Phenotypic characteristics distinguishing bloodstream form and central nervous form of Trypanosoma brucei rhodesiense.

Background: Phenotypic and morphological characteristics distinguishing from bloodstream form (BSF) and central nervous system (CNS) of Trypanosoma brucei rhodesiense ( Tbr ) are poorly understood. Method: To identify these distinguishing characteristics, we separately infected four donor mice with each of five Tbr isolates (KETRI 2537/3537/2656/3459 and EATRO 2291). At 21 days post infection (DPI), donor mice were euthanized, BSF or CNS derived trypanosomes recovered and used for the following studies: 1) determination of morphological characteristics 2) pathogenicity studies using groups of 10 mice per isolate form. We then assessed differences in their lengths and morphology) and other characteristics including pre-patent period (PP), parasitaemia progression, packed cell volume (PCV), body weight, survival times, gross pathology, and histopathology. All analyses of data were conducted using GenStat, UK where p ⩽ 0.05 were considered statistically significant. Differences between and within the means were analyzed using one-way ANOVA. General Linear Model was used to analyze data on the length of the trypanosome. Survival data analysis was carried out employing the Kaplan–Meier method. Results: Morphologically, the CNS forms were predominantly long slender (LS) while BSF forms consisted of a mixture of short stumpy and intermediate forms. The mean length of CNS trypanosomes was 0.6 micrometer longer than their counterpart BSF derived trypanosomes. The PP was significantly (p<0.05) shorter and progression to peak parasitaemia faster (7 vs 9 days) in CNS than BSF derived trypanosomes. PCV declined by 21.6% and 26.9% in BSF and CNS infected mice respectively whereas non-infected control increased by 3.8% at 14 DPI. Body weight changes in BSF and CNS infected mice were (12.7% and 9.2% respectively) and significantly (p <0.05) lower than in non-infected control (27.6%) at 14 DPI. Gross pathology changes (splenomegaly and hepatomegaly) and histopathology changes were pronounced in mice infected with CNS relative to BSF trypanosome forms. Changes in histopathology included congestion, infiltration with inflammatory cells, hemolysis and necrosis, all indicators of differential virulence of the forms. Conclusion: Our study identified higher pathogenicity in CNS relative to BFS derived trypanosome forms in the mouse model. We also identified KETRI 2656 as a suitable isolate for acute menigo- encephalitic studies.


Introduction
Human African Trypanosomiasis (HAT, sleeping sickness) is caused by Trypanosoma brucei rhodesiense (Tbr) and T. b. gambiense (Tbg) species of trypanosomes which are transmitted by tsetse flies of Glossina species [1]. The Tbr is predominant in eastern and southern Africa and causes the acute form while Tbg is predominant in Central and Western Africa and is responsible for the chronic form of HAT [2]. Recognized stages in the clinical presentation of HAT are the hemolymphatic (early) and encephalitic (late) stages that typically manifest as blood Stream form (BSF) and central nervous system, (CNS) forms respectively. The early stage infection is clinically non-specific, manifesting as malaise, headache, arthralgia, generalized weakness, weight loss and anaemia [3]. On the other hand, late stage infection, occasioned by the parasites crossing the blood-brain barrier or blood-cerebrospinal fluid (CSF) barrier and invading the CNS [4,5] clinically manifest as psychiatric, motor, sensory and sleep abnormalities [6].The CNS invasion may be aided by parasite and/or host derived factors [5].
In the CNS, the parasites DNA can be detected as early as six days post infection (DPI) in their replicative slender forms [7,8], with peak infection at 21 DPI [8]. With progression of CNS HAT, changes occur in the cerebrospinal fluid characterized by presence of cytotoxic compounds and reduced cerebrospinal fluid as well glucose levels [9,10], making it more 'hostile' to the trypanosomes [11,12]. This is further contributed by the CNS immune response [13]. Phenotypic effects of these factors on CNS recovered/derived-trypanosomes are poorly understood. To investigate effect of these factors on trypanosomes, we hypothesized that there are distinct morphologic and phenotypic differences between the BSF and CNS forms of T.b. rhodesiense trypanosomes that differentially influence the pathogenicity of either form. We thus compared morphologic and phenotypic changes associated with BSF and CNS derived trypanosomes directly harvested from the blood and the brain, respectively, and evaluated their pathogenicity.

