IHC score meaning brain liver spleen
0 negative 0% 0% 0%
1 mild ≤ 1 % <1 % ≤ 9 %
2 moderate 2 - 9 % 1 – 4 % 10 – 24 %
3 severe ≥ 10 % ≥ 5 % ≥ 25 %
inc inconclusive
pos positive, but proportion not definable
Table 5-1. Definition of immunhistochemistry (IHC) score. Depending on the proportion of positive cells per tissue sample a score for each organ was determined.
40 Manuscript IV
virus, genotype median no. of animals
NY, WT 0 1 deceased mice). Groups with no deceased animals (all survived until the end of experiment) are indicated with an asterisk.
Manuscript IV 41
Variable Mice Isolate P-value Significance Variable Mice Isolate P-value Significance Survival IFNAR -/- UG - DA 0.168218 IHC
Supplemental Table 5-1. Results of statistical analysis. Significant p-values are indicated with an asterisk in the next column.
42 General discussion and conclusion
6 General discussion and conclusion
The worldwide distributed flavivirus WNV has caused increasing public concern in the past decades. It is now endemic in the Americas and Europe and causes life threatening neurologic disease in humans and substantial economic losses. Furthermore, since its introduction to the American continent it has caused substantial population declines in certain wild bird species (LaDeau et al. 2007). While in the past lineage 1 has caused sporadic epidemics in Europe without noticeable wild bird mortality, this was indeed a characteristic sign of its activity in North America (Lanciotti et al. 1999). Additionally, the newly introduced lineage 2 in Europe affected especially birds of prey (Erdelyi et al. 2007; Bakonyi et al. 2013). In Europe both WNV lineages 1 and 2 can now be found (Barzon et al. 2011; Danis et al. 2011a; European Centre for Disease Prevention and Control 2011; Savini et al. 2012). Large outbreaks involving humans have occurred in Croatia, Greece, Italy, Serbia, Turkey and Romania in recent years (Papa et al. 2010; Sirbu et al. 2011; Kalaycioglu et al. 2012; Barzon et al. 2013;
Popovic et al. 2013; Kurolt et al. 2014). Human WNV infections during the last transmission season of 2015 were noticed in Italy, Hungary, Romania, Austria, Bulgaria, France and Portugal (European Centre for Disease Prevention and Control 2015). In Germany there is no evidence of an active WNV infection at the moment; WNV antibodies have been detected in migratory birds only (Ziegler et al. 2015).
It seemed that, among others, large falcons were potentially threatened by WNV (N. M.
Nemeth et al. 2009b; Wodak et al. 2011), but detailed data on the susceptibility and pathogenicity in this species were not available. Furthermore, humans handling these birds, thus coming into close contact with them, could be exposed to zoonotic risks. Therefore, our aim was to characterize the pathogenesis of North American lineage 1 and European lineage 2 isolates comparatively in large falcons to clarify, if large falcons are susceptible to these WNV isolates and could play a role in the WNV transmission cycle (manuscript I).
Subsequently, we analyzed two commercially available equine WNV vaccines and two DNA vaccines at research level for their safety, immunogenicity and efficacy in large falcons using different vaccination schemes (manuscript II and III). Furthermore, a European lineage 2 strain was characterized in comparison to known high virulent lineage 1 isolates and a low
General discussion and conclusion 43
virulent lineage 2 isolate from Africa by using adult wild-type and interferon type I (alpha/beta) receptor knock-out mice (manuscript IV).
