Chronic infection with Mycobacterium lepraemurium induces alterations in the hippocampus associated with memory loss

Murine leprosy, caused by Mycobacterium lepraemurium (MLM), is a chronic disease that closely resembles human leprosy. Even though this disease does not directly involve the nervous system, we investigated a possible effect on working memory during this chronic infection in Balb/c mice. We evaluated alterations in the dorsal region of the hippocampus and measured peripheral levels of cytokines at 40, 80, and 120 days post-infection. To evaluate working memory, we used the T-maze while a morphometric analysis was conducted in the hippocampus regions CA1, CA2, CA3, and dentate gyrus (DG) to measure morphological changes. In addition, a neurochemical analysis was performed by HPLC. Our results show that, at 40 days post-infection, there was an increase in the bacillary load in the liver and spleen associated to increased levels of IL-4, working memory deterioration, and changes in hippocampal morphology, including degeneration in the four subregions analyzed. Also, we found a decrease in neurotransmitter levels at the same time of infection. Although MLM does not directly infect the nervous system, these findings suggest a possible functional link between the immune system and the central nervous system.


Results
Histopathological changes in liver and spleen of MLM-infected mice. Infection with MLM produces a chronic granulomatous disease in susceptible mouse strains such as BALB/c. In the present study, inoculation of 20 × 10 6 bacilli by intravenous route gave rise to progressive granulomatous lesions in liver and spleen parenchyma, with a concomitant steady increase in the number of bacilli (Fig. 1). Mice in the control group did not present morphological alterations in these organs (Fig. 1A,B). By day 40 post-inoculation, early granulomas formed by mononuclear cells (macrophages and lymphocytes) were clearly evident, although they showed scanty acid-fast bacilli by Ziehl-Neelsen staining (Fig. 1C,D). By day 80 post-infection (PI), granulomas expanded and tended to confluence with progressive increase of bacillary loads. At this time, granulomas made-up of bacilli-loaded macrophages were large lesions that replaced most of the liver and spleen parenchyma (Fig. 1E,F). By day 120, the granuloma fraction completely substituted the liver and spleen parenchyma and bacilli became uncountable (Fig. 1G,H).
Cytokine measurement in serum. Serum cytokines were measured using the Luminex ® LABScan 100 system (Bead mouse Th1/Th2 6-Plex Panel) (Table 1). IL-2 was not detected in the control group nor in the MLM-infected group throughout the entire period of infection (40 to 120 days). In contrast, IL-12 was detected at all infection times, yet, a significant statistical difference compared to the control group was only observed at 120 days of infection (F (10.51) = 22,3 p < 0.01). IFN-γ was not detected in the control group nor in the MLM-infected groups (40,80, and 120 days). Type-2 cytokines were not detected in the control group yet the infected animals showed increased levels of IL-4 at post-infection days 80 (49.05 ± 11.17 pg/mL) and 120 (73.99 ± 3.75 pg/mL) (F (138.8) = 36,3 p < 0.0001). Interleukin 5 (IL-5) and IL-10 were not detected in any group at any PI time.
Immunohistochemical cytokine detection. In the control group, IL-12 was only detected in macrophages of the splenic white pulp. By day 80 PI, IL-12 was observed in Kupffer cells and clusters of macrophages in the spleen white pulp. By days 80 and 120 PI, diffuse staining was observed in hepatic parenchymal Micrograph correspond to sections of liver and spleen from a mouse at 40 days PI (D). Scarce MLM in the cytoplasm of macrophages (arrows); some macrophages appear vacuolated and contain bacilli (inset). Liver (E) and spleen (F) sections from a mouse at 80 days PI show numerous macrophages (globi) with abundant MLM. Liver (G) and spleen (H) sections from a mouse at 120 days PI, large and coalescent granulomas substitute extensive areas of the liver and spleen (H/E, Ziehl-Neelsen stains, all micrographs X40, inset X400 magnification).
SCIENtIFIC REPORtS | (2018) 8:9063 | DOI:10.1038/s41598-018-27352-x macrophages and the red pulp of the spleen (Fig. 3). By day 80 PI, incipient granulomas were detected in the hepatic parenchyma and staining for IL-4 was observed in isolated and small clusters of lymphocytes. Multifocal IL-4 staining was also observed in lymphocytes of the spleen parenchyma. By day 120 PI, staining for IL-4 was intense mainly in the granuloma lymphoid infiltrate but also in the splenic parenchyma.

