**2. Results**

### *2.1. ASC Attenuates B. abortus Pathogenesis and Is Critical for Host Survival*

To explore canonical and non-canonical inflammasome activation to *Brucella abortus*, we used mice that lack either the ASC adaptor protein (*Asc*−/−*)* or the non-canonical inflammasome-associated caspase, caspase-11 (*Caspase-11*−/−). After intraperitoneally injecting mice with 1 × 10<sup>5</sup> colony forming units (CFUs) of *B. abortus*, we monitored mortality in all mouse groups for a 24-day period (Figure 1A). At Day 7, there was a 26.3% decrease in the survival of the *Asc*−/− mice. In these animals, the weight loss in 5 of the 19 mice exceeded 20%, and several of the *Asc*−/− mice developed clinical parameters associated with disease progression, such as decreased body condition, that required euthanasia (Figure 1B). However, there was no decrease in survival for the wildtype (WT) and *Caspase-11*−/− mice. These mortality data sugges<sup>t</sup> that the canonical inflammasome plays a more critical role in host survival compared to the non-canonical inflammasome and caspase-11.

**Figure 1.** *Asc*−/− and *Caspase-11*−/− mortality and morbidity. *Asc*−/− (*n* = 19), *Caspase-11*−/− (*n* = 19), and C57BL/6 WT (*n* = 26) mice were injected intraperitoneally with 1 × 10<sup>5</sup> *B. abortus* CFUs and assessed daily for excessive weight loss (>20%) warranting euthanasia according to the Institutional Animal Care and Use Committee (IACUC). (**A**) Mortality graph based on (**B**) a morbidity assessment of *Asc*−/<sup>−</sup>, *Capsase-11*−/<sup>−</sup>, and WT mice. All the mice were weighed for a 24-day period, with morbidity warranting euthanasia in *Asc*−/− mice only at Day 7. \* *p* < 0.05, \*\*\*\* *p* < 0.0001.

### *2.2. ASC Contributes to Inflammatory Pathogenesis during B. abortus Infection*

To further elucidate the role of ASC and caspase-11 activation *in vivo*, we conducted a histopathological analysis on the liver and spleen from wildtype, *Asc*−/<sup>−</sup>, and *Caspase-11*−/− mice three days post-infection. Histopathology indicated that all the infected mouse groups exhibited elevated extramedullary hematopoiesis (EMH) and inflammation in the liver and spleen (Figure 2A,B).

**Figure 2.** Inflammation in the liver and spleen. A total of 20 *Asc*−/− (*n* = 6 uninfected [U], 14 infected [I]), 27 *Caspase-11*−/− (*n* = 8 U, 19 I), and 34 C57BL/6 WT (*n* = 11 U, 23 I) livers and spleens were evaluated by histopathology 3 d.p.i. (**A**) H&E stained histological slides of the liver and spleen from WT, *Asc*−/<sup>−</sup>, and *Caspase-11*−/− mice. All spleen images were taken at 4× power and all liver images were taken at 40× power. Inflammation and extramedullary hematopoiesis (EMH) were the dominant features observed in the histopathology evaluation. (**B**) Bar graphs of WT, *Asc*−/<sup>−</sup>, and *Caspase-11*−/− histopathology composite scores were generated based on inflammation and EMH in the liver and spleen. \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, \*\*\*\* *p* < 0.0001.

Wildtype mice exhibited higher EMH and inflammation scores compared to the *Asc*−/− animals in the spleen. This trend was not observed in the *Caspase-11*−/− mice in either the liver or the spleen. Together, these data sugges<sup>t</sup> that ASC augments splenic inflammation and plays a key role in *B. abortus*-mediated pathology in the spleen, which is a target organ in this mouse model. The failure to mount a vigorous immune response to *B. abortus* in the *Asc*−/− mice is likely associated with the increased morbidity and mortality.

### *2.3. Bacterial Load Is Significantly Increased in the Absence of ASC and Decreased in the Absence of Caspase-11*

Bacterial loads were determined in the spleen and liver of *B. abortus*-infected animals 3 d.p.i (Figure 3). Between wildtype, *Asc*−/<sup>−</sup>, and *Caspase-11*−/− mice, there was no significant difference in the weight of the spleens used for analysis.

**Figure 3.** *Brucella* CFUs in the liver and spleen of *Asc*−/− and *Caspase-11*−/− mice. Three d.p.i. liver and spleens from infected C57BL/6 WT (*n* = 23), **(A)** *Asc*−/− (*n* = 14), and **(B)** *Caspase-11*−/− (*n* = 19) mice were homogenized and counted for CFUs/gram. \* *p* < 0.05, \*\* *p* < 0.01.

