Model of severe malaria in young mice suggests unique response of CD4 T cells

Abstract Severe malaria occurs most in young children but is poorly understood due to the absence of a developmentally‐equivalent rodent model to study the pathogenesis of the disease. Though functional and quantitative deficiencies in innate response and a biased T helper 1 (Th1) response are reported in newborn pups, there is little information available about this intermediate stage of the adaptive immune system in murine neonates. To fill this gap in knowledge, we have developed a mouse model of severe malaria in young mice using 15‐day old mice (pups) infected with Plasmodium chabaudi. We observe similar parasite growth pattern in pups and adults, with a 60% mortality and a decrease in the growth rate of the surviving young mice. Using a battery of behavioral assays, we observed neurological symptoms in pups that do not occur in infected wildtype adults. CD4+ T cells were activated and differentiated to an effector T cell (Teff) phenotype in both adult and pups. However, there were relatively fewer and less terminally differentiated pup CD4+ Teff than adult Teff. Interestingly, despite less activation, the pup Teff expressed higher T‐bet than adults' cells. These data suggest that Th1 cells are functional in pups during Plasmodium infection but develop slowly.

In establishing a model for malaria disease in younger mice, we consulted the literature on maturation of mice. A mathematical model has been widely used which relies on the developmental similarities between mice and humans to calculate analogies between mouse and human ages. In this formula, fusion of growth plates in the scapula is used as a marker of transition from adolescence to adulthood, when humans attain sexual maturity. This makes it possible to determine maturation status by dividing the average sexual maturity for humans in days, by that of mice. Based on this model, it is suggested that 2.60 mouse days are equivalent to one human year. 6 Thus, the calculations provided by this model suggest that 14-15 mouse days is developmentally analogous to five human years, a common cutoff used in malaria susceptibility of children, as most cerebral malaria deaths from P. falciparum occur before that age. Establishment of such a model will enable the determination of immune development and response to Plasmodium infections in young mice. Such information can provide beneficial knowledge for improved design of malaria treatments and vaccines targeted for children.
Immunity to severe malaria develops gradually as children in highly endemic areas sustain multiple infections with P. falciparum.
Clinical immunity reduces disease severity and likelihood death from the infection, 7 while parasite immunity depends on the accumulation of antibodies specific to parasite strains a person is exposed to over time. 8 Immunity to parasite growth requires both B and T cells, with CD4 + T cells central to regulation of both antibody responses and phagocytic killing of parasites. While Th1 responses are beneficial for parasite killing, they also are implicated in severe malaria. 9 In some cases, this is accompanied by increases in FoxP3 + regulatory CD4 + T cells, though Tregs decrease after multiple incidences of malaria. 10 These regulatory mechanisms create a balance between pathogenspecific immunity and immune-mediated pathology. 11 These immune changes have been studied in P. chabaudi infection. 12,13 While P. chabaudi has been well-established as a reliable model of pathogenesis and immunity, there should be more investigations of the role of age on immunity and pathogenesis.
Both young mice and children have a lower overall immune response to various stimuli including Plasmodium parasites. [14][15][16] However, the mechanisms at play in determining increased severity to disease in young people are not clear, and it is not certain if this is a defect in their ability to develop immunity, or an increase in the propensity to cerebral malaria, which T cells contribute to. 17 In this current study, we utilize C57BL/6 mice to describe the immune response to P. chabaudi infection in a young rodent model. Such a model would be beneficial in understanding the development of immunity to malaria in young children.

| SHIRPA tests to assess strength and sickness
The mouse is transferred quickly into a new environment with little human contact in order to observe the immediate reaction. For transfer arousal, a score of 5 showed that the mouse did not freeze when in a new environment while a lower score indicates a longer pause before moving. For tail and pelvic elevations, the mouse is observed during forward motion, as it explores its new environment. A score of 2 indicates that the tail is elevated, a 1 indicates that it is extended horizontally, and a score of 0 shows that the tail is being dragged. If the pelvic region was more than 3 mm in elevation the score of 3 was assigned, 2 represented a normal elevation of 3 mm, 1 showed that the pelvic elevation was barely touching the floor and 0 was the pelvis was flattened on the ground. Mice were in the beaker for 5 min before body position was recorded. A score of 4 indicated that the mouse was standing on their hind legs, a score of 3 meant that the mouse was sitting, a score of 2 meant that the mouse was lying prone, and a score of 1 indicates that the mouse was lying on its side. When the mouse is placed in the beaker, it is observed for spontaneous activity. A score of 3 shows that the mouse had a rapid and darting movement, a score of 2 meant vigorous grooming and moderate movement, 1 represents casual grooming or slow movement. When a mouse was resting or showed no movement it was given a score of 0. As the mouse was exploring the new environment the pelvic elevation was observed.