Working memory (T-maze test).
For the assessment of working memory, five mice of each group were tested. Each test consisted of 12 sessions with an inter-test period not longer than 2 min; the maximal time-lapse for each test was 120 seconds (sec). Figure 4A illustrates the latency time (sec), i.e., the time taken by each mouse of each group to complete the test. Latency time was 10.26 ± 1.12 sec in the control group, 48.8 ± 1.12 sec in the 40-day PI group, 92.18 ± 3.63 sec in the 80-day PI group, and 106 ± 4.62 sec in the 120-day PI group. Values at 40-, 80-and 120-PI days greatly differed from the value in the control group (F (86.59) = 298,3 p < 0.0001).
To evaluate short-term memory associated with reward, the percentage of attempts leading to a successful hit during the 12 sessions was calculated. The efficiency in reaching the arm with the reward was 78% in the control group, 64.8% in the 40-day PI group, 26.8% in the 80-day PI group, and 16.6% in the 120-day PI group, the last two values being statistically different from the one in the control group (F (27.81) = 42,3 p < 0.001) (Fig. 4B).
Time taken to succeed in reaching the award was also calculated (Fig. 4C). Animals in the control group took a shorter time to complete the test compared to the 40-day PI group (9.72 ± 1.18 vs 34.19 ± 3.32 sec), the 80-day PI group (76.37 ± 6.73 sec), and the 120-day PI group (73.66 ± 11.56 sec), with differences highly significant (F (39.95) = 140,3 p < 0.0001).
Spontaneous locomotor activity. The spontaneous locomotor activity test was applied to the animals of the four groups to detect possible damage in regions other than the hippocampus, e.g., corpus striatum and basal ganglia 26,27 . No altered locomotor activity was observed in any of the four groups during the 30-min test (Fig. 5).
Histological study in the hippocampus. To evaluate histological damage in the hippocampus, several parameters were taken into account, particularly those indicative of cell death such as basophilic pyknotic nuclei, acidophilic cytoplasm, and fragmented nuclei in regions CA1, CA2, CA3, and DG (bregma regions −1.82 and −2.46 mm) 28 . The general morphology of hippocampus in the control group was well preserved, with neurons of normal characteristics. By day 40 PI, neurons with pyknotic nuclei and acidophilic cytoplasm were observed in the four regions analyzed. By day 80, the number of pyknotic-acidophilic neurons increased and the thickness of the layer regions decreased. At 120 days PI, these alterations had further progressed. The histologic changes observed in the hippocampus of the MLM-infected animals are characteristic of non-specific degenerative processes (Fig. 6B). The most remarkable result that emerged from the analysis was the lack of cell infiltrates and bacilli in the brain parenchyma that might suggest direct microbial infection in the hippocampus (Fig. 6B). The percentage of neurons with morphological alterations indicative of cell damage in CA1, CA2, CA3, and DG, increased progressively and changes were highly significant when compared to the histology in the control group (p < 0.001) (Fig. 7).
Neurotransmitters quantification by HPLC. Figure 8A shows serotonin concentration (5-HT) in the different study-groups. The level of 5-HT was 0.069 ± 0.003 pmol/mg protein in the control group, 0.014 ± 0.001 pmol/mg in the 40-day PI group, 0.033 ± 0.001 pmol/mg in the 80-day PI group, and 0.036 ± 0.005 pmol/mg in the 120-day PI group. Values in the infected groups were significant lower than the value in the control group (F (39.86) = 23,3 p < 0.001). Figure 8B shows the results on the levels of dopamine (DA) in the four study-groups. These were 0.476 ± 0.02 pmol/mg in the control group, 0.026 ± 0.001 pmol/mg in the 40-day PI group, 0.031 ± 0.004 pmol/mg in the 80-day PI group, and 0.094 ± 0.02 pmol/mg in the 120-day PI group. Values in the infected groups were significantly lower than the value in the control group (F (145.7) = 29,3 p < 0.001). shows focal positivity in splenic macrophages in the control group. At 80 and 120 days PI hepatic parenchyma exhibits focal positivity in Kupffer and multifocal macrophages. In spleen, multifocal staining is observed in macrophages in the splenic white pulp (arrows) X400. IL-4 immunolabeling is negative in the control group but at day 80 PI IL-4 staining is present on incipient granulomas in liver and it appears multifocally distributed in lymphocytes in the splenic red pulp (arrows). Diffuse IL-4 positivity is observed in extensive inflammatory hepatic infiltrates at day 120 PI. Diffuse IL-4 staining is also observed on spleen lymphocytes (box). X200 and X400 micrographs.  Figure 8C depicts the levels of epinephrine (EP) in the four study-groups. These levels were 2.75 ± 0.15 pmol/ mg in the control group, 1.12 ± 0.11 pmol/mg in the 40-day PI group, 0.35 ± 0.01 pmol/mg in the 80-day PI group, and 0.5 ± 0.07 pmol/mg in the 120-day group. Values in the infected groups were statistically different from the value in the control group (F (116.9) = 29,3 p < 0.001). Figure 8D shows the levels of norepinephrine (NE) in the four study-groups. These values were 0.16043 ± 0.024 pmol/mg in the control group, 0.0016 ± 0.0009 pmol/mg in the 80-day PI group, and 0.03762 ± 0.015 pmol/mg in the 120-day PI group. NE was not detected in the 40-day PI group. Statistically significant differences were observed between control and MLM-infected groups (F (26.43) = 31,3 p < 0.001).