In *Asc*−/− mice, we observed significantly elevated *B. abortus* CFUs in the liver and a trending increase in the spleen. In the *Caspase-11*−/− mice, we observed a similar trending increase in CFUs in the liver. However, in the spleen *Caspase-11*−/− mice had significantly decreased bacterial CFUs compared to the wildtype. The increased bacteria load in *Asc*−/− mice is consistent with the increased morbidity and reduced inflammation in Figures 1 and 2, and further illustrates the critical role of ASC in the host response to *B. abortus*. Likewise, these *Caspase-11*−/− data sugges<sup>t</sup> a significant, but variable, role in controlling the *B. abortus* bacteria burden that does not appear to impact the overall host morbidity or inflammation.

### *2.4. B. abortus Initiates a Weak Inflammasome-Mediated Inflammatory Cytokine Response*

Inflammasome activation results in the cleavage and processing of IL-1β and IL-18. To evaluate the generation of pro-IL-1β, so-called "Signal 1", at the transcript level, we conducted quantitative real-time PCR in liver and spleen homogenates (Figure 4). We observed significantly increased fold changes in *Il1*β in the livers of *Asc*−/− and wildtype mice infected with *B. abortus* versus the uninfected mice (Figure 4A). Within infected mouse groups in the liver, we also found that infected *Asc*−/− mice had a significantly decreased fold change of *Il1*β compared to wildtype mice (Figure 4A). However, when assessing the total IL-1β (uncleaved and cleaved) protein, we did not see a significant difference between the infected wildtype and *Asc*−/− mice.

**Figure 4.** Inflammatory signaling in the liver and spleen of *Asc*−/− and *Caspase-11*−/− mice. Three d.p.i. livers and spleens of 34 C57BL/6 WT (*n* = 11 U, 23 I), (**A**) 20 *Asc*−/− (*n* = 6 U, 14 I), and (**B**) 27 *Caspase-11*−/− (*n* = 8 U, 19 I) mice were homogenized and analyzed for their *Il1*β RNA fold change using RT-PCR and IL-1β protein concentration through ELISA. \* *p* < 0.05.

In experiments with *Caspase-11*−/− mice, we also found an elevated RNA fold change between the infected and uninfected groups (Figure 4B), but there was no significant difference between the infected wildtype and *Caspase-11*−/− mice for *Il1*β transcription (Figure 4B). However, there was a minimal, but statistically significant, decrease in IL-1β protein in the livers from *Caspase-11*−/− mice compared to the wildtype animals.

### *2.5. B. abortus Infection Attenuated IL-1*β *and Induced a Strong ASC-Dependent Pyroptosis Response in Macrophages*

Due to the significant phenotype observed in the *Asc*−/− mice, we next sought to better define the underlying mechanism using ex vivo bone marrow-derived macrophages (BMDMs). Intracellular bacterial replication and survival was evaluated over a 48 h period in *B. abortus*-infected *Asc*−/− and wildtype BMDMs. Under these conditions, we observed a significant decrease in *B. abortus* growth in the *Asc*−/− macrophages compared with the wildtype BMDMs after 24 h (Figure 5A). By 48 h, the *B. abortus* replication and survival was no longer detectable, while increasing in the wildtype BMDMs. These results were unexpected based on ourin vivo findings and sugges<sup>t</sup> a possible disconnect between pathogen clearance, inflammasome function, and pyroptosis in BMDMs.

We next measured the RNA fold change and protein concentration of IL-1β (Figure 5B,C). Within 2 h of *B. abortus* exposure, we observed elevated *Il1*β transcription in the infected wildtype and *Asc*−/− macrophages. Transcription was statistically significant, but only slightly higher, in the *Asc*−/− cells compared to the wildtype BMDMs (Figure 5B). At 24 h and 48 h, we observed a significant decrease in *Il1*β transcription in both groups of infected mice, with significantly more *Il1*β in the *Asc*−/− BMDMs at 24 h compared to the wildtype (Figure 5B). Complementing the transcription data, we also evaluated the IL-1β protein levels in the cell supernatant using ELISA (Figure 5C). These levels were significantly attenuated in the *B. abortus*-infected cells (Figure 5C). We also observed a significant decrease in IL-1β in the *Asc*−/− BMDMs under all conditions (Figure 5C), emphasizing that IL-1β processing is ASC-dependent.

