Deficiency in astrocyte CCL2 production reduces neuroimmune control of Toxoplasma gondii infection

Toxoplasma gondii is an obligate intracellular parasite that infects one-third of the world’s human population and establishes infection in the brain. Cerebral immune cell infiltration is critical for controlling the parasite, but little is known about the molecular cues guiding immune cells to the brain during infection. Activated astrocytes produce CCL2, a chemokine that mediates inflammatory monocyte recruitment to tissues by binding to the CCR2 receptor. We detected elevated CCL2 production in the brains of C57BL/6J mice by 15 days after T. gondii infection. Utilizing confocal microscopy and intracellular flow cytometry, we identified microglia and brain-infiltrating myeloid cells as the main producers of CCL2 during acute infection, and CCL2 was specifically produced in regions of parasite infection in the brain. In contrast, astrocytes became the dominant CCL2 producer during chronic T. gondii infection. To determine the role of astrocyte-derived CCL2 in mobilizing immune cells to the brain and controlling T. gondii infection, we generated GFAP-Cre x CCL2fl/fl mice, in which astrocytes are deficient in CCL2 production. We observed significantly decreased immune cell recruitment and increased parasite burden in the brain during chronic, but not acute, infection of mice deficient in astrocyte CCL2 production, without an effect on peripheral immune responses. To investigate potential mechanisms explaining the reduced control of T. gondii infection, we analyzed key antimicrobial and immune players in host defense against T. gondii and detected a reduction in iNOS+ myeloid cells, and T. gondii-specific CD4+ T cells in the knockout mice. These data uncover a critical role for astrocyte-derived CCL2 in immune cell recruitment and parasite control in the brain during chronic, but not acute, T. gondii infection.


Introduction
An effective neuroinflammatory response against infection requires the accuracy and precision to enable sufficient immune cell infiltration to the brain to control the pathogen while preserving the brain itself.The recruitment of peripheral immune cells to the brain is a highly orchestrated process driven in part by the brain's chemokine milieu.The unique repertoire of chemokine receptors expressed on circulating immune cells allows them to be recruited by distinct combinations of chemokines.In addition, as cells infiltrate the brain, their own chemokine production can further alter the landscape, drawing other immune cells into the tissue.Therefore, the evolving chemokine environment in the brain enables targeted and precisely timed immune cell mobilization to protect against infections of the brain.
Toxoplasma gondii is an obligate intracellular parasite with the unique ability to traverse the blood-brain barrier and infect the brain [1,2].T. gondii is among the most successful parasites, as it can invade almost all nucleated cells, infect virtually all warm-blooded animals, and is estimated to have infected one-third of the global human population [3,4].This parasite is transmitted via ingestion when humans consume water contaminated with oocysts or undercooked meat containing tissue cysts, or can spread vertically via congenital infection [5][6][7][8].After ingestion, T. gondii is able to cross the intestinal epithelium, enter the bloodstream, and disseminate throughout the body, eventually reaching the brain where it establishes a chronic infection [9].
T. gondii has a lytic life cycle, and the rapidly replicating tachyzoite stage during acute infection can cause cellular damage as the parasites lyse out of infected cells and invade neighboring host cells [10].In most immunocompetent individuals, infection is asymptomatic, but T. gondii stage converts into cyst-forming bradyzoites, which are slow-growing and found predominantly in neurons and muscle cells during chronic infection [11,12].However, in immunocompromised people, T. gondii can be potentially fatal due to loss of effective immune control [13], and cerebral toxoplasmosis causes 10% of AIDS-related deaths in Africa [14].
Inflammatory monocytes play a critical role in immune defense against the parasite during both the acute and chronic stages of the infection [27][28][29].These cells express high levels of the chemokine receptor CCR2, which binds to its ligand, the chemoattractant CCL2 [30].CCR2 + monocyte infiltration of the brain coincides with parasite entry to the brain during acute infection, and these cells persist in the brain during the chronic stage of infection [31].However, the factors driving this recruitment are poorly understood.
The production of CCL2 during T. gondii infection is necessary for survival, as mice lacking this chemokine have decreased monocyte recruitment and succumb to acute T. gondii infection [28].During infection, CCL2 recruits CCR2 + monocytes out of the bone marrow into the circulation, and ultimately into injured or inflamed tissues [32].In the periphery in mice, CCL2 is produced during recognition of the T. gondii protein profilin, which leads to the recruitment of Ly6C hi monocytes to sites of infection [33].In vitro, CCL2 production is triggered by the release of the alarmin S100A11 by T. gondii-infected myeloid cells [34].
In the brain, glial cells can produce CCL2 in the context of neuroinflammatory diseases [35,36], and during chronic T. gondii infection, astrocytes comprise 75% of the CCL2-producing cells [37].Additionally, if astrocyte activation is suppressed, the levels of CCL2 and of CCR2 + cells decrease in the brain [38].We aimed to determine the role of astrocyte-derived CCL2 in immune cell recruitment and control of T. gondii from acute to chronic infection of the mouse brain.We found that microglia and brain-infiltrating myeloid cells produce high levels of CCL2 during acute infection in close proximity to the parasites.In contrast, astrocytes become the dominant CCL2 producers during chronic T. gondii infection.By infecting mice specifically deficient in astrocyte CCL2 production (GFAP-Cre x CCL2 fl/fl ), we observed significantly decreased immune cell recruitment to the brain and reduced parasite control during chronic, but not acute, infection.