Selection Of Trypanosomes Isolates
Five Tbr isolates were used in this study (Table 1). They were selected from BioRI (formerly KETRI) cryo-bank and were originally isolated from HAT patients in the three east African countries [14].The isolates had gone between1-8 eight passages in mice. Old and recent isolates were selected for this study. acclimatized the mice to experimental room conditions for two weeks and dewormed them using Noromectin®, Norbrook, UK at 0.2 mg/kg as described by [17] to ensure that they were free of ecto-and endo-parasites. We obtained during this acclimatization period base line data on packed cell volume (PCV) and body weight prior to infection.

Experimental Design
Four (4) donor mice per isolate were immunosuppressed using cyclophosphamide 300 mg/kg for three consecutive days [18] after which the recovered cryopreserved T.b.rhodesiense isolates shown in Table 1were injected for multiplication.The mice were monitored for parasitaemia for 21 days post infection, a period known to be sufficient for development of late stage disease in the mouse model [19]. They were then euthanized using concentrated carbon dioxide (CO 2 ). At least one mL of heart blood trypanosomes were collected in 5 µl of 10% EDTA enough to prevent coagulation and pooled from the four donor's mice while brain tissues from the four donor mice were also pooled and suspended in cold PSG pH 8.0. Brain tissues were washed in 3 Ml PSG pH 8.0 buffer per wash for at least ten times (total 30Ml PSG buffer) and presence of trypanosomes in every wash carefully examined under the microscope. When no trypanosome could be detected, the final buffer wash was discarded, the brain excised using sharp pair of scissors and gently homogenized in PSG pH 8.0. The supernatant containing brain trypanosomes (CNS) from the four donor mice was pooled. Slide smears of CNS trypanosomes of each of the isolates were made from the brain homogenates. Similarly, smears of BSF trypanosomes from each of the Tbr isolates were made from pooled heart blood. The density of trypanosomes in the pooled supernatant collected from the brain homogenates (Table 3) was then quantified using a haemocytometer. Similarly, the density of the BSF trypanosomes in the pooled heart blood ( Table 3) were also quantified and adjusted to the inoculum level of the supernatant containing CNS form trypanosomes using PSG pH 8.0 (Table 3). Either BSF or CNS form of trypanosomes (BSF and CNS) were then intraperitoneally injected into ten experimental mice while five non-infected mice acted as control. The infected mice were monitored for the following parameters: pre-patent period (PP), parasitaemia progression, PCV and body weight once a week, survival, gross pathology, and histopathology. At 30 days post infection, four mice from each group including two non-infected control were sacrificed for gross lesions (lesions and organ weights) and histopathology. The remaining mice were humanly euthanized and disposed.
Pre-patent period (pp) and parasitaemia progression Blood for estimation of parasitaemia levels was collected daily from each mouse using the tail tip amputation method [20]. The pre-patent period and parasitaemia progression were determined using the rapid matching method [21]. The infected mice were monitored for 30 DPI. The end point of the infected mice was determined by observation of clinical signs such as lethargy and hackle hair, as well as PCV drop of approximately 25% with consistent high parasitaemia levels of 1 × 10 9 /mL − 1 for at least three consecutive days.
The animals were sacrificed immediately by CO 2 asphyxiation in accordance with guidelines of the Institutional Animal Care and Use Committee (IUCAC) as described by [22] and recorded as dead animals.

Determination Of Morphology Of Trypanosomes
A drop of blood from mice infected with BSF or CNS derived trypanosomes was collected as infection progressed. Thin smears were then prepared and examined using Leica Packed cell volume (PCV) and body weight These were measured once a week as outlined by Naessens [24] for PCV and by Gitonga [25] for body weight.

Survival
All control mice survived up to the end of the experimental period of 30 days and their survival time data were therefore categorised as censored. The shortest survival time was observed in mice infected with EATRO 2291 at 13.7 ± 2.0 and 12.7 ± 1.5 for BSF or CNS infected mice, respectively ( Fig. 7  weight of each mouse was determined after which the mouse was euthanized, dissected and the brain, spleen, kidneys, liver, lungs and heart collected and weighed using a weighing balance (Mettler Toledo PB 302 ®, Switzerland). These organs were preserved in 10% formalin and thereafter processed for histopathology as previously described [27]..