| SHIRPA tests to assess neurological deficits
The mouse was placed in an open arena and allowed to explore the new environment. The arena had equal length and width in squares.
For locomotor activity, the number of squares all four paws entered was counted over 30 min. For tremor, the mouse was observed when placed in a glass beaker. A score of 2 shows no tremor, a 1 represents a mild tremor, and a 0 shows that an important tremor was observed.
For negative geotaxis, a mouse was placed on a horizontal cage top and when the mouse moved in one direction the top was raised vertically so that the animal was facing downwards. A stopwatch was set for 30 s, and the mouse was observed. A score of 4 indicates that the mouse turned around and climbed up the grid, a 3 represents a mouse that turned around but froze. When a mouse moved but did not turn around it was given a score of 2, 1 was given when the mouse did not move for 30 s, and 0 represents when the mouse fell off grid. For touch escape, a mouse was stroked by a finger while exploring its new environment. If the mouse vigorously escapes the finger stroke, it was given a score of 3, if the mouse had a rapid response to a light stroke it was given a score of 2, and a score of 1 was given to mice that did not escape upon a light finger stroke. For visual placing, the mouse is held by its tail and lowered to the cage top, and the extension of the forelimbs by the animal was observed. A score of 4 showed that the mouse had early vigorous extension around 25 mm above the cage top, 3 represents an extension of forelimbs before vibrissa contact around 18 mm, 2 shows that the extension of forelimbs occurred upon vibrissa contact, 1 means it occurred upon nose contact, and 0 shows that there was no response.

| Monitoring weight
Adult and young mice were infected with 10 6 P. chabaudi parasites for survival and behavioral studies. Mice were weighed daily on a balance in a glass beaker (1 L). Percent weight change was calculated using the original weight before infection on day 0 of the experiment [(Weight todayweight d0)/Weight d0] Â 100%.    Pups that died strikingly had not gained more than 30% weight by the third week, and pups that lived also had significantly reduced growth. This result is consistent with stunted growth observed in children in malaria endemic areas, 20 and suggests that Plasmodium infection directly contributes to stunting. Infected adult mice lost weight at the peak of infection, but regained their normal weight after recovery by third week p.i. Some pups succumbed to the infection, resulting in 60% mortality ( Figure 1C), and these mice failed to gain weight by the third week of infection. Parasitemia trends showed an increase in pups, but this did not reach significance when compared to parasitemia in adult mice ( Figure 1D). Pups that died also had an additional recrudescence at day 14 p.ii. These data suggest that malaria infection leads to reduced growth rate of pups during infection.

| Statistical analysis
Malaria can lead to neurological complications and behavior problems in children. [21][22][23] Cerebral malaria can be detected in mice as behavioral changes. 19   . Fraction indicate number of dead mice on the indicated day and + indicate all mice dead on that day. Groups were compared using One-Way ANOVA followed by Tukey post-test with *p > 0.05 considered significant. * compares infected pups that survived to control pups, + compares infected pups that died to infected pups that survived, # comparing control pups to infected pups that died, and @ compares infected adults to infected pups that died. Fraction shows number of mice that died out of total mice  We previously showed a linear differentiation of effector cells into three stages based on CD27 and CD62L expression. 26 With lower Teff populations in the pups in Figure 3 above, we wondered if pup cells had any defects in the progression of effector T cell differentiation.
While we observed high proportions of Teff Early and Teff Intermediate in the pup cells ( Figure 4A-C), the numbers were not significantly different between the infected groups ( Figure 4E,F). The infected adult mice had significantly higher percentages and number of Teff Late (Figure 4D,G).
To measure functional differences in the effector CD4 + T cell activation, we tested proliferation-inducing (IL-2) and protective (IL-10, IFNγ, and TNF) cytokine production during Plasmodium infection. [27][28][29] Using intracellular cytokine staining and gating strategy as shown in ( Figure 5A), we observed a distinct CD4 + T population of double IFNγ/IL-10 producers that are well-defined to be protective from severe immunopathology. 12 While the IFNγ + and IFNγ/IL-10 double producers were lower in the infected pups compared to infected adults, they did not reach significance difference in both proportions and number ( Figure 5B-D). Interestingly, the IL-10 single producers were similar in both infected groups of pups and adults ( Figure 5E).
When we determined the proliferation cytokine IL-2 together with TNFα production, we observed that apart from IL-2 alone, which was higher in both the infected and uninfected pups, the  , Teff Intermediate (CD62L À CD27 + ), and Teff Late (CD62L À CD27 À ), respectively, gated on effector T cells (CD127 À CD44 + CD4 + ). Data represent two independent experiments with 3-4 mice per group. The error bars represent standard error of the mean (SEM) analyzed using One-Way ANOVA followed by Tukey's post-test. *p < 0.05, **p < 0.01, ***p < 0.001.  Figure 6B). This increase was found to be primarily due to extended high levels of T-bet in terminally differentiated Teff Late , while the Teff Early and Teff int were similar in both pups and adult mice ( Figure 6C). Similar to previous reports, 10 we observed a reduction in Foxp3 + CD4 T cells in both the infected pup and adult mice ( Figure 6D).

