Background
Severe malarial anaemia (SMA) is a major factor contributing to malarial morbidity in humans (for review see [
1,
2]) and is a significant pathological feature of most rodent malaria infections [
3]. During malaria infection, destruction of red blood cells (RBCs) from parasite replication is likely to contribute to the observed anaemia. However there also appears to be an interruption in the normal flow of replacement RBCs, as well as the appearance of abnormal RBCs (dyserythropoiesis) and premature destruction of uninfected RBCs (erythrophagocytosis) [
4‐
6] that will alter the total number and relative effectiveness of RBCs in the circulation (reviewed by [
7]). Although it is known that the severity of anaemia in malaria infection can be discordant with parasite replication rates, the relative contributions of RBC destruction through parasite replication and the altered replacement and persistence of functional RBCs in the circulation in SMA has not been formally assessed.
Similarly to human SMA, dyserythropoiesis and erythrophagocytosis have been observed in mouse malaria infection [
8‐
10]. The pro-inflammatory response of the host to the malaria parasite observed in both human infections and mouse models [
11‐
15] is thought to play a role in these processes. Tumour necrosis factor-α (TNF-α) [
16‐
19] and macrophage migration inhibitory factor (MIF) produced by immune cells, in particular macrophages [
20,
21], are thought to contribute to malarial anaemia by suppressing erythroid development. Conversely, anti-inflammatory cytokines such as IL-10 may play a protective role [
12,
22‐
24]. Therefore, it is possible that under the conditions of natural malaria infections, where differences between hosts in the magnitude or skew of the immune response to malaria parasites can be observed, the contribution of the different factors determining the severity of malarial anaemia may also be altered.
The extent to which the number of parasitized RBCs (pRBCs) in the circulation correlates with the severity of malarial anaemia, or whether even in the early phase of a blood stage infection there is evidence of involvement of the host response in RBC destruction or suppression of RBC synthesis, has been examined here using blood stage infections of a mouse model of malaria, Plasmodium chabaudi chabaudi, which replicates in a synchronous manner approximately every 24 hours. Circulating parasite numbers have been compared with haemoglobin (Hb) levels and loss of RBC during P. chabaudi infections in mice. Using RAG2-/- mice lacking T and B cells, and two P. chabaudi clones (AS and CB) of differing virulence, the data indicate that the level of anaemia does not correlate with the number of parasite-infected RBC. Furthermore, the greater loss of RBC in the more virulent infection of P. chabaudi (clone CB) is not the result of a greater parasite load, but rather a decrease in the number of uninfected RBC. Therefore, an increase in the severity of anaemia in P. chabaudi infections of BALB/c mice appears to occur via mechanisms that are independent of parasite replication.
Discussion
The data from the experiments in this study indicate that the acute severe anaemia accompanying a blood stage infection with P. chabaudi in mice is not primarily the result of direct parasite destruction of infected RBCs. In particular, during infections with two clones of P. chabaudi that differ in virulence, the drop in circulating RBC numbers was clearly not associated with the total number of circulating parasites, but rather with the percentage of infected RBCs suggesting that it is availability of new RBC that is limiting in this infection model. As this association also occurred in RAG2-/-immunodeficient mice, the mechanism governing availability of RBCs do not seem to require an adaptive immune response.
The reduced number of uninfected RBCs in acute
P. chabaudi infections could be the result of the host response to the infection, which leads to impaired replacement of RBC (erythropoiesis), the premature destruction of uninfected RBCs (erythrophagocytosis) or results in the production of defective RBCs (dyserythropoiesis) which would be removed by the reticular endothelial system [
7‐
10,
30]. In
P. chabaudi (AS) infections, impaired erythropoiesis has been shown to be due, in part, to the reduced ability of RBC precursors to respond to erythropoietin (EPO), resulting in limited replacement of RBCs and thus exacerbation of anaemia observed during infection [
31]. An increase in the removal of either infected RBC or erythrophagocytosis via opsonizing immunoglobulin cannot wholly explain RBC loss described in this study, as a similar loss of RBC also occurs in
P. chabaudi RAG2-/- mice, in the absence of an acquired immune system. However, it is still possible that complement components activated via the alternate pathway could facilitate erythrophagocytosis by allowing the recognition of erythrocytes via macrophage complement receptors [
32]. Such a mechanism may normally facilitate steady state haemostasis, but may be enhanced under the immune conditions of a malaria infection.
