Infesting ticks on infected natural hosts remains the method most closely resembling the normal acquisition of a virus in the wild. Direct feeding on infected host can be facilitated by using feeding bags (Figures 1a,b) or feeding chambers (Figures 1c,d). However, the maintenance and handling of animal hosts can be expensive and difficult, particularly for wild animals (Bonnet and Liu, 2012). The direct feeding technique also lacks the capacity to quantify the pathogen dose acquired by the tick during or post feeding. The technique may also not be appropriate for virus strains not suited for replication in the vertebrate hosts (Mitzel et al., 2007). In addition, it remains a challenge to synchronize viremia with tick feeding, and for ethical reasons, the use of alternative artificial methods in infecting ticks without the use of laboratory animals is still preferred (Bonnet and Liu, 2012).

Various hosts, mostly small laboratory animals, have already been infected for direct tick acquisition of the virus. Dermacentor andersoni ticks were previously infected by infesting rabbits injected intravenously with large doses of the Powassan virus (Chernesky, 1969). Laboratory mice were also previously used to study severe fever with thrombocytopenia syndrome virus (SFTSV) transmission by Haemaphysalis longicornis (Luo et al., 2015), while Rhipicephalus appendiculatus specimens were infected with the Thogoto virus (THOV) by allowing them to feed on THOV-infected Syrian hamsters (Booth et al., 1989). Transmission of West Nile virus from infected mice to naïve I. ricinus nymphs through direct blood feeding was also previously observed (Lawrie et al., 2004).

Non-viremic transmission, or co-feeding transmission (Figure 2c), is an important transmission mechanism for TBVs established by Jones et al. (1987). It occurs between infected and uninfected ticks when they co-feed in close proximity on susceptible hosts, even when these hosts do not develop viremia (Alekseev and Chunikhin, 1990; Labuda et al., 1993; Jones et al., 1997; Labuda et al., 1997a,b). Though co-feeding is an established natural tick infection method, it requires an animal host for feeding and may not produce high infection rates, as transmission of the virus from infected and uninfected ticks greatly depends on the proximity or distance among feeding ticks.

Co-feeding experiments were mostly conducted in small laboratory or wild animals. Virus transmission experiments using yellow-necked mice (Apodemus flavicollis) and bank voles (Clethrionomys glareolus) (Labuda et al., 1996, 1997b), BALB/c mice (Khasnatinov et al., 2009; Slovák et al., 2014), European hedgehog (Erinaceus europaeus), striped field mouse (A. agrarius) European pine vole (Pitymys subterraneus), and common pheasant (Phaseanus colchicus) (Labuda et al., 1993) were used to study TBEV transmission by I. ricinus.

Co-feeding transmission of the Louping ill virus on I. ricinus was also evaluated in mountain hares (Lepus timidus), New Zealand white rabbits (Oryctolagus cuniculus) and red deer (Cervus elaphus) (Jones et al., 1997). Non-viremic transmission was also established for Thogoto virus (THOV) on R. appendiculatus (Jones et al., 1987, 1997) and CCHFV on Hyalomma truncatum, H. impeltatum (Gordon et al., 1993) and Amblyomma variegatum (Gonzalez et al., 1991) using guinea pigs (Cavia porcellus). Co-feeding transmission was also observed for Bhanja virus and Palma virus in D. marginatus, D. reticulatus, and I. ricinus ticks infested on mice (Labuda et al., 1997a) and Heartland virus in Amblyomma americanum infested on rabbits (Godsey et al., 2016). Lastly, co-feeding transmission of THOV was recently demonstrated in H. longicornis ticks infested on BALB/c mice (Talactac et al., 2018).

Another alternative to tick infestation is through membrane feeding. Membranes from animal and non-animal origin (e.g., silicone membranes) are usually utilized, with variable success, to feed ticks. This method could also be used for studies on the dynamics of pathogen transmission, since it can reduce the variation within a given treatment group because the blood meal from the same donor reduces the variation that may arise from individual tick–host relationships (Krober and Guerin, 2007).

