Hypercholesterolemia Impairs Immunity to Tuberculosis ^▿

Animals.

C57BL/6 and B6.129P2-Apoe^tm1UncN11 (ApoE^−/−) mice were purchased from Taconic Farms or bred at the University of Massachusetts Medical School (UMMS) with breeders purchased from The Jackson Laboratory. ApoE^−/− mice were originally derived by Piedrahita et al. (23) and backcrossed on a C57BL/6 background for more than 10 generations. C57BL/6 OT-II mice were a kind gift from Kenneth Rock, UMMS. The mice were housed within the Animal Medicine facility at UMMS, and the Institutional Animal Care and Use Committee approved these experiments.

Reagents.

The mice were fed either an LC (0.12% cholesterol) or an HC (1.25% cholesterol, 0.5% sodium cholate; Research Diets) diet 2 weeks before and during each experiment. Ovalbumin (OVA)-coated iron beads were a kind gift from Lian Jun Shen, UMMS. Chicken OVA was covalently bound to iron oxide beads according to the manufacturer's protocol (BioMag Amine; Polyscience Inc.). Culture filtrate protein (CFP) was obtained by growing M. tuberculosis Erdman in Sauton's broth for 1 month and then concentrating the culture supernatant and dialyzing it against phosphate-buffered saline (PBS) with a cutoff of 6 kDa.

Serum cholesterol.

Sera were incubated at 60°C for 30 min before being shipped to IDEXX Preclinical Research Services for total-cholesterol measurement.

M. tuberculosis infection.

A frozen stock of M. tuberculosis Erdman, ∼10⁸ CFU/ml in PBS-0.05% Tween 80, was thawed, the concentration was adjusted with PBS-0.05% Tween 80 to deliver ∼50 CFU per mouse, and it was sonicated for 5 min in a cup horn sonifier (Branson Ultrasonics Corporation) and mice were infected in a Glass-Col Inhalation Exposure System (Glas-Col, LLC). In every experiment, two mice were sacrificed 24 h postinfection (p.i.) to confirm the delivered dose.

Bacterial load.

Tissues were homogenized in PBS-0.05% Tween 80, serially diluted 10-fold over 4 to 5 log units, and plated in duplicate on Middlebrook 7H11 agar (Difco, Becton Dickinson) supplemented with oleic acid-albumin-dextrose-catalase, and cultured at 37°C, and the colonies were counted 3 weeks later.

Histopathology.

Lungs were inflated and fixed with 10% buffered formalin for ≥24 h and then processed for staining. Tissue sections were stained with hematoxylin and eosin (H&E) or for inducible nitric oxide synthase (iNOS) production with a 1:100 dilution of anti-iNOS/NOS II rabbit polyclonal immunoglobulin G (Millipore), and bound antibody was detected with a peroxidase-based ABC staining kit according to the manufacturer's protocol (Vector Laboratories). As a negative control, sections were stained in the absence of the primary antibody. All staining was done by the Diabetes and Endocrinology Research Center histopathology core facility at the UMMS.

Fluorescence-activated cell sorter analysis.

Lungs were perfused through the heart with PBS, minced, digested for 30 min at 37°C in collagenase IV (1 mg/ml) and DNase (25 μg/ml) (both from Sigma-Aldrich), and passed through a 40-μm cell strainer, and the remaining red blood cells were lysed with Gey's solution. Viable cells were counted using a hemocytometer and trypan blue dye staining. Lung leukocytes or splenocytes (1 × 10⁶ to 2 × 10⁶) were treated with Fc-blocking monoclonal antibody (clone 2.4G20; BD Bioscience Pharmingen) and then stained with allophycocyanin-cyanine 7-conjugated anti-CD3 (145-2C11), peridinin chlorophyll α-conjugated anti-CD4 (RM4-5), phycoerythrin (PE)-conjugated anti-CD8 (53.6-7), and PE-cyanine 7-conjugated anti-Gr-1 (RB6-8C5) (all from BD Bioscience Pharmingen) and allophycocyanin-conjugated anti-F4/80 (BM8; eBioscience). Flow cytometry was performed on an LSRII flow cytometer (BD Bioscience Pharmingen), 50,000 leukocyte-gated events were collected, and data analysis was done with FlowJo PC (TreeStar, Inc.). Isotype control antibodies were purchased from BD Bioscience Pharmingen and eBioscience.

Lung cytokine production.