Pathogenesis studies in large falcons (manuscript I and non-vaccinated control birds in manuscript II/III)
In manuscript I, our findings showed no differences in clinical signs between North American WNV lineage 1 (NY99) and European lineage 2 (Austria) infected large falcons. Diseased animals showed marked, but nonspecific general disease symptoms such as apathy, ruffled feathers and loss of appetite. Surprisingly, we did not observe any clear neurological signs in the falcons infected for the pathogenesis study (manuscript I), but the control animals used for the vaccination study (manuscript II) developed clear symptoms. Our study revealed differences between the two virus isolates in terms of influence of the infection dose on clinical signs and mortality of the falcons. In lineage 1 infected falcons both animals receiving the high dose succumbed to infection, whereas the other birds survived the experimental course developing only mild to moderate (low dose) or moderate to severe (high dose) disease. However, this was only partly true for the non-vaccinated control falcons of the studies described in manuscript II and III, in which four out of eight animals infected with the high dose succumbed to infection. In contrast, high dose lineage 2 infected birds were only mildly to moderately affected and survived the whole time course of the experiment. This result could be reproduced with two additional animals. On the other hand both animals which succumbed to WNV lineage 2 infection received a low or medium infectious dose. It should be noted, however, that animal numbers per infectious dose were too low to define dose dependent mortality rates accurately. In contrast to our findings, Gyr-Saker (Falco rusticolus x Falco cherrug) hybrid falcons did not develop clinical disease after WNV lineage 1 infection in a previously published study (Busquets et al. 2012). The subcutaneous infection dose used by this research group was comparable to our medium dose. Comparably, in experimentally infected American kestrels neither clinical illness nor death occurred, although the birds developed viraemia (Komar et al. 2003; N. Nemeth et al. 2006). Obviously, development of clinical symptoms and death is not easy to predict and needs further elucidation, as we experienced with the animals infected with WNV lineage 2. Although our needle inoculation WNV infection model was artificial, infection by a WNV infected
44 General discussion and conclusion
mosquito bite would not likely lead to a lower mortality rate or less clinical disease, regarding the fact that Culex species inoculate mouse or chicken with high doses of WNV (over 104 - 106 pfu per bite) (Styer et al. 2007).
Independently of the isolate used, WNV genome shedding per oral and cloacal route and WNV viraemia were present for a long period (until 10 to 21 dpi) with high peak viraemia on 4 to 6 dpi. High viral genome loads could be demonstrated in all organs of deadly affected animals, but also in surviving animals, which were killed at the end of the experiment. In particular brain, spleen, kidney and heart were affected. These results were also seen with the control falcons of the vaccination studies (Manuscript II and III). Hence these organs are most suitable for diagnostic purposes using quantitative real-time RT-PCR (qRT-PCR).
Additionally, all infected falcons seroconverted subsequently.
In a previous study, others concluded from the viraemia profile after experimental WNV lineage 1 infection (corresponding to our medium infection dose) that Gyr-Saker hybrid falcons are no competent reservoir hosts, as peak viraemia was 104 TCID50/ml (Busquets et al. 2012). Comparable to our results American kestrels developed high peak viraemia after WNV lineage 1 infection and were therefore considered as competent reservoir hosts (Komar et al. 2003; N. Nemeth et al. 2006). The viraemic titer was high, but duration of viraemia was shorter (4-5 days) than in our study. Species specific differences might have contributed to this differences (Komar et al. 2003). In agreement with our results, high titer viral shedding was observed in the kestrels for up to ten days, whereas a cell culture based method was used in contrast to the more sensitive qRT-PCR in our study (Komar et al. 2003). The authors also found viral titers in kidney and spleen even after 14d post infection and interpreted this as an indication for persistent infection. Likewise we detected high genome loads and also viral titers (in single birds) 14d post infection in brain and kidney of lineage 1 infected falcons, and persisting high viral genome levels in the brain of lineage 2 infected falcons even after 21d post infection. Comparably, in single falcons in manuscript II and III low viral titers could still be found in the brain three weeks after infection. It is speculative, whether this can already be classified as a persistent infection or whether virus would have been completely cleared after one or two additional weeks. However, persistent infection (with infectious viral particles; viral RNA could be isolated even for longer times) was demonstrated in animal
General discussion and conclusion 45
models using house sparrows (>40 d p.i.), immunocompetent mice (4 month p.i.) and hamsters (>240d p.i.) (Tesh et al. 2005; N. Nemeth et al. 2009a; Appler et al. 2010).