Discussion
The progressive decrease in cell-mediated immunity, CMI (anergy), is characteristic of chronic infectious diseases produced by pathogenic mycobacteria 12,29,30 . Here, we showed that infection of mice with Mycobacterium lepraemurium (MLM) led to a disseminated, multiorgan, infection that did not directly affect CNS 15 .
The immunopathological changes observed in murine leprosy at the late stages of the disease are driven by Th2 anti-inflammatory cytokines that facilitate progression of the disease in a manner similar to what happens in lepromatous leprosy 7 . Like in multibacillary leprosy patients 31 , we found coexistence in serum of IL-12 and IL-4 on day 80 PI, and also the presence of these cytokines in the liver and spleen of MLM-infected animals. The increase in the levels of IL-4 has been related to CMI anergy that develops as the infection progresses 32,33 , this observation coinciding with our results of high bacillary load and positive immunostaining for IL-4 in liver and spleen. The T-maze test was used to evaluate alterations in working memory along the infection 34 . Our results showed a significant increment in the latency-time that correlated with longer times of infection, reduction in the number of successful attempts, and increment in the time required to fulfill a successful event 35 . These results indicated a significant loss in the capacity to accomplish an activity associated to working memory and are in agreement with previous results that demonstrate that administration of LPS or bacteria deteriorate this brain activity [36][37][38][39][40] . The mechanism responsible for this brain deterioration has yet to be elucidated, but the neuroinflammatory effect of persisting high levels of cytokines is a likely cause of it 4,5,41,42 . These alterations have also been documented in models of non-infectious chronic stress [43][44][45] . Using the spontaneous locomotor activity test we excluded the possible existence of lesions in other regions of the brain as a result of the infection 26,27 . Contrastingly, the only alterations found were located in the hippocampus (Figs 4 and 5). Because our findings are based on a limited number of evaluated regions, the results should be interpreted with caution. Systemic chronic infections have repercussion on the CNS activity 3-5 and the persistent peripheral stimuli might lead to production of neurotoxic molecules in the brain, causing neuroinflammation 3 . Therefore, we searched for parenchymal alterations in the dorsal region of the hippocampus considering that it is a target region for infectious and non-infectious chronic stress [46][47][48] . We found pyknotic/acidophil neurons in the CA1, CA2, CA3, and dentate gyrus (DG) regions along with the progress of the infection. This is an important feature considering that the dorsal region of the hippocampus plays a key role in learning and memory processes [49][50][51] . Additionally, CA1, CA3, and DG constitute the classic tripartite-synapse pathway, an important route for memory processes 52 . While inflammatory stimuli, such as systemic LPS administration, impaired working memory affecting CA1, CA3, and DG regions due to ROS-mediated neuroinflammation 46 , CA3 deterioration altered the synaptic transmission to CA1 and the completion of memory processes 47,53 . On the other hand, the neuronal loss in DG is relevant because this region is involved in the early development of spatial memory and the production of new neurons along life 54,55 . Finally, the CA2 region participates in social memory and its deterioration affects recognition, spatial, and contextual memory 56,57 .
Morphological alterations were concomitant to changes in neurotransmitter levels, which are crucial in the optimal function of brain processes including learning and memory [58][59][60][61] . During infection, 5-HT and DA levels in the hippocampus of infected mice were lower than in the hippocampus of uninfected animals. The decrease of these neurotransmitters in the hippocampus has been shown to be associated to memory deficit, neuronal atrophy, and neuronal loss in a model of hepatic encephalopathy 59 . Also, memory and learning impairment has also