Lungs were homogenized in PBS-0.05% Tween 80; an equal volume of cell lysis buffer was added (0.5% Trion X-100, 150 mM NaCl, 15 mM Tris, 1 mM CaCl₂, and 1 mM MgCl₂, pH 7.4); the lungs were vortexed, incubated for 20 min at 4°C, vortexed, and centrifuged for 10 min at 12,000 × g; and the supernatant was sterile filtered. Lung lysates were assayed for interleukin 1α (IL-1α), IL-1β, IL-2, IL-4, IL-10, IL-17, tumor necrosis factor (TNF), macrophage inflammatory protein 1α, and transforming growth factor beta by multiplex enzyme-linked immunosorbent assay (ELISA) (Searchlight; Pierce Biotechnology). Gamma interferon (IFN-γ) in lung lysates was measured with an ELISA kit according to the manufacturer's protocol (R&D Systems).

Ex vivo T-cell restimulation.

Splenocytes were cultured for 48 h at 37°C and 5% CO₂ in medium with 4 μg/ml concanavalin A (Sigma-Aldrich) or 2 μg/ml M. tuberculosis Erdman CFP in 24-well plates with 2 × 10⁶ cells per well in 2 ml. After 48 h, the contents of each well were sterile filtered and assayed for IFN-γ by ELISA.

In vivo antigen presentation.

Leukocytes from the spleen and lymph nodes (LN) of OT-II mice were labeled with 2 μM carboxyfluorescein diacetate succinimidyl ester (CFSE), their viability was confirmed by trypan blue dye exclusion, CFSE labeling was confirmed by fluorescence microscopy, and ∼15 million cells were injected intravenously into recipient mice. Twenty-four hours later, the recipient mice were injected subcutaneously (s.c.) near the left inguinal LN with 100 μl of OVA-coated iron beads (0.1 μg/μl) or with uncoated iron beads as a negative control; 72 h later, leukocytes were isolated from the left inguinal LN. The cells were stained with peridinin chlorophyll α-conjugated anti-CD4 (clone RM4-5) and PE-conjugated anti-CD45.1 (Ly-5.1; clone A20), and 200,000 leukocyte-gated events were collected on a LSRII flow cytometer and analyzed as described above.

Statistical analysis.

An F test was performed to confirm that variances between two groups were not statistically significant. Student's t test for samples with equal variance or unequal variance was performed as appropriate for comparisons between two groups. Comparison between more than two groups was done by analysis of variance, and a Tukey-Kramer multiple-comparison posttest was performed when analysis of variance results were significant. All tests were performed with GraphPad Instat (version 3.05 for Windows 95; GraphPad Software). P values of <0.05 were considered significant.

Hypercholesterolemia increases the TB susceptibility of ApoE^−/− mice.

Mice were fed LC or HC diets in an effort to correlate TB resistance with different levels of serum cholesterol. The mean serum cholesterol for WT LC, WT HC, ApoE^−/− LC, and ApoE^−/− HC mice are shown in Fig. 1A. Serum total cholesterol of >200 mg/dl indicated hypercholesterolemia. The mice were maintained on these diets for 2 weeks before and during an experiment. Serum cholesterol measurements taken before infection and at the conclusion of experiments did not significantly differ within treatment groups (n = 4 or 5) (data not shown). WT and ApoE^−/− mice on LC or HC diets were infected with ∼50 CFU of M. tuberculosis Erdman, and survival was monitored for 3 months (Fig. 1B). All five ApoE^−/− HC mice in this experiment died within 4 weeks after M. tuberculosis infection. By 3 weeks p.i., they appeared hunched, lethargic, poorly groomed, and visibly smaller than their WT LC, WT HC, and ApoE^−/− LC counterparts. On gross inspection, their lungs had large abscess-like masses with extensive inflammation and tissue necrosis visible by histology (Fig. 1C and D). All of the WT LC, WT HC, and ApoE^−/− LC mice survived until the conclusion of the experiment 3 months p.i. and showed no overt signs of illness.

WT and ApoE^−/− HC mice had comparable lung bacterial burdens up to 20 days p.i., suggesting that a defect of innate immunity or an enhanced intrinsic rate of M. tuberculosis replication was not responsible for the TB susceptibility of ApoE^−/− HC mice. By 30 day p.i., the bacterial lung burden reached a plateau of 6.5 log units CFU in WT HC mice while it was still increasing exponentially in ApoE^−/− HC mice, reaching 8.5 log units at the time of death (Fig. 2A). The bacterial lung burden of ApoE^−/− LC mice was ∼0.5 log CFU greater than that of WT LC mice 30 day p.i. (P < 0.05; n = 4) (data not shown). Since control of M. tuberculosis growth in the lungs requires an effective adaptive immune response, these results suggested that adaptive immunity was severely compromised in ApoE^−/− HC mice.

Increased lung inflammation in ApoE^−/− mice with TB.