Statistical analysis
Analysis was done to test the significant differences between the BSF and CNS forms per isolates in PP, parasitaemia progression, PCV, body weights, trypanosomes length and survival in the two groups. The data obtained from the study were summarized as means ± standard error of mean, while the differences between and within the means were analyzed using one-way ANOVA. The analyses were conducted using GenStat, UK where p ⩽ 0.05 were considered statistically significant. General Linear Model in SAS Release 8.02 was used to analyze data on the length of the trypanosome. Differences between any two means were considered significant at p < 0.05. Survival data analysis was carried out employing the Kaplan-Meier method on StatView (SAS Institute, Version 5.0.1) statistical package for determination of survival distribution function. Rank tests of homogeneity were used to determine the effect on host survival time of BSF-and CNS-infected mice [29].

Molecular identification of cryo-bank isolates
The 460 bp SRA gene fragment was amplified confirming all the isolates as T. b.
rhodesiense isolates. Our PCR amplification of the 460 bp bands SRA gene (Gibson et al, 2002) in all our isolates, confirmed their Tbr species status (Fig. 1).  (Fig. 2). This difference was also reflected in the overall mean ± SE pre-patent period, which was 5.2 ± 0.3 and 4.7 ± 0.2 for all the mice that were infected with the BSF and CNS-derived trypanosome trypanosomes, respectively (p < 0.05)

Progression Of Parasitaemia In Mice
Parasitaemia in mice infected with either the BSF or CNS derived trypanosomes increased rapidly attaining on average first peak parasitaemia of 6.1 × 10 8 and 6.8 × 10 8 trypanosomes/mL(anti-log 8.8) of blood for both BSF and CNS derived trypanosomes (Table 2) showing that in general, the parasitaemia patterns were similar. However, the first peak parasitaemia was attained after an average of 7 and 8 days for CNS and BSF trypanosome forms respectively ( Table 2), showing that at the initial stages of the infections, the parasitaemia increase in mice infected with CNS derived trypanosomes was consistently faster than those infected with BSF trypanosomes (p < 0.05). A noticeable decline in the first peak parasitaemia was observed in a majority of the isolates (Fig. 3). In general, however, the parasitaemia induced by CNS or BSF trypanosome forms remained high with minor fluctuations throughout the infection (Fig. 3).  (Fig. 4(iii)) while BSF trypanosomes were a mixture of short stumpy and intermediate form trypanosomes (Fig. 4).However, when CNS parasitaemia density of all the isolates were compared, KETRI 2656 CNS was relatively higher than the rest (Fig. 4 (iii)). The density of trypanosomes in pooled heart blood was twice greater than trypanosomes in brain supernatant (Table 3). On the stained giemsa slides, there were 2-3 more trypanosomes per field in BSF as compared to corresponding CNS trypanosomes from the same isolate (Fig. 4). Mice infected with the two forms of trypanosomes recorded a significant (p < 0.001) decline in PCV within the first peak parasitaemia when compared with the pre-infection data ( Fig. 5 (i) Mice infected with BSF and CNS derived trypanosomes recorded an increase in body weight as infection progressed, an increase that was however, slower than the increase observed in the un-infected control mice (Fig. 6). The mean ± SE body weight at day 0 ranged between 22.4 ± 0.4 (CNS) to 32.5 ± 1.1 (BSF). At 14 DPI, KETRI 2656 BSF infected mice showed the highest increase from 24.3 ± 1.1 (day 0) to 29.5 ± 0.6, equivalent to 21% increase ( Fig. 6 (iii)). However, the increase in the non-infected control over the same period was from 21.7 ± 2.2 at day 0 to 27.7 ± 0.8 equivalent to 28% increase compared to the highest increase recorded by KETRI 2656 (21%) infected mice. Thereafter, all the infected mice continued increasing in body weight with the increase ranging between 7% (EATRO 2291 and KETRI 3459 (CNS) (Fig. 6,(iv &v)) and 24% (KETRI 3537 (BSF)) ( Fig. 6, (ii)). When compared as groups of infected mice, the increase in body weight was significantly lower (p < 0.01) in CNS than in BSF infected mice (