The lack of clear neurological symptoms described in manuscript I might be due to extensive peripheral lesions which defined the clinical picture and led to death in severely infected falcons before neurological lesions could develop. Individual host derived factors might additionally have influenced the clinical severity of infection. Others also completely failed to provoke clinical symptoms and death in experimental infection of falcon and hawk species (American kestrels, red-tailed hawks) without a definite explanation, but concluded that morbidity and death are variable (influenced by hosts, virus strain, immune suppression or underlying illness) (Komar et al. 2003; N. Nemeth et al. 2006; Redig et al. 2011). However, neurological symptoms (ataxia, head tilt, seizures, recumbency) were observed in two of the eight non-vaccinated control birds during WNV vaccination studies of large falcons in manuscript II and III. This was in accordance with previous studies, which showed that in natural WNV infection, neurological symptoms like ataxia, tremors and seizures were characteristic of WNV lineage 2 infections after introduction of this strain to Europe in birds of prey (Erdelyi et al. 2007; Richter et al. 2009) and were also frequently seen in North American raptor species infected with WNV lineage 1 (N. M. Nemeth et al. 2009b). It must be noted that the surviving but markedly diseased animals in our study might have died in the wild, supposedly due to inappetence, weakness and possibly concurring diseases like parasitosis.
In comparison to experimental WNV infection studies in falcons, natural WNV lineage 1 infection was reported in American kestrels (Falco sparverius) and a single peregrine falcon (Falco peregrinus) (N. M. Nemeth et al. 2009b). In ten captive-held gyrfalcons (Falco rusticolus) one case of WNV infection with WNV lineage 2 was found confirming general susceptibility for this lineage (Wodak et al. 2011), which was confirmed by our results in manuscript I also for adult healthy gyrfalcons (and gyrfalcon-hybrids). Birds of prey are assumed to be selectively vulnerable for WNV lineage 2 in Europe, which was proposed to be based on a comparable molecular mechanism like the one responsible for the increased virulence of North American WNV lineage 1 isolates in American crows (A. C. Brault et al.
2007; Wodak et al. 2011). However, our results point out similar infection courses in large
46 General discussion and conclusion
falcons for WNV lineage 1 and 2 isolates (as seen by clinical observation and virological and histopathological analysis) with indicators for individual vulnerability in regards to lethal infectious doses.
In manuscript I, most animals showed moderate to severe alterations with acute suppurative meningoencephalitis, acute to subacute necrotizing myocarditis and non-suppurative arteritis as prominent features. Findings in non-vaccinated controls in manuscripts II and III were comparable, although arteritis was recorded rarely (n=1). Hence in diagnostics, WNV must be considered as differential diagnosis in birds showing only one of the histopathological signs described above. These pathological results are comparable to pathological alterations in experimentally and naturally WNV infected raptor species (hawks, American kestrels, gyrfalcon and gyr-saker hybrid falcons), in which non-suppurative encephalitis, arteritis and myocarditis were frequent alterations in addition to nephritis, hepatitis and myositis (N. Nemeth et al. 2006; Erdelyi et al. 2007; Wodak et al. 2011;
Busquets et al. 2012). Antigen was detected in brain and heart by immunohistochemistry (IHC) in manuscript I. In manuscripts II and III antigen detection by IHC was shown to be dependent of the time-point of assessment, whereas distinct amounts were only recorded in animals, which succumbed to infection. In the birds that survived for three weeks only marginal amounts of antigen could be detected, if any. Thus IHC is not a sensitive detection method for infections that are already overcome and is only valuable for detection in birds at the peak of infection.