However, this method requires chemical and physical stimuli to enhance attachment by hard ticks to membranes (Kuhnert, 1996). Its use may also depend on the length of the hypostome in all life stages of the hard ticks to be studied (Krober and Guerin, 2007). In addition, this type of artificial feeding is more challenging for ixodid ticks, since they require longer time for attachment (de Moura et al., 1997. This method was previously used in infecting I. ricinus, I. hexagonus, D. reticulatus, and R. bursa with the Bluetongue virus (Bouwknegt et al., 2010). The unlikely involvement of I. ricinus and D. reticulatus as biological vectors of African swine fever virus was also shown using membrane feeding (De carvalho Ferreira et al., 2014).

The introduction of pathogens to ixodid ticks via capillary feeding was first attempted by Chabaud (1950). In this technique, the ticks are normally pre-fed on animals, followed by a careful mechanical removal of ticks from the host. Eventually, a capillary tube containing the pathogen is placed over the tick’s mouthparts, and the tick is immobilized on a slide (Burgdorfer, 1957; Bouwknegt et al., 2010). Capillary feeding provides a number of advantages, especially that it mimics the natural route of infection of ticks, and it can estimate the amount of pathogen to be introduced. However, maintaining the integrity of the mouthparts of the ticks after removal is crucial for a successful capillary feeding (Bonnet and Liu, 2012). This technique was previously used in infecting A. variegatum and R. appendiculatus with the Dugbe virus (Steele and Nuttall, 1989) and I. ricinus, I. hexagonus, D. reticulatus, and R. bursa with the Bluetongue virus (Bouwknegt et al., 2010).

Direct injection of the virus inoculum through the cuticle (between the coxa and trochanter) has the advantage of estimating the viral dose received by the ticks (Figure 2a). However, this method bypasses the midgut barrier of ticks during feeding, making it unrepresentative of the natural route of infection for ticks (Mitzel et al., 2007). This technique also requires a microinjector to efficiently introduce the inoculum into the tick and may produce higher tick mortality due to injection injury (Rechav et al., 1999). Previous studies using this technique include the infection of H. longicornis with the Langat virus (Talactac et al., 2016, 2017a), A. variegatum with the Thogoto virus (Kaufman and Nuttall, 1996), I. ricinus with TBEV (Labuda et al., 1993, 1996, 1997b; Khasnatinov et al., 2009; Belova et al., 2012), and the Louping ill virus (Jones et al., 1997). D. marginatus, D. reticulatus, and I. ricinus ticks also previously received percoxal injections with the Bhanja and Palma viruses (Labuda et al., 1997a). Alternatively, anal pore or rectal injection of the virus directly into the gut of the tick can be used (Figure 2b), though it also requires skill to avoid puncturing the gut upon injection. This method has been used to infect H. truncatum with CCHFV (Gonzalez et al., 1989), I. ricinus with TBEV (Belova et al., 2012), and H. longicornis with Langat virus (Talactac et al., 2017b) and THOV (Talactac et al., 2018).

Infection of ticks through immersion provides a low cost and relatively simple artificial method, since it can synchronously infect a large number of ticks with a defined virus stock. The ticks are believed to be infected when they successfully swallowed the immersion medium containing the virus; with the ingested virus ultimately reaching the midgut (Mitzel et al., 2007). The virus can also possibly penetrate the tick’s exoskeletons, especially the immature ones. Larvae and nymphs have less sturdy exoskeleton, since arthropods must be able to hydrolyze the chitin for cuticle degradation and development during the immature stages (You et al., 2003). However, its major limitation is the generation of cohorts of infected ticks with an equal pathogen burden (Kariu et al., 2011). Infection of ticks using this method was previously reported for I. scapularis infected with Langat virus (Tumban et al., 2011; McNally et al., 2012), the dengue virus (Tumban et al., 2011), and TBEV (Mitzel et al., 2007) and for A. americanum infected with the Heartland virus (Godsey et al., 2016).