Histopathology
For the BSF group, vascular congestion of the heart, lungs, and spleen and were observed.
In addition, the lungs from the BSF groups showed marked alveoli wall thickening and Infiltration with lymphocytes ( Fig. 9 (A and B)). Myocardium from both the CNS and BSF groups had focal areas of lymphocytic infiltration ( Fig. 9G and H)). Focal areas of lymphocytic infiltration and tubular epithelium necrosis were also observed in the kidney tissues from both the CNS and BSF groups ( Fig. 9I and J)). The brain tissues from CNS group showed Infiltration of perivascular areas with lymphocytes, and blood vessels congestion while those from the BSF were all normal ( Fig. 9E and F)). Furthermore, marked vascular congestion and infiltration of perivascular areas with lymphocytes were observed in the liver tissues from both BSF andCNS groups ( Fig. 9C and D)). In general, the leisons were more pronounced in CNS than BSF infected mice (supplementary Table 5)

Discussion
The current study compared the phenotypic characteristic of the CNS derived trypanosomes in comparison with their BSF trypanosomes predecessor in mice using five isolates which were confirmed to be T.b.r. by the presence of the SRA gene as previously described [15]. The main finding of this study was that the phenotypic characteristics were isolate independent although in general, CNS derived trypanosomes were significantly more pathogenic than the BSF trypanosomes.
The pre-patent period was significantly shorter in CNS than in BSF derived trypanosomes.
The parasitaemia of CNS-derived trypanosomes also increased at a consistently higher rate, leading to attainment of the first parasitaemia peak within 7 days as compared to 8 days for BSF-derived trypanosomes, indicating a higher virulence of the CNS derived trypanosomes. According to previous studies, isolates with shorter pp and high parasitaemia are considered virulent [30]. However, after the first peak, there was no difference in the parasitaemia, progression suggesting the forms responded similarly to the humoral immune response. The difference observed during the first parasitaemia peak may be attributed to the morphology of the infecting trypanosomes. At 21 DPI, BSF trypanosomes were represented by intermediate and short stumpy (SS) forms, which are non-proliferating and adapted to tsetse infection [31,32]. In contrast, CNS derived trypanosomes were long and slender as was also reported by Wolburg, [33]. Long-slender (LS) forms have also previously been reported to proliferate actively [32]. With exception of mice infected with KETRI 2656 whose CNS derived trypanosomes parasitaemia density was relatively high when compared with the other isolates, Fig. (4(iii)), our study observed low density of the CNS derived trypanosomes and was in agreement with previous observation [12]. However, this observation was made at 21DPI from the brain of mice initially passaged with BSF derived trypanosomes. Our study however did not quantify at 21 DPI the density of CNS trypanosomes from the brain of mice passaged with CNS derived trypanosomes to determine whether there is any density difference between the two sets of CNS derived trypanosomes (CNS trypanosomes derived from the BSF forms and CNS trypanosomes derived from the CNS forms of trypanosomes).
Results on the PCV showed that mice infected with BSF or CNS derived trypanosomes recorded significant decline confirming previous observations associating anaemia with human infection [34] as well as with experimental human trypanosomiasis using vervet monkeys and mice [35,36]. However, data generated from our study indicated that the decline was more significant in mice infected with CNS than in BSF derived trypanosomes.
This difference could be associated with the high parasitaemia registered by the CNS infected mice during the first peak parasitaemia resulting to a high extracellular destruction of erythrocytes as previously reported [37]. In their study, these authors observed increased red cell destruction in the spleen from the third day of patent parasitaemia. In another study on mechanisms of controlling anemia [38], data showed rapid development of anemia until the peak of the first wave of parasitaemia. Importantly, in the current study, after first peak, parasitaemia remained relatively high with no difference in parasitaemia progression between the BSF and CNS derived trypanosomes groups in the remaining days of infection. During this period, the decline in PCV recorded stability ( Fig. 6 (vi)) characteristic of chronic phase anaemia [36].
Our data on body weight changes showed increase in body weight both in infected and non-infected control with days post infection irrespective of isolate form although the increase in the non-infected control was significantly more. However the increase observed with infected mice was in agreement with previous studies with T. congolense or T. brucei trypanosomes [25,38] and T. evansi trypanosomes [ 39] infections. Observations by these authors and by us contradicts previous observations showing relationship between decline in body weight and trypanosomiasis infections using ewes and mice laboratory animals [40,41] suggesting body weight alone as an inappropriate biomarker of pathogenicity. Indeed our results showed a tremendous body weight increase in mice infected with KETRI 2656 which based on its shortest survival time may be classified as very virulent [42]. The increase according to our data was relatively more in mice infected with BSF than CNS derived trypanosomes. It is not clear as to why body weight should increase as infection progressed, but we partly attribute organomegaly particularly seen in the spleen and liver. Indeed, our results showed that at the end of experiment the spleen and liver organs of BSF and CNS infected mice were heavier when compared with the non-infected control. Our findings showed short survival times for mice infected with EATRO 2291 and KETRI 2656 demonstrating the two isolates as the most virulent among isolates used in this study [27,38].
On the morphological length of both the BSF and CNS derived trypanosomes, our results gave a range of 23 ± 1.7 to 31.1 ± 1.7 for the BSF trypanosomes and 23.0 ± 2.1 to 31.3 ± 09 for CNS derived trypanosomes which was in agreement with previous observation [43].
This was not significantly different although CNS derived trypanosomes measured relatively longer than the BSF trypanosomes. This suggests that the CNS derived trypanosomes maintained a higher density of long and slender forms. This increase in length was observed during the early days of mice infection with CNS derived trypanosomes and was maintained throughout the infection period. Whether this difference in length resulted in the difference observed in pathology between the two forms will require further investigations.
Splenomegaly and hepatomegaly have been reported in experimental human African trypanosomiasis using animals [44] and also in humans [45] as some of the symptoms associated with the haemolyphatic stage of infection [46,47]. Results from our study confirmed this observation in both BSF and CNS derived trypanosomes infected mice although these symptoms were more pronounced in the CNS than in BSF trypanosomes infected mice. In a previous study, splenomegaly was associated with the acute and post-acute phase of Trypanosoma lewisi infections of laboratory rats [48] and this may be attributed to the proliferation of Lymphocytes according Benoit [49]. Stijlemans [50], attributed hepatosplenomegaly to an enhanced extramedullary erythropoiesis occurring mainly in the spleen and to a lesser extent in the liver. The difference we observed in our study could be attributed to the severity of CNS over BSF infected mice. In a study by Muchiri [51], the authors observed pronounced splenomegaly and hepatomegaly in T.b.
rhodesiense isolates causing severe as compared to those causing less severe infection.
The present study recorded tissue inflammation in mice infected with the two forms and was in consistent with previous observations [27,51]. Trypanosome-induced inflammation lesions were more intense in CNS than in BSF derived trypanosomes infected mice which could be informed by the enhanced tissue invasion in CNS infected mice. Indeed, results from the present study showed higher density in the brain of mice infected with KETRI 2656 CNS trypanosomes.