To further expand upon our cell death findings, we conducted a LDH assay in our *Asc*−/− and wildtype macrophages to quantify the lactate dehydrogenase enzyme released from dead cells (Figure 5D). We found that, starting at 24 h, there was significantly higher cell death occurring in our wildtype cells compared to the *Asc*−/− macrophages. To further define the mechanism of cell death, we evaluated pyroptosis by determining gasdermin D cleavage using Western blot. We found a significant increase in cleaved gasdermin D 24 h post-infection in the wildtype BMDMs compared to the significantly reduced levels observed in the *Asc*−/− macrophages (Figure 5E). This was confirmed using densitometry (Figure 5E).

### *2.6. B. abortus gDNA is a Potent PAMP Associated with ASC-Dependent Canonical Inflammasome Signaling*

To further define inflammasome activation following *B. abortus* infection, we next evaluated the potential pathogen-associated molecular patterns (PAMPs), focusing on bacterial gDNA. We challenged macrophages with 1 μg of gDNA (2 μg/mL) both externally, by adding gDNA to the cell media, and internally through the Lipofectamine 3000 reagen<sup>t</sup> (Figure 6A). We also added 300 μM of ATP to augmen<sup>t</sup> the IL-1β release. *B. abortus* gDNA induced a significant increase in the *Il1*β gene transcription 24 h post challenge, following either extracellular or intracellular challenge (Figure 6A). We observed significant increases in *Il1*β transcription under several different conditions in gDNA-challenged *Asc*−/− macrophages compared to the wildtype cells. ELISA assessments revealed that IL-1β protein was only released into the supernatant following internal gDNA stimulation in wildtype cells (Figure 6B). This was highly dependent on ASC. The *B. abortus* gDNA challenge resulted in significant increases in IL-1β in the wildtype cells, whereas the levels were below the level of detection in the *Asc*−/− cells (Figure 6B). Together, these data confirm *B. abortus* gDNA as a potent PAMP and sugges<sup>t</sup> that its recognition by the canonical inflammasome, in an ASC-dependent mechanism, underlies host defense. In addition to gDNA, we also evaluated the ability of the canonical

inflammasome to recognize *B. abortus* LPS (1 μg/mL externally). The fold change in *Il1*β RNA at 8 h post challenge indicated that there was elevated transcription in wildtype macrophages over *Asc*−/− cells (Figure 6C). However, there was no IL-1β protein signaling for *Brucella* LPS (Figure 6D).

**Figure 5.** ASC-dependent IL-1β generation and pyroptosis following *B. abortus* infection in BMDMs. Bone marrow-derived macrophages (BMDMs) (500,000 cells per well, *n* = 3 per group) were harvested from C57BL/6 WT and *Asc*−/− mice and challenged with a MOI 100:1 (10<sup>7</sup> CFUs) of *Brucella abortus.* (**A**) BMDMs were measured for CFUs for a 48 h period post-challenge. (**B**) BMDMs were measured for the *Il1*β fold change in RNA through RT-PCR. (**C**) BMDMs were analyzed for IL-1β protein in the supernatant using an ELISA. (**D**) Lactate Dehydrogenase (LDH) was measured through spectrophotometry through a 48 h time period post-challenge. (**E**) BMDMs were lysed at 24 h post-challenge and used for Western blot analysis. BMDM protein (20μg) was probed for cleaved gasdermin D and β-actin as a control for the protein amount. Invitrogen IBright Analysis was used to determine the density ratio between the cleaved gasdermin D over β-actin. \* *p* < 0.05, \*\*\* *p* < 0.001, \*\*\*\* *p* < 0.0001.

**Figure 6.** *B. abortus* gDNA is a potent PAMP and induces ASC-dependent IL-1β production. Bone marrow-derived macrophages (BMDMs) (500,000 cells per well, *n* = 2 per group) were harvested from C57BL/6 WT and *Asc*−/− mice. (**A**,**B**) BMDMs were challenged with 1 μg of gDNA (2μg/mL) externally in media or transfected internally with 300 μM of ATP and harvested after 24 h. BMDMs were analyzed for the IL-1β (**A**) RNA fold change through RT-PCR and (**B**) protein concentration through ELISA. (**C**,**D**) BMDMs were challenged with 1 μg/mL of *Brucella* LPS externally to the media with 300 μM of ATP and harvested after 8 h. BMDMs were analyzed for the IL-1β (**C**) RNA fold change through RT-PCR and (**D**) protein concentration through ELISA. Note that there was no detectable (*n*.d.) protein concentration in (**D**). \* *p* < 0.05, \*\*\*\* *p* < 0.0001.
