frozen mice.

Parasite loads were generally low in the infected mothers (range 0.004–2 per 2,000 erythrocytes), with values decreasing from preparturition to postparturition in all cases (Table 1). Interestingly, the two infected mothers for which we had smears made at capture then showed lower Hepatozoon loads than at preparturition (Table 1). We hypothesize that parasitemia could have increased initially in the laboratory due to the stress of captivity (Sparkman et al. 2014) which, together with a more advanced pregnancy, might have diverted resources away from the mothers' immune systems. After giving birth, and perhaps as they became more habituated to their captive environment, mothers may have been able to allocate more resources to immune function, thus lowering parasitemia.

For laboratory-born offspring, parasite loads per 2,000 erythrocytes were significantly higher (W=108; P=0.006, n=23) at 2 mo (median=1; range=0–64) than after 1 yr (median=0.2; range=0–10, Table 3). Interestingly, newborns showed the highest parasite loads in this study—even higher than their mothers at preparturition—which could be explained by the immature immune systems of the offspring which develop progressively with age (Sparkman and Palacios 2009). Infection prevalence did not differ between the sexes either at 2 mo (female [F]: 11/13; male [M]: 6/10; χ²=3.75, P=0.053) or at 1 yr of age (F: 10/13; M 6/10; χ²=1.066, P=0.3019). Parasite loads, however, were higher on average in female than in male offspring at 2 mo (U=33.5; P=0.045, n=23) but not at 1 yr of age (U=45.5; P=0.194, n=23).

Fitness consequences of Hepatozoon infection in T. elegans are unknown. They have been investigated in only two studies of free-living snakes, with contrasting results. Madsen et al. (2005) reported various fitness costs associated with heavy Hepatozoon sp. infection (up to ~1,000/2,000 erythrocytes) in water pythons (Liasis fuscus), but Brown et al. (2006) found no links between parasitemia and fitness in the keelback (Tropidonophis mairii), despite very high parasitemia levels (up to ~1,280/2,000 erythrocytes). Parasitemia levels in snakes from our study were considerably lower than those reported in the two fitness studies. In fact, in those studies only 1,000–2,000 erythrocytes were counted to determine infection status and estimate parasite loads whereas we scanned more than 500,000 erythrocytes in order to determine these parameters with confidence. Only one other published record exists on hemogregarines—likely Hepatozoon according to the authors—in T. elegans (Clark and Bradford 1969). In that study, only one of eight snakes captured in the field was infected, with only a single parasite found in a 15-min scan of the smear. The relatively low parasite loads observed in T. elegans suggest that fitness costs might not be large, but this remains to be tested. The use of molecular techniques in conjunction with microscopic study (Ujvari et al. 2004; O'Dwyer et al. 2013) will enhance our ability to detect Hepatozoon infection, allowing us to determine species identification, infection prevalence, and evaluate hypotheses. Regarding vertical transmission of Hepatozoon, molecular data would allow us to determine maternal and offspring haplotypes and test for evidence of parasite presence in placental tissues.

That vertical transmission is possible for hematozoan parasites with a complex life cycle in a vertebrate host having limited contact between mother and offspring (i.e., through a rudimentary placenta during gestation) suggests that this might be a common, but currently overlooked, mode of transmission for other disease-causing organisms and other vertebrate hosts in natural ecosystems. This can have important implications for disease ecology, as such a route enables disease transmission even in the absence of vectors and intermediate hosts, likely enhancing the rate of disease spread, especially for diseases caused by parasites with complex life cycles.

We thank M. Manes and M. Brandenburg for field support and C. Baughman for help with slides. This research was supported by grants from the National Science Foundation to A.B. (DEB-0323379, IOS-0922528). All animal handling procedures were carried out in accordance with standard animal care protocols and approved by Iowa State University Animal Care and Use Committee (3-2-5125-J). The State of California Department of Fish and Wildlife granted scientific research permits.