Many pathogens consist of genetically distinct strains. When hosts are simultaneously infected with multiple strains the phenomenon is known as a mixed infection or a co-infection. In mixed infections, strains can interact with each other and these interactions between strains can have important consequences for their transmission and frequency in the pathogen population. Vector-borne pathogens have a complex life cycle that includes both a vertebrate host and an arthropod vector. As a result of this complexity, interactions between strains can occur in both the host and the vector. Interactions between strains in the vertebrate host are expected to influence transmission from the co-infected host to uninfected vectors. Conversely, interactions between strains in the arthropod vector are expected to influence transmission from the co-infected vector to the uninfected host. This thesis used the tick-borne bacterium, Borrelia afzelii, as a model system to investigate how co-infection and interactions between strains influence their transmission and lifetime fitness over the course of the tick-borne life cycle. B. afzelii is a common cause of Lyme disease in Europe, it is transmitted by the castor bean tick (Ixodes ricinus) and it uses small mammals (e.g. rodents) as a reservoir host. An experimental approach with two genetically distinct strains of B. afzelii (one Swiss stain, one Finnish strain) was used to investigate the effects of co-infection in both the host and the vector. In Chapter 1, lab mice were experimentally infected via tick bite with either 1 or 2 strains of B. afzelii. The infected mice were then fed upon by I. ricinus ticks from a laboratory colony to quantify host-to-tick transmission. qPCR was used to determine the presence and abundance of each strain in the ticks. Chapter 1 found that co-infection in the mice reduced the host-to-tick transmission success of the strains. This chapter also found that co-infection reduced the abundance of each strain in the tick. This is one of the first studies to show that co-infection is important for determining the abundance of the pathogen strains in the vector. In the lifecycle of B. afzelii, the bacterium is acquired by larval ticks that blood feed on an infected host. These larvae subsequently moult into nymphs that are responsible for transmitting the bacterium to the next generation of hosts. The bacterium has to persist inside the midgut of the nymph for a long time (8 – 12 months). Chapter 2 investigated whether nymphal ageing (1-month-old vs 4-month-old nymphs) under different environmental conditions (summer vs winter) influenced the interactions between strains in co-infected ticks. The spirochete abundance inside the nymph decreased with nymphal age, but there was no effect of the environmental conditions investigated. In Chapter 3, the presence and abundance of the two strains of B. afzelii were quantified in the tissues of 6 different organs (bladder, left ear, right ear, heart, ankle joint, and dorsal skin) that were harvested from the co-infected and singly infected mice. This study showed that co-infection in the mouse host reduced the prevalence of the Finnish strain in the host tissues (but the Swiss strain was not affected by co-infection). Chapter 3 found a positive relationship between the prevalence (or abundance) of each strain in the mouse tissues and the host-to-tick transmission of each strain. External tissues (e.g. ears) were more important for host-to-tick transmission than internal organs (e.g. bladder). Chapter 3 enhances our understanding of the biology of mixed infections by showing the causal links between co-infection in the host, the distribution and abundance of the strains in host tissues and the subsequent host-to-tick transmission success of the strains. Chapter 4 investigated how co-infection in the arthropod vector influences vector-to-host transmission success. A second infection experiment was performed, where naïve mice were exposed to nymphs that were either co-infected or infected with one of the two strains (i.e., using the nymphs generated in Chapters 1 and 2). The infection status of the mice was then tested using the same qPCR-based methods. Importantly, Chapter 4 confirmed that the negative effect of co-infection in the mouse on host-to-tick transmission (observed in Chapters 1, 2, and 3) had real fitness consequences for subsequent tick-to-host transmission. Ticks that had fed on co-infected mice were much less likely to transmit their strains to the host because these strains were less common inside these co-infected ticks. Chapter 4 did not find evidence that co-infection in the nymph influenced the nymph-to-host transmission success of each strain. This Chapter did find that there was a two-fold difference in nymph-to-host transmission success between the two strains. This work provides evidence for the idea that vector-borne pathogen strains can exhibit trade-offs across the different steps of their complex life cycles. In the co-infected mice, the Swiss strain had higher host-to-tick transmission success than the Finnish strain. Conversely, the Finnish strains had higher spirochete loads in the tick vector and had tick-to-host transmission success. Thus, the Swiss and Finnish strains are specialized on the host versus the vector, respectively.
Natural reassortment of a segmented RNA arbovirus illustrates plasticity of phenotype in the arthropod vector and mammalian host in vivo