The pattern and kinetics of lung leukocyte recruitment were investigated in M. tuberculosis-infected ApoE^−/− HC and WT HC mice. Lung leukocytes were isolated from the left lung lobe and the right caudal lobe by enzymatic digestion, and T-cell (CD3⁺ CD4⁺/CD8⁺), macrophage/monocyte (F4/80⁺), and granulocyte (GR-1⁺ CD3⁻ F4/80⁻) populations were measured by flow cytometry. Signs of increased lung inflammation were detectable in ApoE^−/− HC mice as early as 10 days p.i. and involved a massive influx of granulocytes (Fig. 2B and C). Roughly 45% of the lung leukocytes of ApoE^−/− HC mice at this time point were granulocytes versus <10% for WT HC mice. Most of the granulocytes appeared to be neutrophils, with few basophils or eosinophils visible in H&E-stained lung tissue sections. Despite the massive early influx of neutrophils into the lungs of ApoE^−/− HC mice, extensive lung tissue necrosis and inflammation were not evident by histology until after 20 days p.i. While ApoE^−/− HC lungs had more T cells and macrophages than those of WT HC mice by 30 days p.i., they comprised a smaller proportion of the total lung leukocyte population than in WT HC mice. Uninfected ApoE^−/− HC mice also had a higher proportion of resident lung granulocytes (24% versus 14%) but a lower proportion of T cells (3% versus 7%) than WT HC mice (P < 0.05; n = 5). The exaggerated innate immune response and relatively low T-cell numbers despite the presence of a high antigen burden served as additional indicators that adaptive immunity was impaired in ApoE^−/− HC mice. Differences between WT LC, WT HC, and ApoE^−/− LC mice were not as dramatic as those observed with ApoE^−/− HC mice; therefore, these mice were examined 3 months p.i. The degree of lung inflammation after M. tuberculosis infection coincided with increasing serum cholesterol, as ApoE^−/− LC mice had a modest but statistically significant increase in lung leukocytes compared with WT LC and WT HC mice after 3 months of TB disease (Fig. 3A). There was no difference in the proportions of T-cell, macrophage/monocyte, and granulocyte populations between WT LC, WT HC, and ApoE^−/− LC mice 3 months p.i. (Fig. 3B). However, ApoE^−/− LC mice had significantly more T cells and granulocytes than WT LC and WT HC mice (Fig. 3B).

Lung cytokine production of WT and ApoE^−/− mice with TB.

A Th1 cell-mediated immune response is required for protection against TB, and it is well established that IFN-γ is essential for resistance to TB in mice and humans (6, 10, 18). We assayed lung lysates from WT and ApoE^−/− mice on LC or HC chow for Th1, Th2, and proinflammatory cytokines. The mean IFN-γ levels in the lungs of WT LC, WT HC, and ApoE^−/− LC mice were 268 pg/ml, 253 pg/ml, and 166 mg/ml, respectively, 3 months p.i. Although ApoE^−/− LC mice had significantly lower IFN-γ production than WT LC and HC mice (P < 0.05; n = 4), it is unclear whether this small difference accounts for the fivefold-higher bacterial lung burden of ApoE^−/− LC mice than WT LC mice. Interestingly, there was no difference in lung IFN-γ production between WT HC and ApoE^−/− HC mice 20 days and 30 days p.i. (Fig. 4A). This raised the possibility that the lung macrophages of ApoE^−/− HC mice might have had a reduced capacity to respond to IFN-γ, thereby reducing their ability to kill internalized bacilli. Stimulation of macrophages to produce iNOS is one of the most important antimycobacterial functions of IFN-γ (9, 27, 12). To assess IFN-γ responsiveness, immunohistochemical staining for iNOS was performed on lung tissue sections from WT HC and ApoE^−/− HC mice 10 days, 20 days, and 30 days p.i. (Fig. 4B). Abundant iNOS production was detected in the TB lung lesions of ApoE^−/− HC mice, indicating that their macrophages retained the capacity to respond to IFN-γ.

Levels of Th1, Th2, and proinflammatory cytokines in lung homogenates were similar between WT LC, WT HC, and ApoE^−/− LC mice 3 months p.i. (data not shown). Th1 and Th2 cytokine production levels were also comparable in WT HC and ApoE^−/− HC lungs, and neither group had detectable production of transforming growth factor beta 20 days p.i. Production of the inflammatory cytokines IL-1β, macrophage inflammatory protein 1α, and TNF increased dramatically in the lungs of ApoE^−/− HC mice between 20 days and 30 days p.i. (Fig. 4A). ApoE^−/− HC mice lost ∼7% of their preinfection body weight, consistent with the elevated production of TNF, while WT HC mouse preinfection body weight increased ∼14% by 30 days p.i. (data not shown). Interestingly, IL-17 production levels were similar in WT HC and ApoE^−/− HC mice. IL-17 is associated with neutrophilic inflammation (16) but appears not to play a significant role in the inflammation seen in the lungs of M. tuberculosis Erdman-infected ApoE^−/− HC mice. The overproduction of inflammatory cytokines in ApoE^−/− HC lungs was consistent with the histology and lung leukocyte data and presumably reflected the persistent stimulation of innate immunity in the absence of an effective adaptive immune response.