CCL2 is produced by astrocytes, microglia, and infiltrating myeloid cells during acute T. gondii infection
CCL2 is transcribed in the brain as early as 7 days post-infection (DPI) and persists at 10, 30, and 60 DPI [39].To examine CCL2 protein levels in the brain, wild-type C57BL/6J mice were injected intraperitoneally (i.p.) with 200 type II T. gondii tachyzoites or PBS as a control.The brains were harvested at 15 DPI, which is an acute infection timepoint that represents the peak of CCR2 + monocyte infiltration to the brain in our model [31].Brains were snap frozen and homogenized, and CCL2 levels in brain homogenates were analyzed by ELISA.We detected a significant increase in CCL2 protein levels by 15 DPI (Fig 1A).To investigate the cells that produce CCL2 during acute infection, we used CCL2-RFP reporter mice, in which all cells expressing CCL2 also express RFP [40].The CCL2-RFP mice were injected with PBS or infected with T. gondii as above, and the brains were harvested at 15 DPI and sectioned for confocal microscopy.We stained for GFAP + astrocytes, CCL2-producing (RFP + ) cells, and Iba-1 + cells, which include microglia, macrophages, and mature monocytes, as these cells have been found to produce CCL2 in other neuroinflammatory diseases, including multiple sclerosis [41].We found significantly more GFAP + , Iba-1 + , and CCL2-RFP + cells in the brains of T. gondii-infected mice compared to those of PBS-injected mice (Fig 1B).Most of the CCL2 + cells were Iba-1 + myeloid cells, with some CCL2 + astrocytes at 15 DPI (Fig 1C).The mean fluorescence intensity (MFI) of CCL2-RFP was significantly higher in T. gondii-infected mice than in the same regions of the brains of control PBS-injected mice (Fig 1D).
To further analyze the Iba-1 + CCL2-producing cells we stained brain sections for both Iba-1 and Mac2, which predominantly stains infiltrating myeloid cells [42], but is also expressed in some microglia [43].In the brains of infected mice, we observed a significant increase in Mac2 + cells specifically in FOV containing parasites (Fig 2A and 2B, green circles).By focusing on the CCL2-RFP + cells, we detected most of the CCL2-RFP signal within Iba-1 + Mac2 -microglia (Fig 2C, blue circles).Interestingly, CCL2-RFP signal was also detected within Mac2 + positive cells, suggesting that infiltrating myeloid cells that are recruited to sites of T. gondii infection are also producing CCL2 (Fig 2C, green circles).To further characterize the CCL2-RFP + myeloid cells, we utilized an anti-Ly6C antibody, which stains monocytes, and found CCL2-RFP signal within Ly6C + cells (S1A Fig) .We confirmed these findings by staining with anti-Ly6B.2,which stains both infiltrating neutrophils and monocytes, but not lymphocytes, and also observed colocalization of CCL2-RFP and Ly6B.2 (S1B Fig) .In addition to infiltrating myeloid cells, microglia, and astrocytes, it has been shown that neurons can produce CCL2 in a model of viral CNS infection [44].To examine the relative contribution of neurons to CCL2 signal in the brain, we also stained brain sections with antibodies against the neuronal marker NeuN and compared CCL2 signal in NeuN + neurons to that of Iba-1 + Mac2 -microglia, Mac2 + myeloid cells, and GFAP + astrocytes in FOVs with and without parasites.Although neurons were readily detectable in these FOVs (Figs 2B and S1C), there was little to no CCL2 signal within these cells (Figs 2C and S1C), indicating that neurons are not likely to be a substantial source of CCL2 in the T. gondii-infected brain at this acute timepoint.
We also used flow cytometry to determine the frequencies of immune cells producing CCL2 in the brains of control or T. gondii-infected mice at 15 DPI.We found that the percent of microglia (CD45 int CD11b + ), infiltrating myeloid cells (CD45 hi CD11b + ), including Ly6C hi and Ly6C lo monocytes, and neutrophils (Ly6G + ) producing CCL2-RFP increased in infected compared to control mice (Fig 3C ), consistent with the microscopy data (Fig 2B).On average, 10.33% of microglia, 7.77% of Ly6C hi and 12.71% of Ly6C lo monocytes, and 1.14% of neutrophils were also positive for CCL2-RFP during infection (Fig 3C).We also examined the mean Therefore, during acute infection, immune cells are recruited to the brain and meninges.Although resident microglia comprise the majority of CCL2-producing cells in the brain, Ly6C hi and Ly6C lo monocytes, neutrophils, and astrocytes also produce CCL2 at this timepoint.

