BARTONELLA INFECTIONS

RICHARD BIRTLES

School of Environment and Life Sciences, University of Salford, Greater Manchester, UK

Infections caused by Bartonella species are now widely recognized as being of public health and veterinary impor­tance.

AETIOLOGY

Bartonella species are Gram-negative bacteria with a spe­cific tropism for mammalian erythrocytes. They exploit a wide range of mammalian species, including humans, livestock, companion animals and wildlife, as reservoir hosts. In general, each Bartonella species is adapted to a specific mammal species, and in this host, infections tend to be chronic but asymptomatic. Several Bartonella species are pathogens of veterinary and public health sig­nificance, but disease usually results from ‘accidental’ infection of a non- r eservoir species or when a reservoir host is immunocompromised.

There are currently 24 valid Bartonella taxa, although, as numerous other strains have been partially described, the true breadth of the genus is likely to be considerably wider than is presently recognized. The Bartonella genus was formally removed from the Rickettsiales in the mid 1990s, and more recently, its members have also been distinguished from mycoplasma-like Haemobartonella and Eperythrozoon species.

EPIDEMIOLOGY

GEOGRAPHICAL DISTRIBUTION AND HOSTS

Bartonella species are obligate parasites of mammals and arthropods. The range of mammals parasitized by bar­tonellae is broad, and includes humans, companion animals and livestock as well as wildlife. Bartonellae have been associated with all of the most common mammalian orders, including Artiodactyla, Chiroptera, Carnivora, Diprotodontia, Insectivora, Lagomorpha, Primates and Rodentia.

Bartonellae have been encountered on all continents except Antarctica. In Europe, the medical significance of B. henselae has provoked a large number of surveys of domestic cats (the reservoir host) in many different coun­tries. Surveys for bartonellae in other mammalian species, although less frequently reported, are nonetheless numer­ous. Bartonellae have been recognized by microscopic observation of blood smears drawn from various wildlife species for over 100 years. Contemporary surveys of wild­life for Bartonella species have encountered infections in many different species across Europe. Most surveys have focused on rodents, with infections reported in numerous countries from Ireland, the UK, France and Portugal in the west, to Russia in the east, and from Scandinavia in the north to countries bordering the Mediterranean in the south. Infections have been recorded in numerous rodent taxa, including, most commonly, mice of the genera Apo- demus and Mus, and voles belonging to the genera Myodes and Microtus. Infections in other common and widespread rodents such as Rattus norvegicus and Sciurus vulgaris have also been reported. European shrews (Soricidiae) also serve as reservoir hosts to bartonella infections, as do other small mammals including bats (Chiroptera) and rabbits (Lepori- dae, Lagomorpha). Evidence for infections in mustelids has yet to be produced. There is also scant evidence for Bartonella infections in European large carnivores, although very few have been tested; infections have been described in such animals elsewhere in the world.

The application of molecular methods, most commonly comparative DNA sequence analysis, has demonstrated that, in general, Bartonella species are host-specific. Among the 24 valid Bartonella taxa, only a handful have been shown to exploit more than one mammal species as reser­voir hosts; these Bartonella species (B. doshiae, B. grahamii and B. taylorii) are associated with rodents, and have been shown to be able to parasitize sympatric host species. Fur­thermore, a specific host may be parasitized by more than one Bartonella species; for example, felids appear to serve as reservoir hosts for B. henselae, B. clarridgeiae and pos­sibly B. koehlerae.

ENVIRONMENTAL FACTORS

The epidemiology of bartonella infections in reservoir hosts is generally characterized by seasonality, with infec­tion prevalence rising to a peak in late summer and early autumn before declining during winter and early spring. Different studies have reported varying infection preva­lences, but, in general, a significant proportion of the population is infected. For example, in the UK, the preva­lence of bartonella infections in field voles (Microtus agres- tis) ranges between about 10% in spring to 50% in late summer.

HOST FACTORS

There is no clear evidence that host factors such as age or sex have a significant influence on susceptibility to infec­tion, however, as would be expected, both host population density and the population dynamics of arthropod vectors are important determinants of bartonella transmission. For example, there is an inverse correlation between prevalence of B. henselae infections in domestic cats and latitude, which is thought to reflect the relatively small populations of cat fleas ( Ctenocephaledes felis) in high latitude countries compared with the larger populations that are supported in warmer and wetter climates.

EPIDEMIOLOGICAL ROLE OF WILD ANIMALS

As obligate parasites of mammals, Bartonella species have evolved to exploit their reservoir hosts in a manner that best facilitates their transmission and natural persistence. The current paradigm for bartonellae is a strategy for host exploitation involving chronic, subclinical infections of the blood rather than overt disease or pathology, and such a strategy is clearly appropriate given their dependence on arthropods for transmission; blood- feeding ectoparasites are far more likely to be encountered by healthy, fit individuals.

TRANSMISSION

The route of acquisition of a bartonella infection remains a matter of debate, but the current paradigm is one in which bacteria are shed in the faeces of arthropods onto the skin of the host, then this contaminated material is inoculated through the skin as the host scratches itself to relieve the irritation of being bitten or infested. Direct transmission during blood feeding by an arthropod prob­ably also occurs for some Bartonella species.

