Background
Despite large investments and global interventions over the last century, malaria remains a leading cause of morbidity and mortality in humans with 212 million clinical cases in 2015 and 429,000 deaths estimated worldwide, 70% of which are children under 5 year of age [
1]. The malaria infection can occur with fever and generalized sickness, and progress to severe anaemia or cerebral malaria, or can be asymptomatic. This range of clinical courses can be attributed to the host or to the parasites. In particular, the genetic background and the immune system of the host, the dynamics of transmission and the virulence of the parasites have to be considered [
2,
3]. Hepatic involvement in cases of malarial infection is a well-recognized entity. Malaria hepatocellular dysfunction is characterized by an increase in the serum levels of bilirubin and aminotransferases, exceeding three times the upper limit of normal values. Malarial hepatopathy is associated with a higher incidence of cerebral malaria, shock, acute respiratory distress syndrome (ARDS) and acute kidney injury [
4]. It occurs both in adult and paediatric patients, although it is more common in adults [
3,
4]. One suspected cause of malarial hepatopathy is related to the activation of liver macrophages that phagocytize haemozoin or parasitized red blood cells [
5]. Focal hepatocyte necrosis, cholestasis and granulomatous lesions have been documented in all malarial nodules [
4]. In malaria patients with jaundice, histopathological changes in the form of damaged hepatocytes, congestion of liver cells, haemozoin deposition, inflammatory infiltrates, and cholestasis have been demonstrated [
6,
7], as well as hyperplastic Kupffer cells [
8].
The molecular basis of different manifestations of malaria using two species of murine
Plasmodium, namely
Plasmodium berghei strain NK65, and
Plasmodium chabaudi strain AS have been recently investigated.
Plasmodium berghei NK65-infection results in lethal ARDS [
9], whereas
P. chabaudi AS-infection in C57Bl/6J is self-healing after an initial peak of parasitaemia and recrudescence, and does not induce lung pathology [
10,
11].
Plasmodium berghei NK65 and
P. chabaudi AS parasites produce different amounts of haemozoin (Hz), and it has been demonstrated that Hz is pathogenic in the lungs by inducing pulmonary inflammation [
12,
13]. Moreover, in mice infected by
P. berghei NK65, the development of ARDS is associated with biochemical modifications and lipid alterations of lung tissue and disruption of the molecular organization and lipid composition of alveolar surfactant [
12].
Excess Hz deposition was shown in the livers of
P. berghei NK65-infected mice compared to
P. chabaudi AS mice, while hepatomegaly due to inflammation and cell recruitment was higher in
P. chabaudi AS mice. Just as in the lungs, liver dysfunction with elevated serum ALT and AST levels and production of both pro- and anti-inflammatory cytokines were associated with the amount of haemozoin present in the livers of
P. chabaudi AS-infected mice. It was intriguing that inflammation, measured as cell recruitment and cytokine production, was much less pronounced in
P. berghei NK65 livers that had higher amounts of haemozoin than
P. chabaudi AS livers. Since haemozoin, due to its haem iron content, is strongly pro-oxidant and immunosuppressive [
14,
15] the present work aimed at investigating the extent of oxidative damage, the lipid alterations and the antioxidant defences present in the liver tissue of C57BL/6J mice infected with
P. berghei NK65, compared to mice infected with
P. chabaudi AS.
Discussion
The pathogenesis of hepatic dysfunction in malaria is complex and not completely understood. Histopathological findings show congested and swollen hepatocytes, Kupffer cell hyperplasia, deposition of brown malarial pigment, mononuclear cell infiltration, micro-occlusion by parasitized RBC and, less frequently, steatosis and spotty and submassive necrosis [
2,
7,
13,
16,
28]. Infection, inflammation and oxidative stress have been shown to induce multiple alterations in hepatic lipid and lipoprotein metabolism [
29]. The present study demonstrates that blood stage malaria infection by
P. chabaudi AS or
P. berghei NK65 exerts different effects on the liver of C57BL/6J mice. Hepatomegaly was confirmed in
P. chabaudi AS mice, as previously shown [
16] as well as liver dysfunction suggested by the serum increase of the hepatic enzymes AST and ALT, especially at day 10 post infection. The high AST levels lead to an AST/ALT ratio significantly higher in
P. berghei NK65 compared to
P. chabaudi AS. The AST/ALT index is used to discriminate the type of liver damage observed [
27,
30] and a progressive increase in the AST/ALT ratio correlated with a decrease in liver functions, like in cirrhosis and fibrosis [
31]. Since plasma clearance of AST is modulated by the activity of sinusoidal liver cells, during progressive fibrosis and cirrhosis, the functions of these cells are progressively impaired resulting in a relative increase in AST levels [
32]. In the case of malaria, in
P. berghei NK65-infected mice, an alternative hypothesis can be proposed. Elevated AST levels in absence of elevated ALT, in fact, may indicate haemolysis, since RBC contain AST. This has already been reported in sickle cell disease where the AST/ALT ratio has been used as haemolytic index [
33]. In the models used for this work, the AST/ALT ratio may well indicate initial fibrosis in both species, as confirmed by the OH-proline increase at day 8 post infection, and increased haemolysis in
P. berghei NK65-infected mice that develop higher parasitaemia than
P. chabaudi AS mice. The latter leads to higher levels of Hz in the liver of
P. berghei NK65- than in
P. chabaudi AS-infected mice [
16]. It is well known that Hz plays a crucial role in generating oxidative damage [
12,
14,
15]. The higher content of Hz, free (pro-oxidant) haem derived from intravascular haemolysis and high parasitaemia are likely to contribute to excessive oxidative stress response in
P. berghei NK65 livers. TNF may also participate. A marginal role seems to be played by the inflammatory infiltrate, which is much less abundant in
P. berghei NK65-infected livers than in
P. chabaudi AS mice, in which increased liver weight, histology and the correlation between hepatic inflammation, enzymes and haemozoin were demonstrated [
16].
