Kinetics of the inflammatory response during experimental Babesia rossi infection of beagle dogs

Babesia rossi causes severe morbidity and mortality in dogs in sub-Saharan Africa, and the complications associated with this disease are likely caused by an unfocused, excessive inflammatory response. During this experimental B. rossi study we investigated inflammatory marker and cytokine kinetics during infection and after treatment. We aimed to determine whether infectious dose and treatment would influence the progression of the inflammatory response and clinical disease. Five healthy male beagle dogs were infected with B. rossi, three with a high infectious dose (HD group) and two with a low infectious dose (LD group). Clinical examination, complete blood count (CBC) and C-reactive protein (CRP) were determined daily. Cytokines were quantified on stored plasma collected during the study, using a canine specific cytokine magnetic bead panel (Milliplex©). The experiment was terminated when predetermined endpoints were reached. Parasitemia occurred on day 1 and 3 in the HD group and LD group respectively. The rate of increase in parasitemia in the HD group was significantly faster than that seen in the LD group. Significant differences were found in heart rate, blood pressure, interferon gamma (INFγ), keratinocyte chemoattractant (KC), INFγ-induced protein 10 (IP10), granulocyte-macrophage colony-stimulating factor (GM-CSF), monocyte chemoattractant protein 1 (MCP1), tumor necrosis factor alpha (TNFα), interleukin 2 (IL-2), IL-6, IL-7, IL-8, IL-10 IL-15, IL-18, CRP, neutrophils and monocytes between groups at multiple time points during the course of the infection. Our findings suggest that the initiation of inflammation occurs before the onset of clinical disease in B. rossi infection and infectious dose influences the onset of the inflammatory response. Treatment not only fails to curb the inflammatory response but may enhance it. Finally, we found that there is an imbalance in pro/anti-inflammatory cytokine concentrations during infection which may promote parasite replication.


Introduction 49
Babesia rossi, a virulent Babesia species, causes a severe form of babesiosis in the domestic dog 50 associated with a high rate of morbidity and mortality (1-3). Babesiosis is a complex multi-systemic 51 disease that can be classified as either uncomplicated or complicated (2,4,5). Complicated babesiosis 52 occurs when the pathology noted cannot be attributed purely to the anemia or when the anemia 53 becomes severe enough to perpetuate organ dysfunction (2). B. rossi infection, like Plasmodium 54 falciparum malaria in humans, results in a protozoal sepsis with a severe systemic inflammatory 55 response (6-8). The concept of a 'cytokine storm' is well established in human inflammatory and 56 infectious conditions, such as malaria and sepsis (9). This theory proposes that systemic illness and 57 the course of disease is not solely caused by the microbes themselves but is also the result an 58 unbalanced cytokine response to microbe antigens (10). The disease course seen in B. rossi infections 59 bears a striking resemblance to that seen in falciparum malaria, leading one to hypothesize that a 60 similar 'cytokine storm' may be an essential mechanism in the pathogenesis of this disease (11,12). It 61 is clear that B. rossi initiates a marked inflammatory response characterized by increased circulating 62 markers such as C-reactive protein (CRP) and cytokines including monocyte-chemotactic protein-1 63 (MCP-1), interleukin (IL)-2, IL-6, IL-10, IL-18 and tumor necrosis factor alpha (TNFα) (13-15).

64
Complications associated with this disease are likely the result of an unbalanced inflammatory

177
The demographic characteristics of the experimental group of dogs were as follows: All dogs were 6-178 month-old sterilized male beagle dogs. All 6 dogs tested negative for Babesia, Ehrlichia, Anaplasma

179
and Theileria based on PCR-RLB done prior to the initiation of the experimental study. No significant 180 difference was noted between the LD group and HD group for baseline data for any variable. The HD group had a detectable parasitemia 48 hours earlier than the LD group (Fig 1). There was a For the HD group, increases in CRP concentrations (Fig 3)

264
Like changes seen in MCP-1, one dog in the LD group had increased IL-6 at 192 hours but this did 265 not reach significance. Interleukin-6 and MCP-1 were strongly correlated (r = 0.792, p < 0.001). The

266
HD group had significantly reduced IL-8 concentrations (Fig 9) compared to baseline at 72 hours (p = The next category of cytokines, GM-CSF (Fig 11), TNFα (Fig 12), IL-2 ( Fig 13) and IL-7, had very 290 similar patterns of change and were all markedly increased after treatment in the HD group, 291 particularly in one dog (Table 1). Significant increases in these cytokines were seen after treatment, at    was seen in the HD group after treatment and band neutrophilia in B. rossi infections has been 405 associated with lower hematocrits and blood transfusions, consistent with findings in our study (30).