Astrocyte-derived CCL2 is not required for immune cell recruitment and control of acute T. gondii infection
To elucidate the role of astrocyte-derived CCL2 during infection, we generated mice in which astrocytes were specifically deficient in CCL2 production.We bred CCL2 fl/fl mice to mice expressing Cre recombinase driven by the astrocyte-specific GFAP promoter (GFAP-Cre 77.6 mice [46]) to generate GFAP-Cre CCL2 fl/fl mice.The genotype of these mice was confirmed using PCR on genomic DNA (S4A and S4B Fig) .We first ensured that the knockout of astrocyte-derived CCL2 did not affect immune cell frequencies in the brain or periphery in the absence of infection.GFAP-Cre CCL2 fl/fl and control CCL2 fl/fl mice were intraperitoneally injected with PBS (Fig 4A ), the mice were euthanized, and their spleens, brains, and meninges were harvested 15 days later.We detected no differences in the myeloid or lymphoid immune cell subsets in the spleen (S5A To determine the role of astrocyte-derived CCL2 in early CCR2 + cell recruitment and parasite control, we infected GFAP-Cre CCL2 fl/fl and control CCL2 fl/fl mice as above with 200 type II GFP-expressing T. gondii and collected the spleen, meninges, and brain at 15 DPI (Fig 4A ).By conducting confocal microscopy on brain sections from infected GFAP-Cre CCL2 fl/fl mice, we confirmed that GFAP + astrocytes did not express CCL2-RFP, indicating knockout of astrocyte-derived CCL2 in infected mice (Fig 4C).We also detected a decrease in ccl2 transcripts in the brain in GFAP-Cre CCL2 fl/fl during acute infection (Fig 4D).In contrast, ccr2 transcripts in the brain were similar in control CCL2 fl/fl mice and GFAP-Cre CCL2 fl/fl mice during infection (Fig 4E).We also examined which cells expressed CCR2 in the T. gondii-infected brain and found that the vast majority (>90% on average) of CCR2-RFP-expressing cells were monocytes (S6 Fig) .To determine if any specific immune cell subsets were affected by the loss of astrocyte-derived CCL2, we conducted flow cytometry on brain homogenates during infection and found no differences in immune cell recruitment to the brain (Fig 4F).There were also no differences in immune cell frequencies in the spleen or meninges of mice deficient in astrocyte-derived CCL2 (S5C and S5D Fig) .Finally, to determine if the loss of astrocytederived CCL2 affected the ability of the mice to control T. gondii infection, we measured levels of the parasite B1 gene in the brain by qPCR and detected no difference in B1 levels in the GFAP-Cre CCL2 fl/fl mice compared to CCL2 fl/fl mice at 15 DPI (Fig 4G).With these results, we concluded that astrocyte-derived CCL2 does not affect peripheral immune cell mobilization to the brain or meninges, nor control of parasite burden at this timepoint.