The higher production of ROS in
P. berghei NK65-infected mice was confirmed by the elevated concentration of MDA, which is an index for the loss of structure and cell membrane integrity, and the concomitant depletion of the antioxidant defence system, such as CAT and total GSH. GSH constitutes the first line of defence against free radicals and is a critical determinant of tissue susceptibility to oxidative damage. The increase of SOD, shown also in other models of liver damage [
34], reveals the necessity of minimizing oxidative stress, which largely derives from the production of superoxide anion radicals. These changes seems to be associated with
P. berghei NK65 infection, since they were not present in
P. chabaudi AS-infected mice that showed only an increase of GR at both days post infection.
An important change observed in
P. berghei NK65 vs
P. chabaudi AS mice was the augmentation of the liver content of TG and cholesterol esters that indicates a potent stimulation of hepatic lipogenesis by
P. berghei parasites [
35]. This is in agreement with literature data showing that livers from
P. berghei-infected mice contain lipid droplets and myelin-like figures [
36]. A significant, although limited increase of TG was also present in
P. chabaudi AS infected mice in agreement to the small neutral lipid inclusions observed by Seixas et al. [
37]. Elevated TG are a characteristic of liver pathologies of different aetiology, clinically defined as non-alcoholic fatty liver disease in which TG accumulate and plasma albumin and protein decrease.
On the contrary, higher amounts of PL and free Cho were present in
P. chabaudi AS-infected mice compared to contr or
P. berghei NK65-infected mice. This is possibly related to reticuloendothelial hyperplasia and/or cholestasis, similar to those reported in the same or other models of malaria infection [
7,
38]. The fatty acid distribution in the liver tissue of
P. chabaudi AS or
P. berghei NK65 infected mice at different days after the infection confirms the impairment (possibly induced by ROS) of the elongation/desaturation pathway from linoleic to arachidonic acid and may explain the higher linoleic/arachidonic acid ratio in the liver of
P. berghei NK65-infected mice [
39]. These observations also contribute to explain the changes in the lipid content and composition that was reported in the lungs of
P. berghei NK65 infected mice [
12].
All together, these findings are in agreement with other models of
P. berghei infection and support the hypothesis that ROS may play a critical role in the liver dysfunction caused by
P. berghei NK65 parasites [
35,
40]. Consistent with this hypothesis are also the signs of fibrosis, revealed by the elevated hepatic levels of hydroxyproline at day 8 post infection in
P. berghei NK65-infected mice. Indeed, it is well known that oxidative stress is one of the major stimuli of fibrosis [
41]. Excess of ROS may induce severe liver damage followed by a phase of repair [
42] during which TGF-β1 on one hand limits the proliferative response of hepatocytes, and on the other hand, increases the production of collagen and other extracellular matrix proteins. Submicroscopical signs of fibrosis were evident also in
P. chabaudi AS-infected mice, although at a lower extent. Interestingly, at day 10 post infection, the amount of hydroxyproline decreased with both parasite species, becoming similar to control mice. The hypothesis that inflammation may lead to the activation of matrix metalloproteinases (MMPs) is under investigation. MMPs are proteolytic enzymes able to degrade different proteins of the extracellular matrix (e.g. collagen, laminin, fibronectin) and modulate cytokine and chemokine activity in cases of severe inflammation. Previous data indicated that, in the liver of C57Bl/6J mice infected with
P. berghei ANKA, the activity of MMPs is significantly increased [
43]. Furthermore, several in vitro studies have shown a TNF-dependent induction of MMPs at both mRNA and protein levels in monocytes fed with natural Hz [
16,
44].
Authors’ contributions
DS: Data collection, data analysis and interpretation, drafting the article. KD: Data collection and data interpretation. YC, PC and NL: Data collection. PVDS: Data analysis and interpretation, critical revision of the article, final approval of the version to be published. FOS: Data analysis and interpretation, critical revision of the article. DT: Data interpretation, critical revision of the article, final approval of the version to be published. All authors read and approved the final manuscript.