Conclusion
The phenotypic characteristics of CNS derived-trypanosomes are poorly understood. To investigate these characteristics, we compared morphologic and phenotypic characteristics associated with BSF and CNS derived trypanosomes directly harvested from the blood and the brain, respectively, and evaluated their pathogenicity. our results have demonstrated that CNS derived trypanosomes are relatively more pathogenic than BSF forms when measured in terms of pre-patent period, parasitaemia development, PCV and organ pathology. It is possible that the parasitaemia proliferation within the first peak influences the pathological response considering that in this study, CNS trypanosomes proliferated at a higher rate than the BSF trypanosomes during the first peak parasitaemia. We further identified KETRI 2656 as a suitable isolate for acute menigoencephalitic studies. We recommend further studies to determine the factors influencing the observed variation in pathogenicity between the CNS and BSF forms.

Consent for publication
Not applicable

Availability of data and material
The datasets during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests

Funding
This study received funding from the Kenya Government through Kenya Agricultural and Livestock research organization (KALRO). All the authors except two, John Thuita and Paul Okumu are employee of KALRO. The Organization was involved in the design of the study, data collection, and analysis, interpretation of data and in writing the manuscript.

Authors Contribution;
KN Developed the proposal, conducted the study and prepared the draft manuscript.         Supplementary Files This is a list of supplementary files associated with the primary manuscript. Click to download.