Astrocyte-derived CCL2 induces immune cell recruitment to the brain to control parasite burden during chronic T. gondii infection
We next aimed to determine the importance of astrocyte-derived CCL2 during chronic infection when parasites have converted into the slow growing bradyzoites within cysts, and astrocytes comprise the majority of CCL2 producing cells [38].We infected GFAP-Cre CCL2 fl/fl and control CCL2 fl/fl mice as above and euthanized the mice at 28 DPI.CCL2-RFP was detected in control mice during chronic infection and was significantly reduced in astrocytes in GFAP-Cre CCL2 fl/fl mice (Fig 5A and 5B).To control for the specificity of CCL2 deletion, we confirmed that there was no reduction in CCL2-RFP signal in Iba-1 + Mac2 -microglia, Mac2 + myeloid cells, or NeuN + neurons in the GFAP-Cre CCL2 fl/fl mice (S7A-S7C Fig), as expected.Since astrocytes are the major producers of CCL2 during chronic infection, we examined the extent to which astrocytes contribute to overall CCL2 production in the brain at this timepoint.We measured ccl2 expression in the brain using real-time qPCR and detected a 70% decrease in ccl2 mRNA levels in the brains of the knockout mice during infection (Fig 5C ).
To determine how deficiency in astrocyte-derived CCL2 affects recruitment of CCR2 + cells to the brain, ccr2 transcripts in the brain were examined at 28 DPI.Unlike during acute infection, the deficiency in astrocyte-derived CCL2 led to a 50% decrease in ccr2 mRNA levels in the brain (Fig 5D).To identify the immune cell subsets affected by this decrease in CCL2 levels during chronic T. gondii infection, we examined myeloid cells, granulocytes, and T cells in the brain.There was no change in the frequencies of microglia (CD45 int CD11b + ) in the absence of astrocyte-derived CCL2, as expected.However, the mobilization of infiltrating myeloid cells (CD45 hi CD11b + ), inflammatory monocytes (CD45 hi Ly6C hi ), and neutrophils (Ly6G + ) all declined in the brains of the knockout mice (Figs 5E and S8).We also detected reductions in T cell numbers in the brain, including helper T cells (CD4 + ), cytotoxic T cells (CD8 + ), regulatory T cells (Foxp3 + ), and proliferating T cells (Ki67 + ) (Fig 5F ), indicating a role for astrocytederived CCL2 in mobilizing these cells to the brain during chronic infection.In contrast, in the meninges there was a difference in the number of infiltrating monocytes, but not of granulocytes or T cells in the knockout compared to the control mice (S5F Fig) .Additionally, the frequencies of immune cells in the spleen were unaffected by deficiency in astrocyte-derived CCL2, as expected (S5E Fig) .Consistent with prior studies demonstrating that CCR2 + cells are important for the control of T. gondii infection in the brain [27], the reduction in CCR2 + immune cell infiltration was associated with a doubling of the total number of cysts in the brain (Fig 6A).However, the knockout mice did not exhibit impaired survival or increased weight loss out to 28 DPI (S9A and S9B Fig) .These results indicate that astrocyte-derived CCL2 plays a critical role in the control of T. gondii burden, but not necessarily survival, during chronic infection of the brain.
To determine potential mechanisms explaining the reduced control of T. gondii infection, we analyzed key antimicrobial and immune mechanisms of host defense.As T cells are required for controlling T. gondii infection, we measured T. gondii AS15 tetramer-specific T cells [47].We detected decreased levels of CD4 + T. gondii tetramer + cells in the brain (Fig 6B ).iNOS is increased in myeloid cells during T. gondii infection and plays a critical role in the control of the parasites during chronic infection [23] [26].To determine if this antimicrobial pathway was affected in mice deficient in astrocyte-derived CCL2, we measured iNOS + myeloid cells using flow cytometry and detected a decrease in the number of these cells in the knockout compared to the control mice during infection (Fig 6C).These data demonstrate (G) Assessment of brain parasite burden by qPCR for B1 from CCL2 fl/fl and GFAP-cre CCL2 fl/fl mice at 15 DPI.In (B) n = 4-5 mice per group, in (D and E) n = 8-11 mice per group, in (F) n = 4-7 mice per group, and in (G) n = 8-9 mice per group from two to three independent experiments.Statistical significance was determined by randomized block ANOVA, *p<0.05,ns: not significant.https://doi.org/10.1371/journal.ppat.1011710.g004

PLOS PATHOGENS
Astrocyte CCL2 is required for immune control of chronic T. gondii infection that during chronic infection, astrocyte-derived CCL2 plays a key role in driving parasite-controlling immune cells to the brain.