406
A band neutrophil count of > 0.5 x 10 9 /L at presentation carries an odds ratio for death of 5.9 (2).

407
Interestingly the only dog with a band neutrophil count above this level prior to treatment in our study 408 was the dog that died. In one study there was a higher proportion of dogs with a neutrophilia in the

495
In previous studies on the cytokine changes in B. rossi infections, IL-8 was decreased at presentation 496 when compared to healthy control dogs (13,14). In contrast to these findings, IL-8 increases in B. canis 497 infections and even showed a progressive rise for at least 7 days after treatment (16). Our study 498 demonstrated decreased IL-8 concentrations during the early stages of infection in the HD group 499 followed by a considerable increase shortly before treatment, when parasitemia was very high.

511
Interleukin-10 is essential in the modulation of the inflammatory response and plays a key role in 512 preventing excessive inflammation as well as promoting the resolution of inflammation once the 513 inciting pathogen has been eliminated (54). Although the anti-inflammatory effects of IL-10 are critical, 514 excessive or inappropriately timed production of IL-10 may prevent an effective immune response to 515 an organism, allowing persistence or even unregulated replication in the host (54). This has been seen (56). It is also a potent stimulator of the production of other pro-inflammatory cytokines such as IL-1β,

531
IL-6 and IL-8, serving as co-ordinator in the inflammatory response (46). Increased concentrations 532 were found in natural B. rossi infections and higher levels were associated with increased risk of 533 complicated disease and death (14). Concentrations of TNFα were only increased after treatment in this 534 study and reached very high levels in one dog in the HD group. There was also a strong positive 535 correlation with IL-6, GM-CSF, IL-2 and IL-7. The excessive production of TNFα following treatment may be indicative of a dysregulated immune response. Both IL-2 and IL-7 act on lymphocytes and were 537 positively correlated with each other in this study (57,58). Interleukin-2 did not increase when B. rossi 538 infected dogs were compared to healthy control dogs in one study but higher concentrations were noted 539 in infected dogs presented within 48 hours of clinical illness (13,14). Increases in IL-7 has not been 540 identified in previous studies of B. rossi infections (14). Both cytokines only displayed significant 541 increases after treatment in the current study. Previously, significant reduction in T-helper lymphocytes 542 and cytotoxic T-lymphocytes were identified in complicated B. rossi infections and decreased 543 concentrations of cytotoxic T-lymphocytes was also noted in uncomplicated cases at presentation (28).

544
In that study it was hypothesised that a state of functional immune suppression may be present, and this 545 is supported by the apparent deficiency of cytokines involved in lymphocyte proliferation and activation 546 identified in our study prior to treatment (28). Treatment and subsequent lysis of the parasites may have 547 interrupted the immunosuppressive state and the release of soluble antigens was able to stimulate the 548 adaptive immune response triggering production of these cytokines.  (13). In 554 the current study IP-10 concentrations were mildly increased in both groups.

556
The main limitation of this study was the small sample size. Six dogs we were used in the study, with 557 only 5 dogs being inoculated. Every attempt was made to exclude any confounding or influencing 558 factors. All dogs were the same age, sex and breed with identical vaccination and deworming 559 protocols. They were housed in the same isolation housing and outdoor facilities. Diet, training, 560 experimental procedures, sample collection and human interaction was consistent between all dogs.

561
Due to the small sample size, it is possible that significant differences between the groups may have 562 been missed (type 2 error). Our study has found that infectious dose influenced the onset and dynamics of the inflammatory 565 response. Most variables shared similar kinetic patterns between groups, differing with respect to the 566 timing of the onset of disease only. If the infection in the LD group had been allowed to progress, it is 567 likely these variables would have reached similar degrees of change to those seen HD group. There