In conclusion, WNV lineage 1 (NY99) and lineage 2 (goshawk Austria 2009) isolates both are highly pathogenic in large falcons. Even low dose inoculation led to viraemia, viral genome shedding and clinical symptoms in all animals. In several birds, the viral titer in whole blood exceeded 105 TCID50/ml blood. As the threshold viraemia titer for infectiousness for Culex mosquito species is considered to be 105 pfu per ml, falcons may be competent reservoir hosts for WNV (Turell et al. 2000; Komar et al. 2003). However this experimental model is of limited value for making a definite statement, as mosquito infection studies might reveal a more complete picture and mosquito feeding behavior in the wild is influenced by multiple factors (Kilpatrick et al. 2006; Kilpatrick 2011; Janousek et al. 2014).
However, viral shedding, potential infection of mosquitoes and high viral loads in organs of
General discussion and conclusion 47
large falcons represent a risk for humans handling infected animals or carcasses. Clinically and morphologically, infections with lineage 1 and 2 isolates cannot be distinguished in large falcons, both lead to the same characteristic signs of infection. Dose dependent differences in the severity of infection could only be shown for lineage 1. Nonetheless lineage 2 infections were also deadly for large falcons. This leads to the assumption that individual constitutional factors of the animals also influence the course of infection. Moreover, severe pathological alterations were seen in all animals; even in animals which recovered clinically post mortem examinations revealed long-term viral persistence in organs and severe chronic lesions for example in the heart. This suggests that the lethality rate in wild falcons might be higher than in our experiments, when severe disease, inappetence, apathy and organ lesions coincide with the need for food, bad weather influences and possibly additional impairing factors, e.g.
parasitosis or trauma. Furthermore falcons kept in commercial falconries might achieve only poor flight performance after infection.
Vaccination studies falcons (manuscripts II and III)
In the WNV pathogenesis study described above (manuscript I) we proved that large falcons are potential competent reservoir hosts for WNV and can become substantially ill and may succumb to infection. WNV is endemic in Europe and a potential threat for wild populations of raptors as well as for valuable breeding stocks of large falcons. Furthermore infected falcons pose a zoonotic risk for humans handling the affected birds. Unfortunately, for birds so far no approved vaccine is available. To evaluate the effect of vaccination against WNV in large falcons, a commercially available equine inactivated vaccine and a recombinant live vaccine (manuscript II) as well as two DNA vaccines at research level (manuscript III) were evaluated in terms of safety, immunogenicity and efficacy. For this purpose different vaccination schemes were compared. The pathogenesis study described above failed to define a challenge dose for WNV lineage 2 isolate (goshawk Austria = GO), so vaccine efficacy studies were performed with WNV NY99 (106 TCID50) as challenge virus only. All non-vaccinated control birds were infected successfully as described above. This shows that this infection model is a robust model to evaluate vaccine efficacy regarding viraemia, clinical symptoms and shedding, but prevention of death cannot be evaluated precisely.
48 General discussion and conclusion
The inactivated vaccine was tolerated very well with mild reactions only seen in single birds in pathological examinations (mild phlebitis and lymphocytolysis of locally aggregated follicles, each n=1). With a two shot regimen, only transient seroconversion occurred with low titers. Clinically, only mild improvement was obvious during live viral challenge, with one animal even succumbing to infection. This mirrors laboratory analysis with only mild reduction in viraemia and shedding, titers potentially high enough for reinfection of mosquitoes. Additionally, organ viral loads were not reduced as compared to control falcons.