Discussion
Neuroinflammatory responses must be tightly regulated to enable sufficient immune cell infiltration to the brain during infection but limit excessive immunopathology, which is correlated with neuroinflammatory and neurodegenerative diseases.As toxoplasmic encephalitis continues to account for AIDS-related deaths [14], the drivers of immune cell mobilization to the brain during T. gondii infection warrant further investigation.Prior studies have shown that CCL2 is necessary for host immune protection against T. gondii infection [28].However, whole-body CCL2 knockout mice have reduced levels of circulating immune cells, since CCL2 is required for monocyte egress from the bone marrow.In the brain, CCL2 is correlated with immune cell recruitment, but the role of specific CCL2-producing cells in the brain during T. gondii infection is unknown [38].Our study found that astrocyte-derived CCL2 plays a pivotal role during chronic, but not acute, T. gondii infection in the brain.
In examining the upregulation of CCL2 in the brain during acute T. gondii infection, we found that CCL2-expressing cells were not uniformly dispersed throughout the brain.Rather, clusters of CCL2-producing cells were detectable in the cerebrum, frequently surrounding T. gondii tachyzoites.As monocyte infiltration and localization in the brain is highly associated with parasite clusters [31], T. gondii infection may be a direct driver of CCL2 production, thereby inducing this region-specific monocyte recruitment in the brain.T. gondii is known to trigger CCL2 production in infected host cells through the activities of the dense granule protein GRA25 [48].However, it is also clear that direct invasion of host cells in the brain by T. gondii is not required for CCL2 production.Merritt et al. found that CCL2 expression is upregulated in regions of the brain cells without significant interaction with T. gondii, but was found within 120 μm of a parasite-interacted cell [49].In the periphery, innate immune recognition of T. gondii profilin by TLR11/12 induces CCL2 production and monocyte recruitment [33].The production of CCL2 in the periphery and in human monocytes is also regulated by S100A11 in a RAGE-dependent manner [34].Therefore, T. gondii likely induces a localized upregulation of CCL2 in the brain through a combination of parasite and host factors.Additionally, the nonuniform distribution of CCL2 across the brain may explain why astrocytederived CCL2 was more important for recruiting immune cells to the brain parenchyma than to the meninges.These data suggest that CCL2 production by specific cells in the brain may vary in importance depending on its location.
Interestingly, when we analyzed CCL2 production from acute to chronic infection, we found that the major CCL2-producing cell type transitioned from myeloid cells to astrocytes.It is possible that the cues driving CCL2 production during acute infection differ from those during chronic infection, particularly since the composition of immune cells and cytokines in the brain changes as the parasites undergo stage conversion from rapidly replicating tachyzoites to slow-growing bradyzoites that are encysted within neurons.Microglia activation by TNF-α induces CCL2 production and monocyte recruitment to the brain in a model of hepatic inflammation [50].By 10 days after T. gondii infection, ILCs are recruited to the brain, and these cells are correlated with TNF-α and CCL2 expression in the brain [51].Therefore, during acute T. gondii infection, TNF-α production by ILCs may be inducing CCL2 production in microglia to recruit the monocytes to the brain.In contrast, during chronic infection, CCL2 production by astrocytes is induced by the IL-33 alarmin in close proximity to T. gondii [38,52].Notably, the ability of astrocytes, but not myeloid cells, to recognize IL-33 via its receptor ST2 is key for CCL2 production, immune cell recruitment to the brain, and parasite control during chronic, but not acute infection [38].Therefore, the changes in the neuroimmune landscape between acute and chronic infection likely contribute to distinct cues and cell types producing CCL2 at different stages of the infection.
In infected astrocyte-specific CCL2 knockout mice, ccl2 expression was more significantly reduced during chronic infection compared to acute infection, and a decline in monocytes, neutrophils, and multiple subsets of T cells in the brain was only observed during chronic infection.These data suggest that the reduction in ccl2 during acute infection of the knockout mice was not sufficient to impact immune infiltration of the brain, or alternatively, that during acute infection, other chemokines orchestrate the initial recruitment of immune cells to the brain.Investigating a role for microglia-derived CCL2 in early immune cell recruitment to the brain would necessitate the use of a microglia-specific Cre driver mouse line.There are challenges with this strategy, since many markers for microglia are also expressed by peripheral myeloid cells, including monocytes and macrophages.However, an inducible system, such as the CX3CR1-CreERT2 model has proven successful for inducing reporter gene expression specifically in microglia during T. gondii infection [43].Crossing these mice to mice with floxed ccl2 may be a promising approach for analyzing the importance of CCL2 production by microglia in the control of T. gondii infection in the brain.
One surprising observation was the effect of astrocyte-specific CCL2 deficiency on immune cells that do not express the CCR2 receptor during T. gondii infection, including T cells and neutrophils.It is likely that this effect is indirect: during T. gondii infection, astrocyte-derived CCL2 recruits CCR2 + monocytes, which along with brain-resident cells, may produce chemokines that recruit non-CCR2-expressing T cells and neutrophils, as proposed in other models [53].However, unlike in the brain, Ly6C + monocyte infiltration of the meninges was reduced in mice deficient in astrocyte-derived CCL2 during chronic infection.These findings are notable, since astrocytes comprise the glia limitans, a thin barrier that separates the brain parenchyma from the pia mater of the meninges.These data suggest that astrocytes in the glia limitans contribute to CCL2 production, such that knocking out CCL2 in astrocytes impacts meningeal monocyte infiltration.These findings also indicate that other chemokines produced within the meninges may be the primary chemoattractants for T cell and granulocyte infiltration of the meninges.
Although the decreased immune cell mobilization to the brains of GFAP-cre CCL2 fl/fl mice was accompanied with a near doubling of cysts in the brain compared to control CCL2 fl/fl mice, the survival and weight loss of the mice did not differ between the two genotypes.While a doubling of parasite cysts represents a substantial increase in parasite burden, it may not be sufficient to affect survival of the mice.It is clear that CCR2 blockade in the chronic phase of infection is fatal [27].Astrocyte CCL2 reduces the infiltration of CCR2 + monocytes, but not to a degree that is fatal, suggesting that other sources of CCL2 maintain sufficient numbers of protective immune cells to provide a degree of host resistance against the parasite.