PATHOGENESIS, PATHOLOGY

AND IMMUNITY

Once inside the host, bartonella are thought to migrate to a primary niche, possibly vascular endothelial cells or mac­rophages. From here the bacteria are released into the bloodstream, where they encounter and then invade eryth­rocytes. Inside the red cells, bartonellae initially replicate, but the number of intracellular bacteria soon stabilizes. The integrity of infected erythrocytes does not appear to be significantly affected, and they are thought to remain in circulation for as long as there are uninfected cells, thereby facilitating their uptake by haematophagous insects. Studies of both experimentally and naturally infected reservoir hosts have demonstrated that bartonella infections are chronic, lasting weeks, months or even years. Investigations into the molecular basis of infection have demonstrated the presence of specific factors such as adhe­sions and specialized secretion systems in most Bartonella species that mediate host interaction.

Infection of a reservoir host does not normally appear to provoke pathology. For example, the author has never observed any gross pathological signs associated with Bar­tonella infections during dissections of large numbers of field voles (Microtus agrestis), despite some animals having a very poor body condition and/or having other debilitat­ing conditions or disease. However, usually mild patholo­gies associated with Bartonella infections have been described, albeit very rarely, in those companion animals that serve as reservoir hosts for bartonellae, and following experimental infections. Accidental infection of individu­als that do not belong to reservoir host species appears far more likely to provoke pathology; B. henselae infections in humans most commonly present as cat scratch disease, a syndrome characterized by lymphadenopathy.

The nature and scale of immune response to Bartonella infection by a reservoir host species is not fully under­stood. Hosts do mount a humoral response, but the effectiveness of this in curtailing infection is uncertain. Although some experimental work has suggested a key role for antibodies in the cessation of bacteraemia, natural infections appear to continue when high titres of circulat­ing antibodies are detected; a capability that may, in part, be the result of occupation of erythrocytes, a site of com­plete immune privilege.

Clinical signs and treatment

There are no reports of morbidity or mortality among wildlife species that are clearly linked to Bartonella infec­tion. Indeed, given the high prevalence of infection in apparently healthy reservoir hosts, establishing a link between bartonellae and a specific clinical manifestation is very difficult. Nonetheless, some evidence that Bartonella infections can provoke disease in a minority of individuals among a reservoir host population has emerged from studies on B. henselae, which parasitizes cats; the presence of B. henselae DNA has been demonstrated in relevant tissues of animals with uveitis and endocarditis, and epi­demiological surveys have suggested that seropositive cats are more likely to have kidney disease and urinary tract infections, stomatitis and lymphadenopathy.

DIAGNOSIS

The detection of bartonella infections in reservoir host species is relatively straightforward. The examination of Giemsa or Romanowski-stained blood smears under 1000? magnification reveals characteristic intra-erythrocytic rod­shaped bacteria (Figure 37.2). However, this approach is relatively insensitive, as infection intensity can be low (assays have been described, targeting various genetic loci and with either genus- or species­specificity. The identification of bartonellae to species level is usually achieved by comparative sequence analysis of PCR amplification products; again, various genetic loci have been used for this. Bartonella species are generally inert to most routine biochemical tests, so the differentia­tion of species using phenotypic methods is not possible. However, serological methods, using species-specific poly­clonal or monoclonal antibodies, do offer an alternative approach when PCR is not possible. Testing for antibodies, most often using an immunofluorescence test format is also used in epidemiological studies, although use of this approach for some species of wildlife is compromised by an absence of appropriate conjugates.

public health concern

Bartonella species are established zoonotic pathogens; the feline-associated species B. henselae is associated with an increasing spectrum of clinical syndromes in humans including, most commonly, cat scratch disease, but also endocarditis, ocular syndromes and vascular proliferative diseases such as bacillary angiomatosis and peliosis hepatis. Some of these presentations have also been encountered in companion animals. Reports of human disease caused by other Bartonella species have also been published, but these remain rare; it seems likely that the relative medical importance of specific Bartonella species is tempered by the frequency of contact between humans and its particular reservoir host species. The wildlife- associated species implicated in human disease include B. grahamii, which has been associated with neuroretinitis, B. alsatica, which exploits rabbits as its reservoir host, and has been associated with lymphadenitis and endocar­ditis, and B. elizabethae, which naturally parasitizes rats, and also been associated with endocarditis. Infection may occur via the bite of an infected arthropod, but, a more common route of transmission to humans appears to be via the scratch of a claw contaminated by infected arthro­pod faeces.

SIGNIFICANCE AND IMPLICATIONS IN ANIMAL HEALTH

Bartonella species are also associated with disease in live­stock and companion animals. For example, the deer- associated species B. Schoenbuchensis was isolated from a French cow with no clinical signs, and B. bovis, for which cattle are the reservoirs, has been implicated as a cause of endocarditis in cattle (again suggesting that serving as a reservoir host does not come without a cost). B. henselae is increasingly being recognized as a pathogen of dogs.

MANAGEMENT, CONTROL AND REGULATIONS

Control of infections is difficult, but is mediated primarily through the control of ectoparasite fauna. In domesticated animals at least, the regular application of insecticides/ acaricides is considered the most practical means for infec­tion control. Currently there are no national or interna­tional requirements to report Bartonella infections in Europe, and the diagnosis of Bartonella infections in live­stock does not currently have any consequences for their management.

References for further reading are shown at the end of this chapter¹-^15-17).