Conclusion
This work refers to a murine leprosy model in which the disease does not directly affects the nervous system but is, however, the cause of alterations in the working memory, the morphology of the dorsal region of the hippocampus, and the levels of neurotransmitters. We regard that murine leprosy as an infection model suitable for further exploring the effects of chronic infections on the structure and function of the central nervous system.

Materials and Methods
Mice.

Infection with MLM. Mycobacterium lepraemurium (MLM) is a hardly cultivable mycobacterium, resem-
bling Mycobacterium leprae in this aspect. Therefore, MLM (Hawaii strain) was isolated from liver and spleen of a mouse bearing a 4-month infection, as described elsewhere 62 . Viability of bacteria was assessed using the fluorescein diacetate-ethidium bromide method by Jarnagin et al. 63 , as previously described 62 .
A group of 30 mice was intravenously inoculated with 20 × 10 06 MLM in 50 µL saline solution. Control mice only received saline solution. Mice were sacrificed at the same time of the day to avoid circadian variations.
Histologic and morphometric analysis. Mice of the control group and those infected for 40, 80, and 120 days were euthanized by CO 2 inhalation and immediately heart-perfused with 10% neutral formalin. After 10-min perfusion, the brain, spleen, and liver were collected and preserved in the same fixer solution for 5 days. Then, the organs were trimmed and prepared for paraffin sectioning according to standard procedures. Three-micron thick sections were obtained using a Leica microtome, mounted on poly-L-lysine-coated slides, and stained with hematoxylin-eosin (H&E) for conventional histology, and Ziehl-Neelsen (ZN) stain for acid-fast bacilli. Images were taken in a Nikon Eclipse E8000 microscope (Tokyo, Japan). Bacillary density in the liver and spleen was quantified using Image J software and the results were given as red pixels 12 . Determination cytokines by immunohistochemistry. For immunohistochemistry, 4-µm thick sections of liver and spleen were prepared and collected on poly-L-lysine-coated slides. After deparaffination and quenching of endogenous peroxidase with 3% hydrogen peroxide solution and methanol by 30 min, sections were incubated with rabbit anti-IL-4 and IL-12 antibodies (Santa Cruz Biotechnology diluted 1/200 in PBS by 3 hrs at room temperature and subsequently incubated with goat anti-rabbit IgG-peroxidase diluted 1/500 for 30 min at room temperature. After three washings with PBS-Tween, sections were stained with haematoxylin, mounted with synthetic resin, and examined under the microscope. Images were taken in a Nikon Eclipse E8000 microscope (Tokyo, Japan).
Working memory. We decided to use the T-maze to reduce the stress generated by the averse stimuli employed in other mazes 64,65 . Working memory was evaluated by a standard protocol using a stainless-steel T-maze (OMNIALVA, Instruments, Mexico) consisting of three sections: a left and a right arm (75 × 20 × 40 cm each) and a starting box (30 × 25 × 30 cm). A small circular pot (6-cm diameter and 3-cm depth) was placed at the end of each arm. A food reward (0.5-g, Test-Diet pellet, chocolate flavor) was placed in the pot and could not be seen or smelled by the mouse from the start box. The working memory test was used to evaluate hippocampus-dependent memory 32 . For the habituation process, each mouse was placed in the T-maze after a 24 h fasting period. Mice were laid in the maze for 30 min and allowed to move freely to the aisle of their choice; they eventually found the sweet (award). Two days later, each animal was subjected to 12 trials of rewarded alternation in the T-maze. Mice were allowed a maximum of 120 sec to complete a trial. The T-maze allowed us to evaluate the latency time, percentage of success, and time for a success.