Parasite strains
Mice were infected intraperitoneally with 200 type II Toxoplasma gondii Prugniaud strain tachyzoites or 10 ME49 cysts in 200 μL of 1X PBS.Uninfected control mice were injected intraperitoneally with 200 μL of 1X PBS.Tachyzoites were maintained via serial passage in human foreskin fibroblasts, as described previously [54].T. gondii tissue cysts were obtained from infected Swiss Webster mice.

Flow cytometry
To isolate single cells from mouse brains, the harvested brains were minced and digested using dispase II (Toche Applied Science) diluted in Hepes-buffered saline.To filter out clumps, the mixture was triturated then passed through a 70 μm filter (Falcon), and myelin was removed using a 35% and 75% percoll (GE Healthcare) gradient.Alternatively, cells were isolated by mincing brains in RPMI with 10% FBS (R10).The brains were then homogenized using a 3 mL syringe and 18-gauge needle and digested in a 2 mL mixture of collagenase-dispase (Sigma), and DNAse I (ThermoFisher) in R10.The mixture was triturated then passed through a 70 μm filter (Falcon), and myelin was removed using a spin in 40% percoll.To

PLOS PATHOGENS
isolate single cells from spleens, spleens were homogenized and passed through a 40 μm filter (Falcon), and red blood cells were lysed used ACK lysing buffer (Gibco).To isolate single cells from meninges, the dura was collected from the skull and digested using collagenase D (Roche) and DNase I (Thermo Scientific), and the tissue was passed through a 70 μm filter (Falcon).

ELISA
Brains were snap frozen, homogenized using a mortar and pestle, and resuspended in 1X PBS with protease and phosphatase inhibitor (Thermo Scientific Cat#78444).Cells were centrifuged to pellet debris, and the supernatant was used for ELISA with the Mouse MCP-1 (CCL2) ELISA max kits from Biolegend (Cat#432704).Samples were run in triplicate and analyzed on a Spectra Max Plus 384 nm (Molecular Devices) spectrophotometer.