| Pup CD4 + T cells express high levels of T-bet after Plasmodium chabaudi infection
Taken together with previous studies, these results suggest a unique differentiation state of Th in pups infected with P. chabaudi. Despite the reduced number and less terminal differentiation state of CD4 + Teff in infected pups, the Teff appear to sustain expression of T-bet into the terminal differentiated subset of Teff. If this represents a strongly polarized Th1 state, it suggests a possible mechanism for the increased behavioral symptoms seen in pups.

| DISCUSSION
Malaria is a significant threat to children in endemic areas. To our knowledge, there is no good established young rodent model for our experiments for pups that recovered ( Figure 1B).
Similar to children who suffer more incidence of severe malaria, 38 we observed that more pups than adult mice experience severe disease, as demonstrated by increased mortality. We also tested several behavioral assays designed to indicate sickness behavior, such as that induced by systemic cytokines in response to transient LPS exposure, and specific behavioral indicators of cerebral involvement in (Figure 2). Pups that succumbed to infection had reduced scores on spontaneous activities as well as more complex behaviors such as negative geotaxis, compared to infected adult mice or pups that survived the infection.
Negative geotaxis and tremor are reflective of specific changes seen in cerebral malaria in WT mice infected with P. berghei ANKA, or IL-10 KO infected with P. chabaudi, and are not seen in adult WT mice. 19,24,25 These syndromes include extensive cerebral gliosis and vascular events including hemorrhage, edema, and endothelial congestion and coagulation throughout the brain, indicating neuropathology. 25 Additionally, negative geotaxis, or ataxia, and reflex reaction are specifically associated with cerebral malaria and can distinguish the neuro-specific aspects of P. berghei infection from sickness behavior. We anticipate to test the correlation of these behavioral changes with inflammation and leakage of the blood brain barrier in future studies.  While further work needs to be done to establish the role of CD4 + and CD8 + T cells in the brain of infected pups, as they are required for experimental cerebral malaria; 42 both cell types also make IFNγ that is essential to experimental cerebral malaria in adult mice.
T-bet and IL-12Rβ2, a marker of Th1 commitment are required for experimental cerebral malaria. 42,43 We speculate that the less differentiated pup Teff phenotype in our study, suggest an interesting pup T cell bias toward multipotency, another interesting finding to followup using this new model.
Higher numbers of regulatory T cells (Tregs) are reported in children from high malaria transmission areas. 44 Tregs require transforming growth factor beta (TGFβ) for their development and maintenance, and its presence inhibit inflammatory cytokines important for parasite elimination. 45 This could be one reason why children experience severe disease and poor immune response, resulting in death from malaria infection. In a recent study, it was shown that Treg cell numbers decline over time in children who are heavily exposed to several malaria infections, which correlates with reduced symptoms. 10 Similarly, we observed a decrease in Foxp3 + cells in both infected pups and adult mice in our current study supporting an increase in the Th1 to Treg ratio. 13 One of the most important questions in malaria research is if age is the primary factor influencing susceptibility to cerebral malaria, or if the young age of patients is due to their exposure from birth. In

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1111/pim.12952.