Spontaneous locomotor activity.
To estimate the spontaneous locomotor activity, each mouse was placed in a Plexiglas cage (50 × 50 × 30 cm) surrounded by a set of 16 light-beam photocell pairs appropriately positioned to register the number of light interruptions resulting from the animals' movement 66,67 . The light registered by the photocells was transformed into digital pulses, which were then sent to a desktop computer. Data were processed and analyzed by means of OABiomed software. Animals were adapted to the activity chamber for three sessions of 30 min each, and then they were individually placed into the chamber and monitored for 30 min.
For the behavioral tests, each group was randomly divided into two subgroups, each with five mice. The first group was used for the T-maze test and the second for the spontaneous locomotor activity test. In these experiments, working with five mice allowed us to avoid the stress induced by the fasting period and exposure to an open field in the spontaneous locomotor activity test.
Histologic and injury assessment in the hippocampus. Mice were euthanized by CO 2 inhalation and immediately hart-perfused with isotonic saline solution for 5 min and then with 10% neutral formalin. After a 10-min perfusion, brains were removed and included in brain matrix to a 1.70 mm bregma depth. Then, the brains were trimmed and prepared for paraffin sectioning according to standard procedures. For the morphometric analysis, 3-micron thick sections were collected on poly-L-lysine-coated slides and stained with the H&E and ZN stains. To investigate the presence of bacilli (MLM) in the CNS, brain sections were stained with the ZN stain. Five pictures from each section corresponding to the hippocampus dorsal region (bregma coordinates −1.82 and −2.46) were taken (Nikon Eclipse E8000 microscope,Tokyo, Japan). Five microscope fields of regions CA1, CA2, CA3, and DG were analyzed to calculate the number of cells showing morphological alterations.
Neurotransmitter quantification by HPLC. To quantify serotonin (5-HT), dopamine (DA), norepinephrine (NE), and epinephrine (EP), the hippocampus was homogenized using 400 µl of a solution containing 5% ascorbic acid, 200 mM sodium phosphate, 2.5 mM L-cysteine, and 2.5 mM EDTA. Then, the protein was precipitated by the addition of 100 µl of 0.4 M perchloric acid, followed by incubation at 20 °C for 20 min. Supernatants containing NE, EP, DA, and 5-HT were collected after centrifugation at 12,000 rpm for 10 min (4 °C). NE, EP, DA, and 5-HT concentrations were determined by reversed-phase HPLC (RP-HPLC) in a system integrated by two 515 pumps (Waters ™ ), a degasser AF (Waters ™ ), a 717 autosampler (Waters ™ ), and an X-LC ™ 3120FP fluorescence detector (Jasco, Inc). Instruments were controlled by Millenium 32 software (Waters ™ ). Chromatographic runs were performed using a Jupiter C18 column (300 Å, 5 μ, 4.6 × 250 mm, Phenomenex ® ) at 30 °C. Column was equilibrated with the mobile phase A (MPA) containing 0.1% trifluoroacetic acid. Mobile phase B (MPB) containing 0.1% trifluoroacetic acid in acetonitrile was used to perform a linear gradient until reaching 20% MPB, from min 5 to min 15. Then, 20% MPB was maintained until min 20; the flow rate was 0.8 ml/min. The fluorescence detector was set at gain 100, attenuation 32, response 20 s, and 280 nm and 315 nm for excitation and emission, respectively. The sample injection volume was 50 μl.

Statistical analysis.
Results are presented as the average mean value ± standard error of mean (SEM). Data for bacillary load, cytokine and neurotransmitter levels, hippocampus cell injury, and memory loss were analyzed by one-way ANOVA and Tukey's test for multiple comparison. Graphs were produced by using GraphPad Prism 4 software. A difference was considered significant when p < 0.05.