Infection with the more virulent CB clone of
P. chabaudi resulted in a greater reduction in RBC number than the less virulent clone AS, despite comparable numbers of parasites (Figure
2c) and invasion rates (data not shown). Across the AS and CB-infected BALB/c mice, pRBC numbers peaked between 1.4 -4 × 10
9 pRBCs with a concomitant loss of 7.4 – 13.6 × 10
9 RBCs in circulation (Table
1). Therefore, anaemia in
P. chabaudi infections cannot be entirely due to haemolysis of RBCs due to parasite replication. Since
P. chabaudi can sequester in the liver [
33], it is possible that the circulatory pRBCs are not entirely representative of parasite population numbers. However, this was controlled for by measuring numbers of pRBCs at 3 pm in the normal light cycle, a time when the sequestering schizont stage of
P. chabaudi is not present in this synchronous malarial infection [
34,
35]. Alternatively, it is possible that selective invasion of
P. chabaudi for young normocytes and reticulocytes may erode the replenishment of circulating RBCs. However current evidence suggests that
P. chabaudi only invades younger cells when more mature normocytes are in short supply [
36]. Additionally, if this explained the clonal differences in the severity of anaemia, we should concomitantly see differences in parasite load between the clones, with an extended parasitaemia in one of the clones when normocytes are in short supply. It seems more likely that it is the differential induction of the host response by the CB clone that is responsible for the greater degree of anaemia observed in this infection.
Rodent malaria parasites can activate mouse macrophages [
37,
38] to produce the cytokines IL-12 [
37], TNF-α and MIF [
20] probably via recognition of parasite ligands by pattern-recognition receptors (PRRs) [
39]. These cytokines have all been implicated in the immunopathogenesis of anaemia in mouse malaria infections [
3]. It is thus possible that the response of innate cells such as macrophages, dendritic cells or NK cells could contribute to the host response and therefore to anaemia. It would therefore be of great interest to determine whether impaired erythropoiesis or erythrophagocytosis in CB-infected mice is greater than that measured in AS-infected animals. Since the peak percentage of parasitaemia occurred earlier in CB than in AS infections it might be that an impaired response to EPO by erythroid progenitor cells is detectable at an earlier time point in CB-infected animals. A detailed examination of the nature of the host response to the different
P. chabaudi clones would clarify whether cells of the innate system play a role in the development of anaemia.
The discrepancy between the total number of circulating parasites and the percentage parasitaemia in the clonal infections described here highlights the need for caution in the interpretation of percentage of infected RBC as a measure of parasitaemia. Percentage parasitaemia is confounded by the level of anaemia and thus can be misleading as a determinant of parasite numbers. It appears that the cause of the differences in anaemia between the two clones is due to differences in the number of uninfected RBCs in circulation in CB-infected and AS-infected mice.
An interesting finding of this study was that a reduction in the number of circulating RBC did not necessarily correlate with a drop in the levels of Hb in the blood. Infected RAG2-/- mice had lower circulating numbers of RBCs than infected BALB/c, yet displayed similar levels of Hb. Furthermore, there was no correlation between the drop in RBC numbers and Hb in mice infected with either of the two
P. chabaudi clones. This finding may be related to different levels of circulating reticulocytes between RAG2/- and BALB/c animals, and between BALB/c animals infected with clone AS or CB. Reticulocytes are elevated in
P. chabaudi AS infection as a response to the drop in circulating RBC number [
36]. Although we did not measure circulating reticulocyte numbers in the infections in this study, we would predict that both RAG2-/- AS-infected animals, and BALB/c CB infected mice would have a greater impairment of haematopoiesis and thus a lesser number of reticulocytes in the circulation than AS-infected BALB/c mice. If reticulocytes have reduced haemoglobin levels until they fully mature into RBCs, then it follows that the AS-infected animals will have lower circulating Hb levels relative to the number of circulating RBCs compared with AS-infected RAG2-/- or CB-infected BALB/c mice. Regardless, these results emphasize that any investigation of malarial anaemia should include more than one measurement of anaemia.
Competing interests
The authors declare that they have no competing interests
Authors' contributions
The data was collected and analysed by TL. Both authors contributed to the experimental design, interpretation of the data and writing of the manuscript. Both authors read and approved the final manuscript.