Response to ex vivo antigen stimulation by WT and ApoE^−/− mice with TB.

Any delay in the adaptive immune response to low-dose aerosol M. tuberculosis challenge in mice results in a higher bacterial lung burden before bacterial growth is restricted. To monitor the kinetics of the adaptive immune response to M. tuberculosis in WT HC and ApoE^−/− HC mice, splenocytes were stimulated ex vivo for 48 h with M. tuberculosis CFP, and IFN-γ release was measured by ELISA. Splenocytes incubated with medium or concanavalin A were used as negative and positive controls, respectively. A strong IFN-γ response by T cells from WT HC mice to ex vivo CFP stimulation was seen as early as 20 days p.i., while the response by T cells from ApoE^−/− HC mice was weak at this time point (Fig. 5A). By 30 days p.i., the CFP-stimulated IFN-γ response of splenocytes from ApoE^−/− HC mice was comparable to that of WT HC mice. The spleen bacterial burdens were comparable between WT HC and ApoE^−/− HC mice 30 days p.i., 4.7 log units and 4.9 log units, respectively (n = 4), indicating that differences in IFN-γ responses were not caused by unequal antigen loads. The number of splenocytes in WT HC mice increased during M. tuberculosis infection but remained unchanged in ApoE^−/− HC mice (Fig. 5B). ApoE^−/− HC mice had more granulocytes in the spleen than WT HC mice despite having fewer total splenocytes (Fig. 5C). At 20 days and 30 days p.i., WT HC mice had almost fivefold more T cells than ApoE^−/− HC mice. Taken together, these results indicate that by 30 days p.i., IFN-γ production on a per-T-cell basis was actually higher in the spleens of ApoE^−/− HC mice than in WT HC mice. Uninfected WT HC and ApoE^−/− HC mice had similar numbers of splenocytes, but ApoE^−/− HC mice had slightly more granulocytes (data not shown). T cells from WT LC and ApoE^−/− LC mice 3 months p.i. produced equivalent amounts of IFN-γ after ex vivo stimulation with CFP (data not shown). WT LC and ApoE^−/− LC mice had comparable numbers of splenocytes with no differences in the proportion or number of T cells, macrophages/monocytes, or granulocytes (data not shown).

In vivo antigen presentation by WT and ApoE^−/− mice.

The M. tuberculosis-specific T-cell response of ApoE^−/− HC mice was delayed compared to that of WT HC mice, suggesting that inefficient priming of the adaptive immune response might be responsible for their increased TB susceptibility. To test this possibility, an adoptive-transfer experiment was done using leukocytes from OT-II T-cell receptor (TCR) transgenic mice. These mice express a TCR that recognizes OVA peptide (amino acids 323 to 339) in the context of a major histocompatibility complex class II I-A^b molecule (3). Their cells were distinguishable from WT and ApoE^−/− leukocytes by the allelic marker CD45.1. CFSE-labeled OT-II leukocytes were injected intravenously into WT HC and ApoE^−/− HC mice, and the next day, OVA-coated or uncoated iron beads were injected s.c. near the left inguinal LN. After 3 days, leukocytes were harvested from the left inguinal LN, and the proliferative response of OT-II T cells was measured by CFSE dilution. OT-II T cells from WT HC mice injected with OVA-coated iron beads underwent about five rounds of cell division, while significantly less proliferation of OT-II T cells was seen in identically treated ApoE^−/− HC mice (Fig. 6). No proliferation of OT-II T cells was seen in WT HC or ApoE^−/− HC mice injected with uncoated iron beads. The frequency of CD4⁺ CD45.1⁺ OT-II T cells in the inguinal LN of ApoE^−/− HC mice receiving OVA-coated iron beads was comparable to that in WT HC and ApoE^−/− HC mice that received uncoated iron beads (data not shown). This indicated equivalent seeding and survival of OT-II T cells in WT HC and ApoE^−/− HC inguinal LN. These results support the conclusion that priming of the adaptive immune response is significantly impaired in ApoE^−/− HC mice prior to M. tuberculosis infection.