Moreover, all animals revealed distinct pathological alterations despite vaccination. In conclusion, this easy-to-produce, on the hand available and safe vaccine is not suitable for the protection of large falcons against WNV, if given twice. In contrast, using a more frequent vaccination three shot regimen, a distinct and more constant seroconversion was obvious in all animals. Consequently, a good improvement in clinical symptoms, shedding and viraemia was achieved (low level). However, WNV-typical histopathological alterations or viral RNA in organs were still observed in single birds, but in most cases to a lower degree than in control animals. Altogether a robust partial protection was achieved, attenuating, but not preventing WNV infection in large falcons with transient viraemia below mosquito infectiousness level, which is considered to be 105 pfu/ml (if 1 TCID50 ~ 0.7 pfu) (Turell et al. 2000; Komar et al. 2003; Carter u. Saunders 2007). Frequent intramuscular vaccination might be problematic in commercial falconries as well as in wildlife vaccination programs due to monetary and organizational reasons. Similar studies with different species were conducted previously by others showing a failure of seroconversion in red-tailed hawk (Buteo jamaicensis), which is in contrast to our results and might be explained with the reduced vaccination dose used in that study or with species-specific differences between hawks and falcons (Nusbaum et al. 2003). In contrast this vaccine achieved 58% seroconversion rate at a low titer in a mixed group including raptor species and corvids, however only if the full 1 ml dose was applied thrice (Johnson 2005). Multiple full dose applications seem to be necessary for a good seroconversion rate in birds, which corresponds with our results. Inactivated WNV vaccines were also evaluated with overall promising results in other birds species like cranes (with lack of seroconversion, but immune-priming and partial protection in a challenge trial), flamingos (variable seroconversion), geese (inducing robust seroconversion and protection against lethal intracranial challenge model) and penguins (good seroconversion, but no
General discussion and conclusion 49
challenge experiment) (Samina et al. 2005; Okeson et al. 2007; Samina et al. 2007; M. R.
Davis et al. 2008; Olsen et al. 2009).
Application of recombinant canarypox-vector based vaccine yielded a generally lower serological response than application of the inactivated vaccine. Two shots had only short-term effects, but three shots led to seroconversion in the majority of the group. Although no indication of replication of the canarypox vector was seen, clinical acceptance was problematic. This was confirmed in pathological examination, as extensive granulomatous myositis at the vaccination site occurred in all vaccinated birds receiving recombinant vaccine. In the recombinant three shot group one animal died one week post vaccination most probably due to severe endoparasitosis. We do not believe this was a side effect of vaccination, although stress due to change of housing and vaccination might have contributed to a multifactorial systemic break down in this single bird. Canarypox virus vector based vaccines are successfully used in mammal species like horses, dogs and cats. In mammalian species this vector is non-replicative, despite expression of the desired foreign proteins. The basis of the Recombitek vaccine is the ALVAC vector, which is recovered from a live attenuated canarypox virus vaccine for canary birds by plaque-purification. ALVAC is replicative but non pathogenic in canaries, very host restricted and has been tested to be safe in chickens and crows, although it replicates in vitro in chicken embryo fibroblasts (Minke et al. 2004; Poulet et al. 2007). Hence replication of the recombinant vaccine virus in large falcons is unlikely, but extensive granulomatous inflammations at the vaccination sites could indicate that. Beside that, these alterations could also be due to the adjuvant (polyacrylic acid) used, although it was safely used in vaccination of geese before (Gelfi et al. 2010). Further studies could clarify this. In accordance with our results previous reports described similar alterations at the vaccination site in Western scrub-jays (Aphelocoma californica) and attributed these to a possible replication of the vaccine vector (Wheeler et al. 2011). Such adverse reactions might limit the use of Recombitek® in valuable performance falcons as well as in wildlife capacities, where inflammations in the pectoral muscle might have a negative impact on survival. Since no shedding of the vaccine occurred, environmental safety is adequate.
50 General discussion and conclusion
All animals in both recombinant vaccinated groups survived the three weeks challenge period and clinical disease was improved in comparison to the control birds and was moderate (two shot vaccination) or short and mild (three shot vaccination). Shedding and viraemia were reduced significantly with the recombinant three shot regimen leading to the best results (low or even prevented WNV viraemia and shedding). A potential WNV reinfection to mosquitoes was prevented after triple administration of Recombitek®. Organ viral loads were reduced
All animals in both recombinant vaccinated groups survived the three weeks challenge period and clinical disease was improved in comparison to the control birds and was moderate (two shot vaccination) or short and mild (three shot vaccination). Shedding and viraemia were reduced significantly with the recombinant three shot regimen leading to the best results (low or even prevented WNV viraemia and shedding). A potential WNV reinfection to mosquitoes was prevented after triple administration of Recombitek®. Organ viral loads were reduced