T. gondii cyst counts
100 mg of brain tissue was minced in 2 mL complete RPMI.The tissue was then passed through a 18-gauge and a 22-gauge needle.30 μL of the final brain homogenate was mounted on a microscope slide, and cysts were enumerated using a Brightfield DM 2000 LED microscope (Leica).Total cyst burden of the full brain was deduced from these numbers.

Statistical analyses
Statistics were conducted using t-test, ANOVA, or randomized block ANOVA (on two groups at identical time points).Graphs were generated in Prism using software version 9.3.1.The number of samples per group and statistical test utilized can be found in the figure legend for each figure.Significance was represented as follows: ns = not significant, *p<0.05,**p<0.01,***p<0.001,****p<0.0001.Error bars in all figures represent standard deviations.

Fig 2 .Fig 3 .
Fig 2. Production of CCL2 by cell type during acute T. gondii infection.CCL2-RFP mice were injected with PBS or infected with PRU strain T. gondii, and brains were harvested at 15 DPI.(A) Representative confocal microscopy of T. gondii (white), Mac2 + myeloid cells (green), CCL2-RFP (red), and Iba1 + myeloid cells (blue) in brains of PBS control mice or T. gondii-infected mice in FOV with or without parasites.(B) Percent area of each cell type within FOV with or without parasites from infected mice.(C) Percent area of each cell type within CCL2 + area in FOV with or without parasites from infected mice.n = 8-25 FOV from 4-5 mice per group from 4-5 independent experiments.Statistical significance was determined by two-way ANOVA.****p<0.0001,ns: not significant.https://doi.org/10.1371/journal.ppat.1011710.g002 Fig), the meninges (S5B Fig), nor the brain (Fig 4B).

Fig 4 .
Fig 4. Immune cell recruitment to the brain and parasite burden during acute T. gondii infection of mice deficient in astrocyte-derived CCL2.(A) Experimental design, created with Biorender.(B) CCL2 fll/fl (open circles) and GFAP-cre CCL2 fl/fl (closed circles) mice were injected with PBS, and immune cell numbers in the brain were quantified by flow cytometry 15 days later.(C) Representative confocal microscopy of T. gondii (white), GFAP + astrocytes (green), CCL2-RFP (red), and Iba1 + myeloid cells (blue) in GFAP-cre CCL2 fl/fl mice at 15 DPI with PRU strain T. gondii.(D and E) qPCR for ccl2 (D) or ccr2 (E) transcripts from brains of CCL2 fl/fl and GFAP-cre CCL2 fl/fl mice at 15 DPI.Transcripts are normalized to gapdh.(F) Quantification of brain immune cells by flow cytometry from CCL2 fl/fl and GFAP-cre CCL2 fl/fl mice at 15 DPI.

Fig 5 .
Fig 5. Astrocyte-derived CCL2 drives immune cell recruitment to the brain during chronic T. gondii infection.CCL2 fl/fl and GFAP-cre CCL2 fl/fl mice were infected with T. gondii and examined at 28 DPI.(A) Representative confocal microscopy of GFAP + astrocytes (green), CCL2-RFP (red), and Iba1 + myeloid cells (blue) in CCL2 fl/fl and GFAP-cre CCL2 fl/fl mice at 28 DPI with PRU strain.(B) Percent of CCL2 + GFAP + cells of total GFAP + cells per FOV was averaged for each mouse.(C and D) qPCR for ccl2 (C) or ccr2 (D) transcripts from brains of CCL2 fl/fl and GFAPcre CCL2 fl/fl mice at 28 DPI with ME49 strain.Transcripts are normalized to gapdh and shown relative to the mean transcript level of the CCL2 fl/fl mice.(E) Quantification of brain myeloid immune cells by flow cytometry of CCL2 fl/fl and GFAP-cre CCL2 fl/fl mice at 28 DPI with ME49 strain.(F) Quantification of brain T cells by flow cytometry at 28 DPI with ME49 strain.In (B) n = 5-7 mice per group from three experiments.In (C and D) n = 3-4 mice per group, and in (E and F) n = 7-8 mice per group from two experiments.Statistical significance was determined by Student's ttest (B-D) or randomized block ANOVA (E-F).*p<0.05,**p<0.01,***p<0.001,****p<0.0001,ns: not significant.https://doi.org/10.1371/journal.ppat.1011710.g005